The Architecture of Manufacturing Networks Integrating the Coordination Perspective
DISSERTATION of the University of St. Gallen, School of Management, Economics, Law, Social Sciences and International Affairs to obtain the title of Doctor of Philosophy in Management
submitted by Andreas Mundt from Germany
Approved on the application of Prof. Dr. Thomas Friedli and Prof. Dr. Elgar Fleisch
Dissertation no. 4097
Difo-Druck GmbH, Bamberg 2012
The University of St. Gallen, School of Management, Economics, Law, Social Sciences and International Affairs hereby consents to the printing of the present dissertation, without hereby expressing any opinion on the views herein expressed. St. Gallen, October 29, 2012
The President:
Prof. Dr. Thomas Bieger
Für meine Eltern
VORWORT Die vorliegende Dissertation entstand während meiner Tätigkeit als wissenschaftlicher Mitarbeiter am Lehrstuhl für Produktionsmanagement des Instituts für Technologiemanagement der Universität St.Gallen. Von 2009 bis 2012 hatte ich die Möglichkeit, in zahlreichen Industrie- und Forschungsprojekten Einblicke in die Managementpraxis produzierender Unternehmen zu erhalten und in ihrem Umfeld gestaltend tätig zu werden. Die dort gesammelten Erkenntnisse bilden das Fundament dieser Arbeit; ich bin mir jedoch sicher, noch lange und weit darüber hinaus von dem Erfahrungsschatz dieser Jahre zehren und profitieren zu können. Allen, die dazu beigetragen und diese Zeit zu einer unvergesslichen gemacht haben, gilt mein Dank. Besonders hervorheben möchte ich meinen Referenten Prof. Dr. Thomas Friedli, der durch seinen offenen Führungsstil und das unerschütterliche Vertrauen in seine Assistenten nicht nur das Gelingen dieser Arbeit ermöglichte, sondern den Forschungsansatz der „alten“ St.Galler Schule konsequent lebt und in dessen Sinne Freiräume für wertvolle Einblicke in den unternehmerischen Alltag geschaffen hat. Ebenfalls danke ich Prof. Dr. Elgar Fleisch für die Übernahme des Korreferats. Betonen möchte ich zudem das kollegiale und inspirierende Umfeld am Lehrstuhl, welches mir stets intellektuellen Anreiz bot. Meine Assistentenzeit war geprägt durch den Umgang mit verschiedensten Charakteren, die alle auf ihre Art bereichernd waren. Hervorheben möchte ich Stefan Thomas, durch dessen Kreativität und konzeptionelles Verständnis ein neuer Blick auf das Management globaler Produktionsnetzwerke erst möglich wurde, Georg Oschmann, dessen motivierendes Wesen in jeder Hinsicht eine Unterstützung war, sowie Reto Ziegler, Simone Heinzen, Saskia Gütter, Matthias Götzfried, Richard Lützner, Jakob Ebeling, Daniel Bellm, Fabian Liebetrau, Lukas Budde, Carolin Ubieto und Dr. Maike Scherrer. Im Weiteren danke ich Iris Holzleitner für ihre akribische Lektoratsarbeit. Mein letzter und besonderer Dank richtet sich an mein engstes privates Umfeld: an meine Mutter Inge Mundt und an meine Partnerin Renate Policzer. Eure Geduld und euer steter Zuspruch haben mir Kraft und das nötige Durchhaltevermögen gegeben. Ich hoffe, noch lange darauf zurückgreifen zu dürfen.
St. Gallen im Oktober 2012
Andreas Mundt
KURZZUSAMMENFASSUNG Das Konstrukt „Produktionsnetzwerk“ begründet einen systemtheoretischen Blickwinkel auf die global verteilten Aktivitäten produzierender Unternehmen. Geht es darum, diese als Wettbewerbsvorteil zu nutzen und damit das volle Potential des eigenen Netzwerks zu realisieren, scheitern in der Praxis jedoch viele. Gründe hierfür liegen in einem mangelnden Gesamtverständnis bzgl. der Eigenschaften solcher Netzwerke sowie in fehlenden Gestaltungs- und Managementansätzen. Auch die bestehende Forschung greift hierbei zu kurz. Abgesehen von der Ausgestaltung der physischen Materialflüsse liefert sie nur bedingt Ansätze zur Analyse, Entwicklung und Optimierung eines globalen Produktionsverbunds. Insbesondere aus Perspektive des Operations Managements wird eine aggregierte Sichtweise auf die einzelnen Standorte, auf ihre Kompetenzen, strategischen Vorteile und individuellen Beiträge für das Gesamtnetzwerk, aber auch die Berücksichtigung von Koordinationsmechanismen zur Organisation ihres Zusammenspiels nur gestreift. Vorliegende Arbeit trägt zur Schliessung dieser Lücke bei. Sie liefert Managern Unterstützung in der strategischen und konzeptionellen Ausgestaltung ihres Produktionsnetzwerks; nicht nur bzgl. dessen Konfiguration, sondern vor allem bzgl. der zentralen Koordinationsentscheidungen. Hierzu werden die Entscheidungsdimensionen der Netzwerkkoordination aus der Literatur isoliert und praxisseitig validiert. Die einzelnen Dimensionen werden durch sogenannte Managementframeworks operationalisiert. Diese Frameworks sind als Werkzeuge zu verstehen, um strategische Optionen für das Design von Koordinationsmechanismen im Netzwerk zu stimulieren und konzeptionell auszugestalten. Die Netzwerkkoordination wird zudem mit der Netzwerkkonfiguration verknüpft; mit der Organisationsstruktur, der Netzwerkstruktur und der Netzwerkspezialisierung. Letztere wird durch einen innovativen Ansatz für die Ableitung eines Standortrollenportfolios erweitert, welcher es erlaubt, die Rollen der Standorte im Netzwerk systematisch aufzubauen und abzubilden. Dabei wird argumentiert, wie konfigurative Änderungen des Portfolios die Koordination beeinflussen. Das Standortrollenportfolio und die einzelnen Koordinationsframeworks werden in einer integralen Netzwerkarchitektur verankert, welche eine ganzheitliche Sichtweise auf das Produktionsnetzwerk aus Perspektive des übergeordneten Netzwerkmanagements begründet. Abschliessend wird die Architektur in einen „Suggested Practice Ansatz“ für das Netzwerkdesign und Management überführt. Im Sinne eines diskursiven Vorgehens adressiert dieser Ansatz die Kernelemente zur Analyse von Produktionsnetzwerken und die strategischen Entscheidungen für deren konzeptionelle (Neu-)Gestaltung.
SUMMARY A manufacturing network is a valuable construct that constitutes a systemic view on companies’ globally dispersed operations. However, many multinationals fail in leveraging their networks’ full potential, lacking both a holistic understanding and systematic design and management approach. Likewise, except for the design of material flows, research is poor in providing adequate solutions to analyse, develop and improve manufacturing networks. Especially from an operations management perspective, an aggregated view on the network’s plants, their competencies, strategic reasons, and idiosyncratic contributions, as well as an elaboration of the coordination layer in terms of how to organise the interplay between the scattered plants has been addressed very simplified. This study seeks to overcome these limitations, supporting today’s operations leaders in the conceptual and strategic design and management of their intra-company manufacturing networks; not only from a configuration perspective, but, in particular, by integrating the coordination layer. First, the central coordination decision dimensions that network managers face are isolated from operations management literature and validated by practical discussion. Second, each coordination decision dimension is operationalised by a distinct management framework. The frameworks serve as tools stimulating conceptual thinking and the derivation of strategic options for the design of coordination mechanisms. Third, the coordination layer is put in a wider context, linking it with the network configuration, i.e., its organisation, structure, and specialisation. A novel approach for the design of a plant role portfolio is promoted – a framework to systematically create and keep track of the strategic roles the sites play in the network. It is further demonstrated how changes in the configuration of the plant role portfolio affect the network coordination. Fourth, a holistic network management architecture is derived, anchoring the plant role portfolio and the single coordination frameworks. It provides an integral view on manufacturing networks from the operations manager’s superordinate perspective. Finally, the architecture is transformed into a “suggested practice approach” for strategic network design and management. Rather than with a restrictive process relying on prescriptive steps, operations managers are equipped with a discursive approach focusing on the main elements to analyse, and the central decisions to make when conceptually (re-)designing their network.
TABLE OF CONTENTS
V
TABLE OF CONTENTS TABLE OF CONTENTS ........................................................................................................ V LIST OF FIGURES ............................................................................................................VIII LIST OF TABLES .................................................................................................................. X LIST OF ABBREVIATIONS ................................................................................................ XI 1
INTRODUCTION ............................................................................................................ 1 1.1 Motivation & Relevance ................................................................................................ 1 1.1.1 Research Motivation .............................................................................................. 1 1.1.2 Practical Relevance ................................................................................................ 3 1.1.3 Theoretical Gaps .................................................................................................... 5 1.2 Research Foundation & Question .................................................................................. 6 1.2.1 Research Background & Foundation ..................................................................... 6 1.2.2 Research Question .................................................................................................. 8 1.3 Research Methodology & Design .................................................................................. 8 1.3.1 Research Grounding & Process.............................................................................. 8 1.3.2 Research Methodology, Theory Building, Data Collection & Analysis .............. 10 1.4 Study Structure ............................................................................................................. 15
2
UNDERSTANDING MANUFACTURING NETWORKS ........................................ 17 2.1 Analysing the Knowledge Base on Manufacturing Networks ..................................... 18 2.1.1 Structuring the Knowledge Base .......................................................................... 18 2.1.2 Reviewing the Knowledge Base .......................................................................... 20 2.2 Manufacturing Site vs. Network Perspective ............................................................... 25 2.2.1 Manufacturing Site Perspective ........................................................................... 25 2.2.2 Manufacturing Network Perspective .................................................................... 27 2.2.3 Discussing the Site & Network Perspectives ....................................................... 27 2.3 Decision Categories & Layers of Manufacturing Networks ........................................ 29 2.3.1 Manufacturing (Site & Network) Strategy ........................................................... 29 2.3.2 Structural & Infrastructural Decision Categories ................................................. 33 2.3.3 Manufacturing Network Configuration Layer ..................................................... 34 2.3.4 Manufacturing Network Coordination Layer ....................................................... 38 2.3.5 Discussing the Decision Categories & Layers ..................................................... 40 2.4 Network Design & Management Approaches ............................................................. 42 2.4.1 Network Management Frameworks ..................................................................... 42 2.4.2 Network Design, Management & Optimisation Approaches............................... 45 2.4.3 Discussing the Design, Management & Optimisation Approaches ..................... 52
VI
TABLE OF CONTENTS
2.5 Summary & Discussion ................................................................................................ 53 2.5.1 Implications from Literature ................................................................................ 53 2.5.2 Refinement of Research Question ........................................................................ 55 2.5.3 Derivation of Heuristic Research Framework ...................................................... 56 3
DESIGNING THE NETWORK COORDINATION LAYER ................................... 58 3.1 Defining the Decision Dimensions of Network Coordination ..................................... 58 3.1.1 Defining Coordination.......................................................................................... 58 3.1.2 Identifying the Network Coordination Decision Dimensions .............................. 60 3.2 From Decision Dimensions to Coordination Frameworks ........................................... 64 3.2.1 Methodical Approach ........................................................................................... 64 3.2.2 Case Study Selection & Outline ........................................................................... 65 3.2.3 Survey & Interview Outline ................................................................................. 67 3.3 Developing the Coordination Frameworks .................................................................. 71 3.3.1 The Centralisation & Standardisation Framework ............................................... 71 3.3.2 The Resource Sharing Framework ....................................................................... 83 3.3.3 The Incentive System Framework........................................................................ 91 3.3.4 The Information & Knowledge Sharing Framework ......................................... 101 3.4 Summary & Discussion .............................................................................................. 113 3.4.1 Findings from the Design of the Coordination Layer ........................................ 113 3.4.2 Implications for a Network Management Architecture...................................... 116
4
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE ....................................................................................................... 118 4.1 Introducing an Integral Architecture for Network Management ............................... 118 4.1.1 Elaborating the Network Configuration Layer................................................... 118 4.1.2 Transforming the Research Framework into a Management Architecture ........ 127 4.2 Working with the Network Management Architecture .............................................. 129 4.2.1 Case Study Selection & Outline ......................................................................... 129 4.2.2 The “Elevator NW” Case ................................................................................... 131 4.2.2.1 Network Analysis & Target Setting ........................................................... 131 4.2.2.2 Scenario Development ............................................................................... 146 4.2.2.3 Conceptual Network (Re-)Design .............................................................. 147 4.2.3 The “Chocolate NW” Case ................................................................................ 158 4.2.3.1 Network Analysis & Target Setting ........................................................... 158 4.2.3.2 Scenario Development ............................................................................... 163 4.2.3.3 Conceptual Network (Re-)Design .............................................................. 168 4.3 Summary & Discussion .............................................................................................. 169 4.3.1 Findings from the Application of the Management Architecture ...................... 169
TABLE OF CONTENTS
4.3.2 5
VII
Implications for a Strategic Design & Management Approach ......................... 172
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH ............................................................. 174 5.1 Presenting a Strategic Network Design & Management Approach ........................... 174 5.2 Summary & Discussion .............................................................................................. 181
6
SUMMARY & OUTLOOK......................................................................................... 185 6.1 Critical Reflection ...................................................................................................... 185 6.2 Contribution to Theory & Practice ............................................................................. 187 6.3 Limitations & Further Research ................................................................................. 189
REFERENCES ..................................................................................................................... 192 APPENDIX A: CASE STUDIES & INTERVIEWS ......................................................... 208 APPENDIX B: QUESTIONNAIRE ................................................................................... 209 B.1 B.2 B.3
Survey Questionnaire (Excerpt of Questions) ........................................................ 209 Measures for the Calculation of the Network Capability Level & Conformance .. 220 Measures for the Calculation of the Performance Level on Strategic Manufacturing Priorities ......................................................................... 221
CURRICULUM VITAE ...................................................................................................... 222
VIII
LIST OF FIGURES
LIST OF FIGURES Fig. 1: Network capability level vs. performance level on strategic manufacturing priorities .. 4 Fig. 2: Reasons for reallocating foreign facilities ...................................................................... 5 Fig. 3: X-GO Model ................................................................................................................... 6 Fig. 4: Forms of FIT ................................................................................................................... 7 Fig. 5: Research process ............................................................................................................. 9 Fig. 6: Research stages ............................................................................................................. 11 Fig. 7: Study structure .............................................................................................................. 16 Fig. 8: Literature analysis approach ......................................................................................... 17 Fig. 9: Classification of value networks ................................................................................... 19 Fig. 10: Literature screening results ......................................................................................... 22 Fig. 11: The roles of foreign factories ...................................................................................... 26 Fig. 12: Perspectives on value networks .................................................................................. 27 Fig. 13: Factory-network capability matrix.............................................................................. 32 Fig. 14: Overview of the interactive evolutions of plants & the manufacturing network........ 37 Fig. 15: Framework for international manufacturing network configuration & capability profile grid ................................................................................................. 43 Fig. 16: Manufacturing strategy framework for a manufacturing network.............................. 44 Fig. 17: Heuristic research framework ..................................................................................... 57 Fig. 18: General information about the X-GO survey sample ................................................. 67 Fig. 19: Network characteristics of the X-GO survey sample ................................................. 68 Fig. 20: Network capability level & conformance vs. performance level on strategic manufacturing priorities.......................................................................... 69 Fig. 21: Centralisation & standardisation framework .............................................................. 71 Fig. 22: Resource allocation & sharing framework ................................................................. 83 Fig. 23: Incentive system framework ....................................................................................... 91 Fig. 24: Performance categories & reward types ..................................................................... 97 Fig. 25: Information & knowledge sharing framework ......................................................... 101 Fig. 26: Production know-how & primary transfer mechanisms ........................................... 104 Fig. 27: “Mountain model” .................................................................................................... 120 Fig. 28: Plant role portfolio & “site type matrix” .................................................................. 122 Fig. 29: History & principles of organisation structures ........................................................ 125 Fig. 30: Network management architecture ........................................................................... 127 Fig. 31: Selection of the networks for the in-depth case studies ............................................ 130 Fig. 32: Elevator NW: Network structure .............................................................................. 132 Fig. 33: Elevator NW: Example of a supplier profile ............................................................ 135 Fig. 34: Elevator NW: “Site type matrix” per technology line (AS-IS) ................................ 136
LIST OF FIGURES
IX
Fig. 35: Elevator NW: Plant role portfolio per technology line (AS-IS) ............................... 137 Fig. 36: Elevator NW: Centralisation & standardisation framework (AS-IS) ....................... 139 Fig. 37: Elevator NW: Resource allocation & sharing framework (AS-IS) .......................... 142 Fig. 38: Elevator NW: Information & knowledge sharing framework (AS-IS) .................... 143 Fig. 39: Elevator NW: Network capability profile (TO-BE) ................................................. 146 Fig. 40: Elevator NW: Plant role portfolio for the Sheet Metal technology line (TO-BE) .... 147 Fig. 41: Elevator NW: Lead factory authority for the Sheet Metal technology line (TO-BE) .......................................................................................... 149 Fig. 42: Elevator NW: Plant role portfolio per technology line (TO-BE) ............................. 150 Fig. 43: Elevator NW: Centralisation & standardisation framework (TO-BE) ..................... 153 Fig. 44: Elevator NW: Resource allocation & sharing framework (TO-BE)......................... 154 Fig. 45: Elevator NW: Information & knowledge sharing framework - internal (TO-BE) ... 155 Fig. 46: Elevator NW: Information & knowledge sharing framework - external (TO-BE) .. 156 Fig. 47: Chocolate NW: Network capability profile (AS-IS vs. TO-BE) .............................. 162 Fig. 48: Evaluation criteria ..................................................................................................... 169 Fig. 49: Aspects of striving for FIT........................................................................................ 173 Fig. 50: “Suggested practice approach” for strategic network design & management .......... 175
X
LIST OF TABLES
LIST OF TABLES Tab. 1: Overview of the empirical data base ............................................................................ 14 Tab. 2: Literature screening set-up ........................................................................................... 21 Tab. 3: Literature overview ...................................................................................................... 24 Tab. 4: Strategic network capabilities ...................................................................................... 31 Tab. 5: Manufacturing system’s decision categories & dimensions / sub-systems ................. 33 Tab. 6: Generic network types & multiplant strategies ............................................................ 36 Tab. 7: Overview of selected network design, management & optimisation approaches ....... 51 Tab. 8: Coordination mechanisms in multinational corporations ............................................ 60 Tab. 9: Definitions & decision dimensions of manufacturing network coordination .............. 62 Tab. 10: Outline of case study networks used in designing the coordination frameworks ...... 66 Tab. 11: Outline of interviewed networks & corresponding network categories .................... 70 Tab. 12: Responsibility areas & categories .............................................................................. 73 Tab. 13: Findings on the centralisation & standardisation framework .................................... 82 Tab. 14: Findings on the resource allocation & sharing framework ........................................ 90 Tab. 15: Findings on the incentive system framework .......................................................... 100 Tab. 16: Information & knowledge categories....................................................................... 103 Tab. 17: Findings on the information & knowledge sharing framework ............................... 112 Tab. 18: Operationalisation of the coordination decision dimensions ................................... 113 Tab. 19: Network coordination vs. coordination from an organisation theory’s perspective................................................................................................. 115 Tab. 20: Elevator NW: Network characteristics .................................................................... 134 Tab. 21: Elevator NW: Process landscape ............................................................................. 134 Tab. 22: Chocolate NW: Network characteristics .................................................................. 159 Tab. 23: Chocolate NW: Scenario development via tension lines ......................................... 166 Tab. 24: Chocolate NW: Scenario assessment ....................................................................... 167 Tab. 25: Comparison of the “suggested practice approach” (part I) ...................................... 181 Tab. 26: Comparison of the “suggested practice approach” (part II)..................................... 182 Tab. 27: Results & their contribution to theory & practice .................................................... 188
XI
LIST OF ABBREVIATIONS
LIST OF ABBREVIATIONS BU
Business Unit
ca.
Circa
CAD
Computer Aided Design
EBIT
Earnings before Interest and Taxes
ed.
Edition
Ed.
Editor
e.g.
Exempli gratia (for example)
ERP
Enterprise Resource Planning
et al.
Et alii
etc.
Et cetera
Ex.
Example
GDP
Gross Domestic Product
HR
Human Resource
HSG
University of St.Gallen (Universität St.Gallen)
i.e.
Id est (that is)
IfM
Institute for Manufacturing (University of Cambridge)
Imp.
Implication
ISIC
International Standard Industrial Classification
IT
Information Technology
ITEM-HSG
Institute for Technology Management (University of St.Gallen)
KPI(s)
Key Performance Indicator(s)
MNC
Multinational Corporations
n.a. (n/a)
Not available
No.
Number
NW
Network
p. / pp.
Page / Pages
p.a.
Per annum
PARTS
Players, Added Value, Rules, Tactics, Scope
Q.
Question
XII
LIST OF ABBREVIATIONS
R&D
Research & Development
S&OP
Sales & Operations Planning
SC
Supply Chain
SCM
Supply Chain Management
SP.
Supplier (in the Elevator NW case study)
tbd
To be defined
TECTEM
Transfer Center for Technology Management of the Institute of Technology Management at the University of St.Gallen
X-GO
Excellence in Global Operations (research project and survey)
INTRODUCTION
1
1 Introduction 1.1 Motivation & Relevance 1.1.1 Research Motivation Streamlining global operations unlocks new opportunities for international manufacturers in reaching competitive superiority. Effectively exploiting the heterogeneous market and resource advantages at a company’s individual plants and managing their interlinked worldwide manufacturing activities becomes today’s imperative. For this, the concept of manufacturing networks to structure and study globally dispersed operations provides a valuable construct. Such networks are understood as scattered factories with matrix connections where the single factories are affecting each other, and thus cannot be managed in isolation (Shi and Gregory, 1998; Rudberg and Olhager, 2003). They have been addressed from different perspectives: While some research primarily focuses on location decisions (e.g., Schill, 1990; Kinkel, 2004) or the strategic roles of single plants (e.g., Ferdows, 1997a; Vereecke et al., 2006), other considers the network as holistic system (e.g., Feldmann et al., 2010; Cheng et al., 2011). Nonetheless, many companies intending to design and improve this system fail in fully leveraging its capabilities. Newer surveys estimate a cost-saving potential of streamlined manufacturing networks of up to 45% of manufacturing costs while most companies struggle with realising less than 10% (Jacob and Strube, 2008). One reason for this might be the lack of support and guidance for a systematic network management. Traditionally, operations management has focused on optimising a single plant’s or even the shop floor’s microcosm. The concepts of lean manufacturing and operational excellence, for instance, postulate a philosophy of waste reduction and continuous improvement typically embodied in the toolkit of a company’s production system; and literature and practice is rich in giving examples of how applying these toolkits boosts a production site’s performance. Yet, it is obviously more than just the global roll-out of a production system that drives a well-functioning network. Instead, the uncoordinated striving for local optima actually amplifies the risk of missing the global optimum. Accordingly, operations leaders require a common language, tailored tools, methods, and frameworks for analysing, designing, and optimising their manufacturing networks – embedded into a holistic system perspective and management paradigm; or according to Cheng et al. (2011): “… managers need analytical views and contingent thinking on relevant factors of the plant, the network, and even the company in order to make
2
INTRODUCTION
relevant decisions about the transformation of the manufacturing network (…) specific persons who can proactively coordinate the network’s nodes and flows are needed at the company’s headquarters. They must take all factors into consideration, integrate them as a whole, audit them frequently, distinguish or predict their every (possible) change, address the implications of their interactions, and then design product/process flows carefully in order to improve the whole system proactively and continuously among the plants within the same network” (Cheng et al., 2011, p. 1328). Quantitative methods, such as mathematical programming and simulation that support product allocation decisions and material flow optimisation, are a first step towards a network perspective. However, their benefit is limited as long as an overall strategic view is missing: Considering their depth of detail, there is a risk of shifting the scope from essential strategic questions to an overemphasis on time and cost aspects, making these tools beneficial for the evaluation of network design options, but not for their creation. Strategic approaches providing discursive guidance to the previous creation phase of these options can bridge this gap. Nonetheless, research on such approaches has been rare (e.g., Shi and Gregory, 1998; Grallert et al., 2010). Moreover, among the few approaches promoted in research and practice, most have a clear focus on the network's configuration, i.e., the design of the manufacturing footprint in terms of the sites’ global dispersion, their location, and the shape of the internal supply chain. The network coordination, i.e., the organisation and steering of the interplay between the single sites and between the site and the central management level, has been either neglected or addressed very simplified. 1 Thus, although coordination is widely accepted as key managerial task (e.g., Porter, 1986; Shi, 2003; Cerrato, 2006), the statement of Pontrandolfo and Okogbaa (1999) that “… less attention has been devoted to the coordination issues and seldom have studies (…) addressed the coordination of manufacturing activities using any meaningful systematic approaches” (Pontrandolfo and Okogbaa, 1999, p. 5) seems still valid. This study aims at tackling these drawbacks. It strives to support today’s operations leaders in the strategic design and management of their intra-company manufacturing networks; not only from a configuration perspective but also by explicitly integrating the coordination layer. To ensure operational feasibility, the promotion of a network architecture and strategic design and management approach that anchors both layers is aspired, too. 1
A literature screening of 674 papers in the field of operations management between 1986 and 2011 revealed 18 papers on "coordination" vs. 64 papers with "configuration" focus. These findings are supported by the literature analysis of De Toni and Parussini (2010) identifying 65 papers addressing manufacturing networks; 17 of those cover coordination and knowledge exchange, and 40 concentrate on configuration issues.
INTRODUCTION
3
1.1.2 Practical Relevance There is little to add about the importance of global manufacturing. Economic studies, such as the World Investment Report of the United Nations (2011), impressively demonstrate the internationalisation of production; and even the global crisis could not stop this evolution. To cite just some indicators: 2 • With about $16 trillion, the estimated added value generated by transnational corporations in 2010 accounted for more than a quarter of global gross domestic product (GDP) in that year. • Around 40% of this value was generated by foreign affiliates, compared to about 35% in 2005. In total volume, their value added even exceeded the precrisis level. Foreign affiliates contributed to more than one-tenth of global GDP and one-third of the world exports. • The total sales volume of foreign affiliates in 2010 almost recovered the volume of 2008. It reflected a growth of 9.1% with respect to 2009. • Finally, the employment by foreign affiliates has grown steadily since the 1990s, accounting for more than 68 million workers in 2010. In this context, the understanding of a transnational corporation’s production activities as a network of interlinked sites (Rudberg and Olhager, 2003) becomes the predominating form for analysing, organising, and managing manufacturing. Although the fast growing globalisation of production is beyond controversy, the management of the manufacturing network indeed remains challenging (Rudberg and West, 2008). Evidence for this comes from different sources. A recent cross-industry survey on “Excellence in Global Operations (X-GO)” conducted by the University of St.Gallen 3, for instance, contrasted the average performance level of the participating networks along a set of distinct network capabilities – termed overall network capability level – with their average performance on strategic manufacturing priorities (covering price, quality, delivery speed, and flexibility, among others). 4 The comparison, as shown in Fig. 1, illustrates the positive impact of the network capability level on the performance on strategic manufacturing priorities, indicating the strategic lever of a proper network management. However, it also highlights the unexploited potential in the network capabilities for a severe a share of participants.
2
For more details, see the World Investment Report of the United Nations (U.N., 2011, pp. 24). The survey was conducted between December 2010 and May 2011 by the Transfer Centre for Technology Management (TECTEM) of the University of St.Gallen. For more details on the survey, see Section 3.2.3. 4 For a summary of the network capabilities and strategic manufacturing priorities evaluated, see Section 2.3.1 and the Appendix B.2 and B.3. For more details on the calculation of the overall network capability level and the overall performance level on strategic manufacturing priorities, see Section 3.2.3. 3
4
INTRODUCTION Overall network capability level
Overall network capability level as deviation from mean
Below average
Above average
n = 48
Overall performance level on strategic manufacturing priorities
Below average
Above average
Overall performance level on strategic manufacturing priorities as deviation from mean
n = 41
Fig. 1: Network capability level vs. performance level on strategic manufacturing priorities 5
Reasons for that unexploited potential are manifold and can not only be located in the network configuration. Klassen and Whybark (1994), for instance, conducted a Delphi survey addressing a panel of consultants, academics, and professionals to identify the main barriers for an effective management of global operations; six of the eleven most cited barriers were related to coordination. Among those were the management of the factory networks, i.e., the definition of the sites’ strategic purpose and the design of their relationship, the organisational structure comprising the establishment of reporting and communication channels, the balancing of autonomy in the network, the transfer of management skills, as well as the establishment of suitable performance measures. Newer studies confirm that companies have not yet tackled these challenges sufficiently. Kinkel and Maloca (2009), for example, analysed the reasons for German manufacturers reallocating their foreign affiliates. As shown in Fig. 2, the underestimated coordination effort of the headquarters is ranked number five. Anecdotic evidence from a series of interviews with operations managers underlines these findings. The Seals NW, which is the global manufacturing network of a German mechanical engineering company, for instance, lost control due to insufficient transparency, missing rules, and unclear responsibilities. Since barriers for manufacturing are very low in the company's core business and information exchange is poor, the service centres in the network started to build up their own production competencies for serving their local markets. Choosing the right balance between 5
For a detailed description of the evaluation, see Section 3.2.3 and Fig. 20. Besides, it has to be noted that the additional dimension “conformance with aspired network capabilities” in Fig. 20 has not been considered in Fig. 1.
INTRODUCTION
5
centralisation and autonomy, but also defining standards to maintain parental control has been neglected. Similar are the experiences of the Excitation NW, a manufacturer of electrical devices. The local affiliate in South America independently intended to open up a new production site without the headquarters even noticing; a fact termed “Globalisation 2.0” by the central operations manager. Quality
Flexibility, delivery capability Labour costs
33
Transportation / Logistics costs
20
Coordination / Control effort
Availability & fluctuation of qualified worforce
Know-how erosion / IP rights
Proximitiy to domestic R&D
2
5
43
68
32
19
Share of reallocating companies citing the reason
100%
Fig. 2: Reasons for reallocating foreign facilities (adapted from Kinkel and Maloca (2009))
These and other impressions from industry support the believe that when (re-)designing their manufacturing network, most companies, if at all, focus on the global footprint and on configuration decisions, neglecting or postponing the design of the coordination layer. The same is true for the management of existing networks. While the understanding of material flows and logistic transactions between sites is often well-grounded, in many cases, an overview of information and knowledge flows as well as of the sites’ competencies and responsibilities is missing. 1.1.3 Theoretical Gaps Following Shi and Gregory (1998), Rudberg and West (2008), and Miltenburg (2009), practice meets academia with the need to solve the addressed challenges by novel approaches, methods, and processes to understand, design, and improve manufacturing networks. Albeit, the summary of theoretical gaps as assessed in the literature review in Chapter 2 confirms that existing contributions do not yet meet these requirements sufficiently: • First, current literature lacks a holistic understanding of the nature and architecture of manufacturing networks; especially from a superordinate network management perspective. This, in turn, has led to a limited number of tools, frameworks, and approaches supporting the network manager in the analysis, design, and optimisation of the network.
6
INTRODUCTION
• Second, the few existent strategic management approaches, processes, and frameworks identified reveal major limitations. They generally lack evidence of applicability and utility. Further, they focus on network configuration, either completely neglecting the coordination layer, or tackling it rather superficially. • Third, not only is the coordination layer of the network neglected by existing management approaches, but it is also generally underrepresented in research. Especially its definition, the constitution of its elements and their linkage, as well as the interfaces to network configuration have been widely unexplored.
1.2 Research Foundation & Question 1.2.1 Research Background & Foundation A first step towards a holistic understanding of manufacturing networks is provided by the X-GO model 6 as developed by the Institute of Technology Management at the University of St.Gallen. Although of descriptive nature only, the model, as sketched in Fig. 3, gives valuable insights.
Fig. 3: X-GO Model (adapted from Friedli et al. (2011) and Thomas (2013))
Generally, the model links network management with ideas of contingency theory. Contingency research in operations management is based on the assumption that external FIT between an organisation and its environment, as well as internal FIT between the structures and processes within the organisation result in better 6
For more detailed information on the X-GO model, its construction, and elements, see Thomas (2013).
INTRODUCTION
7
performance (Friedli, 2006). As highlighted in Fig. 4, Drazin and Van de Ven (1985) differentiate between three forms of FIT: selection, interaction, and system. Selection Context
Response variable
Interaction Context
Context
Response variable
System
Performance
Response variables
Performance
Fig. 4: Forms of FIT (adapted from Sousa and Vos (2008) based on Drazin and Van de Ven (1985))
• According to the selection approach “… fit is seen as a basic assumption underlying congruence propositions between the organizational context and response variables” (Sousa and Voss, 2008, p. 706). Likewise, the selection approach does not evaluate the impact of the context-response relationship on performance variables (Sousa and Voss, 2008). • In the interaction approach, FIT is seen as the interaction of pairs of organisational context and response variables which, in turn, affects performance (Sousa and Voss, 2008). • For the system approach, FIT is understood as consistency between sets of contextual variables and interdependent response variables which, in turn, affects performance characteristics (Sousa and Voss, 2008). The system view is enhanced by the concept of equifinality proposing various equally efficient ways to achieve FIT based on different combinations and characteristics of the response variables (Doty et al., 1993; Sousa and Voss, 2008). The X-GO model follows this system approach. It is built upon the separation of a network into three main layers: strategy, configuration, and coordination. Each layer is broken into distinct decision dimensions and variables. A proper network design requires FIT between the layers and contextual factors which leads to a superior network performance; but FIT has to be achieved on different levels. First, with respect to a single decision dimension of a certain layer, its variables have to be aligned. Second, with respect to one single layer, the decision dimensions belonging to this very layer have to be aligned. With respect to the network as a closed system, decision dimensions within one and the other layers have to be aligned accordingly, but they also have to be matched with internal contextual factors, like the products manufactured, the manufacturing processes conducted, etc. Finally, with respect to an
8
INTRODUCTION
open network system, this alignment is also subject to external contextual factors, such as customers, competitor, industry trends, etc. Thereby, equifinality postulates that any alignment is not exceptional but can be realised by several equally efficient and consistent set-ups of the layers, the decision dimensions, or the variables. Summing up, the basic assumptions reflected by the X-GO model, i.e., the separation of the network into layers, decision dimensions, and variables, and the link with contingency theory, i.e., the call for a contingent set-up of the single elements, constitute the backbone for the current research. 1.2.2 Research Question Condensing the practical relevance, the theoretical gaps, and the foundation in contingency theory, the guiding research question for this study is as follows: Q. 1: How can the strategic coordination of intra-company manufacturing networks be supported systematically and methodically from a network level perspective? In order to answer the main research question, the following second-order questions shall be addressed predefining and structuring the research approach. Q. 1.1: What are the main decision dimensions and variables characterising the strategic coordination of intra-company manufacturing networks? Q. 1.2: How can decision making along these dimensions be supported systematically and methodically, elaborating the network coordination layer? Q. 1.3: What are contextual factors to be considered for decision making along these dimensions, especially with regards to the network configuration? Q. 1.4: How can decision making support be integrated into a generic network management architecture and design approach?
1.3 Research Methodology & Design 1.3.1 Research Grounding & Process This study is motivated by practical problems of manufacturing companies acting in a dispersed environment. It aims at contributing to the knowledge base by creating practical solutions for their business reality, namely to support operations managers in the systematic and methodical management of their network. Thereby, this research follows Ulrich’s (1984) tradition of business administration as system-oriented applied management science, targeting the “… designing, controlling and further developing
INTRODUCTION
9
(of) purpose-oriented socio-technical organisations” (Ulrich, 1984; Rüegg-Stürm, 2005, p. 11). In this context, the development of generic and applicable normative models to design and change the social reality becomes the primary intention (Ulrich, 1984). The complexity of a manufacturing network, as the excerpt of social reality that is looked at, is accounted for by giving up the aspiration of full controllability, thus turning away from the irrefutable logic of natural sciences (e.g., Giddens, 1984; Friedli, 2006) and from searching for the universal “truth”. Given the poor holistic understanding and foundation of manufacturing networks in theory and practice, this study does not focus on framing and empirically testing of theory-driven hypotheses (Kubicek, 1977; Tomczak, 1992). 7 Instead, it seeks for new theory building with regards to the design of social realities in today’s manufacturing organisations (Ulrich, 1984). Thereby, it focusses on qualitative exploration and application rather than on quantitative explanation. Accordingly, it is grounded in the social constructionism position 8, striving not for quantitatively demonstrating causalities and interdependencies but for establishing a sound understanding of the reality; for this, the researcher is not seen as independent observer but as integral component of the phenomena studied (Easterby-Smith et al., 2002).
Initial findings from literature (Preliminary) theoretical knowledge Initial findings from practice
Practical problems
Questions addressed to practice
Research as iterative learning process
Differentiation, abstraction
Theoretical work
Data collection Practical phenomena
Critical reflection
Empirical work
Fig. 5: Research process (adapted from Kubicek (1977), Tomczak (1992), and Gassmann (1999)) 7
Kubicek (1977) already considered the testing of arbitrary hypotheses only derived from theory but without practical foundation as questionable for the academic progress. 8 The constructivism position accounts for the complexity of the reality to be observed, considering the researcher as integrated part of it. Thereby, it strives for a general understanding of the reality by mainly theoretical abstraction and qualitative data analysis. The positivism position, as its counterpart, considers the social world as existing externally and independent of the assumptions and beliefs of the researcher. This reality has to be simplified for analysis, concentrating on causalities by objective measures and quantitative methods mainly (Easterby-Smith et al., 2002).
10
INTRODUCTION
Consequently, a highly iterative learning approach underlines the research process. It is based on Kubicek’s (1977) iterative heuristic as adapted in Fig. 5. Starting with an initial understanding of a particular problem, questions are raised and addressed to practice. Empirical data is collected to answer these questions, and the critical reflection of this data raises new questions, thus driving a systematic accumulation of experiences and their incremental transformation into theoretical knowledge. 1.3.2 Research Methodology, Theory Building, Data Collection & Analysis To answer the research question, qualitative case study research is chosen as main strategy. Preferring this strategy to others, particularly to experiments, quantitative surveys, archival analyses, and histories (Yin, 2003), is motivated by several reasons: • It is motivated by the constructionism position targeting a general and holistic understanding of manufacturing networks and their design by theoretical abstraction and not by reducing networks to simple elements and conducting (statistical) hypothesis testing on them (Easterby-Smith et al., 2002). • It is motivated by the nature of the main and second-order research questions comprising “how” and exploratory “what”, rather than “who, where, how many, and how much” questions. In line with Yin (2003) “… the most important condition for differentiating among the various research strategies is to identify the type of research question being asked. In general, “what” questions may (…) be exploratory (in which case any of the strategies could be used) (…) “how” and “why” questions are likely to favor use of case studies, experiments, or histories” (Yin, 2003, p. 7). Given the study’s focus on contemporary events, i.e., the design of the network instead of its historical evolvement, histories become impractical as strategy. Given the complexity and time horizon to implement and observe any changes in the network, the researcher’s control and manipulation of behavioural events, too, is hard to realise, which makes case study research preferable to experiments (Yin, 2003). • Finally, qualitative social research is appropriate when not much is known of a given issue, or it is recommended when results are known or assumed but should be looked at from a different point of view (Eisenhardt, 1989; Stuart et al., 2002; Yin, 2003). For this study, both applies. Anticipating the result from the literature analysis in Chapter 2, integral approaches for the design and management of manufacturing networks are novel and not yet well-founded, particularly with respect to the integration of the network coordination. Nonetheless, coordination itself is nothing new from an organisation theory’s perspective. However, transforming this perspective to the idiosyncrasies of manufacturing networks calls for changing the point of view.
INTRODUCTION
11
Overall, the research stages, including the methodology applied and the main results targeted, can be characterised as outlined in Fig. 6. Although in fact of highly iterative nature, for the sake of lucidity, the progress has been retrospectively sketched as stepwise approach. As argued above, the employed approach combines a set-up of multiple case studies enriched by input from a cross-industry survey. Decision dimensions & management frameworks
Research question(s) 0
Initial interviews & literature screening • Identification of practical relevance • Initial formulation of research question(s)
Research Question(s)
Q.1 :
1
2
Desk research
3
Qualitative case studies
• Derivation of gaps and implications from literature
• Conceptual design and testing of management frameworks
• Conceptual design of heuristic research framework
• Refinement of decision dimensions and characteristics
• Refinement of research question(s)
• Identification of coord. (config.) decision dimen. and characteristics
Quantitative study & qualitative interviews • Validation and refinement of management frameworks • Discussion of contextual impact
• Discussion of contextual impact
Design approach
Network architecture 4
5
Qualitative case studies • Transformation of heuristic research framework into network management architecture • Validation of network management architecture
Conceptual design
• Transformation of network management architecture into strategic network design and management approach
Q.1.1: Q.1.2: Q.1.3: Q.1.4:
Fig. 6: Research stages
STAGE 1 – Based on an initial understanding of the practical and theoretical relevance of the research goal and questions gained in stage 0, an extensive review of manufacturing network literature in operations management is carried out in stage 1. The critical reflection of the literature allows positioning the present research in the knowledge base, to sharpen the final research question, and to conceptually derive a heuristic research framework setting and limiting the scope for the subsequent work. The literature review also enables deeper insights into the three central layers of manufacturing networks, i.e., strategy, configuration, and, in particular, coordination. The coordination layer, in turn, is broken down into a set of distinct decision dimensions and variables. At the end of this stage, the research question is frozen. STAGE 2 – The decision dimensions represent the levers to design the coordination layer. To provide operational feasibility, each coordination dimension is conceptually transformed into a portfolio-based “management framework”. The design of the frameworks is based on close cooperation with five case study networks to ensure both a sound grounding in academia and demonstrated applicability in practice. The case networks have been primary selected with respect to their structural differences,
12
INTRODUCTION
aiming to achieve a high degree of generalisability of the frameworks. 9 Multidisciplinary workshops with the operations management teams of each network have been conducted, serving as platform to discuss and test the frameworks and to complete the underlying decision dimensions and variables. The highly iterative approach is in line with the “process research approach” as suggested by Shi and Gregory (1998), who postulate an action research 10 process for strategic and operational decision making (Shi and Gregory, 1998; Shi, 2003). According to them, frameworks in combination with in-depth discussions facilitate the research process since they (1) offer structure and guidance for discussion, (2) provide traceability of the information and data collected, and (3) ensure a combination of theoretical exploration and practical validity (Platts and Gregory, 1990; Shi and Gregory, 1998; Shi, 2003). The approach is also in line with Eisenhardt’s (1989) recommendation of using a priori constructs at the early phase of a theory building process. However here, the frameworks do not only support the interaction with practice; they also constitute a central piece of results. At the end of this stage, research question 1.1, i.e., the identification of decision dimensions and variables characterising the strategic coordination of intra-company manufacturing networks, is answered. STAGE 3 – Results from the design phase are complemented by a cross-industry survey on manufacturing networks. The purpose of the survey is not to quantitatively test hypotheses on the frameworks but to put them in a wider industry context, identifying obstacles, practical approaches, methods, and “success-stories” for the network coordination. Additional semi-structured interviews have been conducted with representatives of the global operations management team of three participants, selected based on findings from the survey. 11 Moreover, applying different quantitative and qualitative methods but not changing the object of study allows for triangulation (Denzin, 1978). In order to synthesise the ample amount of information from different sources, “anecdotic evidence” is preferred to a detailed presentation of the data to underline the frameworks and their logic. This strategy enables a more condensed and target-oriented integration of the findings. At the end of this stage, research question 1.2, i.e., the systematic and methodical support for decision making along the coordination layer, is answered by providing management frameworks for manufacturing network coordination. 9
For the characteristics of the case networks and the reasoning on their selection, see Section 3.2.1 and 3.2.2. Action research, as postulated by Lewin (1946), describes a theory building approach based on a researcher’s close interaction with practice. It comprises an iterative cycle of three phases: planning, execution, and reconnaissance or fact finding. 11 For the characteristics of the interviewed networks and the reasoning on their selection, see Section 3.2.3. 10
INTRODUCTION
13
STAGE 4 – The coordination frameworks can be understood as the bricks of a generic network management architecture. Since such architecture has to be holistic in nature, not only the coordination but also the configuration layer has to be integrated. For that, distinct steps of stage 1 to 3 have had to be repeated with respect to the elaboration of the main configuration decision dimensions. Finally, the network architecture is presented, and its applicability and utility are demonstrated by two in-depth case studies. 12 Restricting the number of case studies to two is feasible since the management frameworks themselves have been validated before in stage 3. Hence, at this point, only the proof of validity for the whole architecture as well as its integration into a systematic and procedural network design and management approach is missing. At the end of this stage, research question 1.3 and the first part of research question 1.4, i.e., the provision of a generic network design and management architecture integrating the coordination dimensions but also the contextual impact of the coordination layer, are answered. STAGE 5 – The final stage transforms the management architecture into a “suggested practice approach” for strategic network design and management. Benefitting from the results of the previous stages, it condenses the experiences of the “research journey”. At the end of this stage, the second part of research question 1.4, i.e., the provision not only of a design and management architecture but also of a procedural approach, is answered. The overall empirical data base for this study is given in Tab. 1. • It comprises the cooperation with five case study networks for the design and testing of the management frameworks with overall 231 workshop hours; most of this time was dedicated to the design of the coordination frameworks, some of it went into the completion of the configuration layer. • It comprises data from a cross-industry survey on excellence in global operations. The survey was conducted by the University of St.Gallen during December 2010 to May 2011. Approximately 550 companies with manufacturing networks were contacted via e-mail and phone, more than 250 questionnaires were placed, and 56 networks of eleven different industries participated. 13
12 13
For the characteristics of the two in-depth case studies and the reasoning on their selection, see Section 4.2.1. For more details on the survey, see Section 3.2.3 and Appendix B.1.
14
INTRODUCTION
Head of Global Product Line
Drives NW
Head of Global Product Line
Seals NW
Edgeband NW
Profile NW
Floor Care NW Dental NW
Head of Production** Manager Global Production Head of Operations Manager Production Head of SCM** Manager Business Development Head of Division** Head of Business Unit** Head of Operations Manager Production Manager Technics Head of Product Line Assistant to COO Site Head Head of Global Production
Head of Global Production Head of SCM Site Head** Assistant to Site Head Elevator NW Head of Engineering Head of SCM Head of International Operations Manager Production Chocolate NW Manager SCM Manager Controlling 56 Respondents from the quantitative survey Overall Number Total Duration Pet Food NW
op
or ks h W
Excitation NW
#
Functions involved
#
Company / NW*
In
te rv ie w s/
Di sc us
si o
n Fr s am de e w ve ork lo pm Ar en ch t& i te tec te st tu st in re in g g Im pl i de c at sig ion n sf & or m th gt e . a st pp rat ro eg ac ic h
• It comprises three one day on-site visits and discussions with selected networks from the survey sample to verify the findings and to get more information on their network management in practice. • It comprises two in-depth case studies, with twelve, respectively seven workshops, to apply the management architecture for analysing and conceptually (re-)designing the network and to gain implications for the construction of the strategic network management approach.
13 (avg. 3 h) 8 (avg. 3 h) 8 (avg. 7 h)
x x x
9 (avg. 7 h)
x
7 (avg. 7 h)
x
1 (ca. 7 h)
(x)
1 (ca. 7 h) 1 (ca. 7 h)
(x) (x) 12 (avg. 3h)
(x)
x
x
7 (avg. 4h)
(x)
x
x
x 3 21 h
67 295 h
* NW = Network ** Participation only sporadic
Tab. 1: Overview of the empirical data base 14
14
For a more detailed overview of the case studies and interviews, see Appendix A.
INTRODUCTION
15
To guarantee for rigor and reliability of the data obtained, and to make a detailed presentation of the findings easier, all companies’ identities are kept anonymous. Data gathering itself relied on field-work using multiple sources, ranging from qualitative data obtained by workshops, interviews, company presentations, and strategy papers, to quantitative data obtained by the survey. As indicated by the list of persons involved, workshops and interviews were conducted with representatives from the middle and top management of the companies’ operations function. Any interaction with the case networks and the interview partners was carried out in a team of at least two researches; improved creativity due to complementary perspectives and different perceptions of the involved researchers benefitted especially the design phase of the frameworks, but it also strengthens the confidence in the findings (Eisenhardt, 1989). Multiple data sources and evaluators, in turn, allowed for data and investor triangulation (Flick, 2002).
1.4 Study Structure The structure of this study is highlighted in Fig. 7. It is organised along the above outlined research stages comprising six chapters: • Chapter 1 “Introduction”: The present chapter introduces the research motivation, the practical and theoretical relevance, as well as the research foundation. These pillars are synthesised by the guiding research question for the study. How to actually answer the raised question is sketched by the research methodology and design. • Chapter 2 “Understanding Manufacturing Networks”: The second chapter presents the results from an extensive literature analysis on manufacturing networks. It constitutes a common understanding of manufacturing networks in general, addresses the content of the network strategy, configuration, and coordination, and reviews existing network design and management approaches. The chapter further concretises the theoretical gaps and derives implications for the current work ending in the construction of a heuristic research framework. • Chapter 3 “Designing the Network Coordination Layer”: The third chapter focuses on the elaboration of the network coordination layer. The central coordination decision dimensions are isolated and operationalised by single management frameworks. The proceeding is based on both literature and empirical data. The chapter is completed by contrasting the findings with the organisation theory’s perspective on coordination, by discussing contextual impacts on the coordination layer, and by deriving implications for the development of a network management architecture and design approach.
16
INTRODUCTION
• Chapter 4 “From the Coordination Layer to a Management Architecture”: The fourth chapter accounts for the linkages between network configuration and coordination. It operationalises the configuration decisions dimensions and anchors both layers in the final network management architecture. The applicability of the architecture is validated and tested by two in-depth case studies. The chapter finishes by discussing findings with respect to a strategic network design and management approach. • Chapter 5 “From a Management Architecture to a Strategic Design & Management Approach”: The fifth chapter condenses previous results and findings by transforming the network architecture into a “suggested practice approach” for strategic network design and management. This approach is positioned against existing methods and processes. • Chapter 6 “Summary & Outlook”: The final chapter summarises and critically reflects the study’s results with respect to their academic and practical contribution. The study ends highlighting both limitations of the current work and possibilities for further research. 1
Introduction 2
Understanding Manufacturing Networks 3
4
Analysing the Knowledge Base on Manuf. Networks
Research Foundation & Question
Manufacturing Site vs. Network Perspective
From the Coordination Layer to a Management Architecture
Introducing an Integral Architecture for Network Management
Summary & Outlook
Decision Categories & Layers of Manuf. Networks
Summary & Discussion
Defining the Decision Dimensions of Network Coordination
Architecture to a Strategic Design & Management Approach
Research Methodology & Design
Study Structure
Designing the Network Coordination Layer
5 From a Management
6
Motivation & Relevance
From Decision Dimensions to Coordination Frameworks
Network Design & Management Approaches Developing the Coordination Frameworks
Summary & Discussion
Working with the Network Management Architecture
Summary & Discussion
Presenting a Strategic Network Design & Management Approach Summary & Discussion Critical Reflection
Contribution to Theory & Practice
Limitations & Further Research
Fig. 7: Study structure
17
UNDERSTANDING MANUFACTURING NETWORKS
2 Understanding Manufacturing Networks The following chapter is dedicated to a literature analysis on manufacturing networks. Due to the holistic aspiration of this study, the analysis is not restricted to network coordination only but reviews the current state of research on manufacturing networks in general: its structure and elements, its design, optimisation, and management. The purpose of this chapter is to (1) provide a fundamental understanding of manufacturing networks, (2) concretise the theoretical gaps as summarised in Section 1.1.3, (3) position the current study within the operations’ management knowledge base, (4) give an overview of existing frameworks, approaches, and methods serving as backbone and building blocks for this work, and to (5) synthesise findings by deriving implications from literature. The analysis follows a defined process outlined in Fig. 8. Literature analysis 1
2 Literature screening • Definition of literature search pattern
• Journal screening for relevant literature • Quantitative evaluation of screening results
3 Literature review • In-depth review based on the literature list of De Toni and Parussini (2010)
• Complementation of the original list based on the screening results
• Extension of the original list by other literature formats and scope
4 Literature discussion
• Discussion of most relevant literature contributions
a) Manuf. site vs. network perspect.
b) Decision categories and layers c) Network design and management approaches
Synthesis of findings
• Derivation of implications from literature
• Design of heuristic research framework
Fig. 8: Literature analysis approach
• Screening: First, based on a discussion of prevailing perspectives to cluster and structure research on manufacturing networks, a search pattern is defined as starting point for a literature screening. Aside from slight modifications, it adapts the structure of De Toni and Parussini (2010), one of the most recent reviews of international manufacturing networks. The screening itself was carried out by a research team at the Institute of Technology Management at the University of St.Gallen in order to explore the field of manufacturing network management. It was conducted quantitatively only covering 674 papers in 30 renowned journals over the last 25 years.
18
UNDERSTANDING MANUFACTURING NETWORKS
• Review: Second, an in-depth literature review is conducted. This step also builds upon the groundwork of De Toni and Parussini (2010). Their results are critically revised with respect to the current research interest, removing papers and adding some others identified as relevant in the screening process. De Toni and Parussini’s (2010) review is complemented by contributions not yet being captured by the defined search pattern. Thematically, these comprise integral management models for manufacturing networks as well as network design and optimisation approaches and methods. Additionally, the review is extended by other formats than journal articles (e.g., books, book sections, and conference proceedings) and by central contributions from research on multinational organisations / corporations. • Discussion: Third, constituting the central part of this chapter, selected contributions are presented in detail, and implications for the subsequent research work are derived. • Synthesis: Fourth, implications from the analysis are summarised and transformed into a heuristic research framework. The chapter is organised as follows: Section 2.1 defines the scope and set-up of the outlined literature screening process and review. Sections 2.2 to 2.4 discuss those contributions considered most important for this study. Section 2.5 summarises the results, highlights the identified research gaps and implications, and finishes with the introduction of the heuristic research framework.
2.1 Analysing the Knowledge Base on Manufacturing Networks 2.1.1 Structuring the Knowledge Base Up to now, there has been no consensus on how to confine and structure research on manufacturing networks. Typically, two main perspectives have been pursued: one based on the level of investigation, i.e., the focal unit, and one based on the three main network decision layers: strategy, configuration, and coordination. Other authors complement these perspectives by the network’s geographic dispersion (Bartlett and Ghoshal, 1989; Shi and Gregory, 1998; Miltenburg, 2009). Focal Unit As outlined in Fig. 9, Rudberg and Olhager (2003) classify research on so-called value networks according to the number of organisations involved and the number of sites per organisation. They distinguish four different levels of analysis: plant, intra-firm / intra-company network, supply chain, and inter-firm / inter-company network.
19
UNDERSTANDING MANUFACTURING NETWORKS
Multiple 3
Number of organisations in network
Single
Supply Chain
(multi-organisation, single-site)
1
Plant
(single-organisation, single-site)
Single
4
Inter-firm network
(multi-organisation, multi-site)
2
Intra-firm network
(single-organisation, multi-site)
Number of sites per organisation
Multiple
Fig. 9: Classification of value networks (adapted from Rudberg and Olhager (2003))
A (1) single plant constitutes the smallest value adding entity. The (2) intra-firm / intra company network broadens this perspective. It can be understood as integral system – a single organisation’s plants are viewed as scattered nodes that are linked by matrix connections where each node affects another; thus, plants cannot be managed in isolation (Shi and Gregory, 1998; Rudberg and Olhager, 2003). The (3) supply chain perspective comprises several independent organisations, with one or a few plants each, primarily connected by material flows. Finally, the (4) inter-firm / inter-company network is a combination of both the intra-firm network and the supply chain with multiple organisations and sites involved (Rudberg and Olhager, 2003). Geographic Dispersion The network’s geographic dispersion reflects the global spread of the manufacturing plants. According to Miltenburg (2009), there are four degrees of geographic dispersion based on the location of plants from a headquarters’ perspective: national, regional, multinational, and worldwide. A national and regional dispersion is what he calls “simple network” while plants scattered on a multinational or worldwide level constitute a “complex network”. This distinction is linked with seven generic management strategies for international manufacturing based on the “pressure for globalisation”, i.e., the necessity to design, produce, and sell goods on a worldwide level, and the “pressure for local responsiveness”, which is the necessity to adapt products and operations to local needs (Shi and Gregory, 1998; Miltenburg, 2009). Decision Layers Another way to structure research on manufacturing networks is along the main decision categories or layers as reflected by the X-GO model in Section 1.2.1: manufacturing (network) strategy, configuration, and coordination (e.g., Porter, 1986;
20
UNDERSTANDING MANUFACTURING NETWORKS
Rudberg and Olhager, 2003; De Toni and Parussini, 2010; Cheng et al., 2011).15 Manufacturing strategy is about specifying manufacturing's strategic priorities in order to leverage competitive advantage in business strategy (Wheelwright, 1984). It can be supported by both site and network level, each providing distinct capabilities (Shi and Gregory, 1998; Miltenburg, 2009). Configuration deals with the structure and physical layout of sites and the network, including decisions on the number of sites, their global dispersion, competencies, capacities, and technology (De Toni and Parussini, 2010). Coordination addresses the organisation and management of the global activities. Decisions have to be made on how plants should interact, autonomy is assigned, resources are allocated and shared, and knowledge and information are exchanged. Thus, coordination refers to the question of how to link and integrate the plants to support strategic business objectives (Meijboom and Vos, 1997; Cheng et al., 2011). 2.1.2 Reviewing the Knowledge Base Combining the introduced perspectives defines the search pattern that is applied to structure and to organise the literature analysis. Literature Screening The literature screening was carried out on 30 renowned journals relevant in operations management research. 16 Due to certain access restrictions, the covered time span varies between 1986 and 2011. The screening was conducted for defined search terms in abstract, title, and key-words. Terms were made up combining the perspectives of the introduced search pattern. Thereby, some cuts were made: First, since the research interest focuses on single organisations, only plants and intra-firm networks were considered as focal unit; the term "supply chain" was not searched for. Second, while the terminology regarding the focal unit and decision layers is mostly uncontroversial, there are multiple ways denoting the degree of “internationalisation” 17; hence, differences in the geographic dispersion were not explicitly considered. To conclude, the search terms applied reflect combinations between the focal unit, i.e., “network” and “plant”, and decision layers, i.e., “strategy”, “configuration”, and “coordination”, complemented by “manufacturing” and “production”. Tab. 2 summarises the screening set-up. 15
A more detailed discussion on the distinct decision layers (categories) will be provided in Section 2.3. The screening covers the journal sample as introduced by Petersen et al. (2011). They applied a meta-analysis on operations management literature coming up with a set of 32 most important journals. This set was reduced by excluding journals with a clearly different focus, like quality and service management or mathematical programming. Moreover, some others were added which were already known to contain central articles. 17 For example: national, domestic, regional, multi-domestic, international, transnational, global, and worldwide. 16
21
UNDERSTANDING MANUFACTURING NETWORKS
Key-words Network:
ma nufa cturi ng network, producti on network, pl a nt network, i ntra orga ni za ti ona l network, i ntra orga ni s a ti ona l network
Strategy: s tra tegi c ma nufa cturi ng, network s tra tegy, pl a nt s tra tegy, ma nufa cturi ng s tra tegy, producti on s tra tegy
Configuration: network confi gura ti on, pl a nt confi gura ti on, ma nufa cturi ng confi gura ti on, producti on confi gura ti on
Coordination: network coordi na ti on, pl a nt coordi na ti on, ma nufa cturi ng coordi na ti on, producti on coordi na ti on
Filtering criteria and search target Filtering Criteria: none Target: title, abstract, key-words
Journals
Period
Aca demy of Ma na gement Journa l Aca demy of Ma na gement Revi ew Ca l i forni a Ma na gement Revi ew Deci s i on Sci ences Europea n Journa l of Opera ti ona l Res ea rch* Ha rva rd Bus i nes s Revi ew IIE Tra ns a cti ons Indus tri a l Ma na gement & Da ta Sys tems Interfa ces Interna ti ona l Journa l of Logi s ti cs Res ea rch a nd Appl i ca ti ons Interna ti ona l Journa l of Opera ti ons & Producti on Ma na gement** Interna ti ona l Journa l of Phys i ca l Di s tri buti on & Logi s ti cs Ma na gement Interna ti ona l Journa l of Producti on Economi cs
1989-2011 1980-2011 1986-2011 1990-2011 1995-2004 1990-2011 1995-2011 1992-2011 1990-2011 1998-2011 1994-2011 1986-2011 1999-2011
Interna ti ona l Journa l of Producti on Res ea rch**
1986-2011
Journa l Journa l Journa l Journa l Journa l Journa l Journa l
1985-2011 1993-2006 1992-2011 1985-2011 2008-2011 1987-2011 1986-2011
of Interna ti ona l Bus i nes s Studi es of Ma nufa cturi ng Sys tems * of Ma nufa cturi ng Technol ogy Ma na gement of Opera ti ons Ma na gement of Opti mi za ti on i n Indus tri a l Engi neeri ng*** of Suppl y Cha i n Ma na gement of the Opera ti ona l Res ea rch Soci ety
Ma na gement Sci ence
1986-2011
Omega Producti on a nd Inventory Ma na gement Journa l * Producti on a nd Opera ti ons Ma na gement Producti on Pl a nni ng & Control Sl oa n Ma na gement Revi ew Stra tegi c Ma na gement Journa l Suppl y Cha i n Ma na gement Suppl y Cha i n Ma na gement Revi ew
1995-2011 1987-2002 1997-2011 1990-2011 1990-2011 1986-2011 1996-2011 2003-2011
* Acces s res tri cted ** Excl udi ng "ma nufa cturi ng s tra tegy" & "producti on s tra tegy" *** Former: Journa l of Indus tri a l Engi neeri ng
Tab. 2: Literature screening set-up
The results of the screening are outlined in Fig. 10, which shows the absolute number of papers for each search term family. Altogether, 674 papers were identified covering the full set of search terms per period.
22
UNDERSTANDING MANUFACTURING NETWORKS
Coordination
Configuration
Strategy
Network
Overall = 674 (multiple assignments possible) 169
191 165
127
39
1986 1990
24
1991 1995
1996 2000
2001 2005
2006 2010
2011
Fig. 10: Literature screening results
Overall, results show a growing number of articles except for 2001 to 2005. Especially for papers with a focus on “networks” in general, the number has steadily been increasing over time, hence confirming the importance of this level of investigation. Regarding the three different decision layers, with 532 contributions in total, the sample is clearly dominated by literature on strategy, followed by configuration with 65 papers, which is obviously getting more prevalent especially within the last decade. With 18 contributions in total, coordination, in turn, is clearly underrepresented in each of the investigated periods. Literature Review A comprehensive literature review of international manufacturing networks is provided by De Toni and Parussini (2010). They carried out a broad search in five most known databases, combining quantitative screening with qualitative filtering, that is, reading of abstracts and removing contributions lacking an international manufacturing scope or management perspective. Their findings are similarly clustered along the network decision layers, providing a good starting point for an indepth analysis. Therefore, their results are revised critically with respect to the current research interests and modified accordingly. Additionally, the review is complemented by important contributions not yet being captured due to a different format or a divergent scope, i.e., those comprising integral management models or frameworks for manufacturing networks, touching network design and optimisation approaches, or representing research on multinational organisations / corporations. Tab. 3 summarises the results; contributions added to the original review are marked with an asterisk (*).
23
UNDERSTANDING MANUFACTURING NETWORKS
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67
Abele et al.* Aikens* Baden-Fuller & Stopford Bartlett & Ghoshal* Bartlett & Ghoshal Bartmess & Cerny* Bartol & Srivastava* Bhatnagar & Chandra* Birkinshaw* Bolisani & Scarso Bonaglia et al. Canel & Khumawala Canel & Khumawala* Cerrato* Cheng et al.* Cheng et al.* Chew et al.* Christodoulou et al.* Coe et al Colotla* Colotla* Colotla et al. Dasu & de la Torre De Toni et al. Diederichs et al.* DuBois et al. Dyer & Nobeoka* Ernst & Kim* Ernst & Kim Feldmann* Feldmann & Olhager* Feldmann & Olhager*
2008 1985 1991 1989 2000 1993 2002 1993 2001 1996 2007 1996 2001 2006 2008 2011 1990 2007 2008 2002 2003 2003 1997 1992 2008 1993 2000 2001 2002 2011 2009a 2009b
B J J B J J J J J J J J J J C J BS B J C B J J J BS J J C J B W C
Global Production: A Handbook for Strategy and Implementation European Journal of Operational Research Strategic Management Journal Managing Across Borders: The Transnational Solution Harvard Business Review California Management Review Journal of Leadership & Organization Studies European Journal of Operational Research California Management Review International Journal of Operations & Production Management Journal of World Business International Journal of Operations & Production Management International Journal of Operations & Production Management International Business Review 9th CINet Conference (Valencia) International Journal of Operations & Production Management In Kaplan 1990: Measures for Manufacturing Excellence (Chap. 5) Making the Right Things in the Right Places Journal of Economic Geography 7th Cambridge International Manufacturing Symposium Operation and Performance of Internat. Manufacturing Networks International Journal of Operations & Production Management Management Science International Journal of Operations & Production Management In Abele et al. 2008: Global Production (Chap. 7) Journal of International Business Studies Strategic Management Journal Nelson & Winter Conference Research Policy A Strategic Perspective on Plants in Manufacturing Networks Linköping University Working Paper 16th EUROMA Conference (Goteborg)
Feldmann & Olhager* 2011 C 18th EUROMA Conference (Cambridge) Feldmann et al. 2009 J Production Planning & Control Feldmann et al.* 2010 C 17th EUROMA Conference (Porto) Ferdows 1997a J Harvard Business Review Ferdows 1997b J Production and Operations Management Ferdows 2003 J Industrial Engineering Ferdows 2006 J Production and Operations Management Ferdows 2009 J GCG Georgetown University - Universia Ferdows et al. 2004 J Harvard Business Review Friedli et al.* 2011 J Zeitschrift für wirtschaftlichen Fabrikbetrieb Fusco & Spring* 2003 J Integrated Manufacturing Systems Ghoshal & Bartlett* 1988 J Journal of International Business Studies Grallert et al.* 2010 C 15th Cambridge International Manufacturing Symposium Gray et al. 2009 J Decision Sciences Gulati et al.* 2000 J Strategic Management Journal Gupta & Govindarajan* 1991 J Academy of Management Review Jaehne et al.* 2009 J International Journal of Production Research Jacob* 2006 B Quantitative Optimierung dynamischer Produktionsnetzwerke Justus* 2009 B Management globaler Produktionsnetzwerke Karlsson & Sköld 2007 J Journal of Manufacturing Technology Management Ketokivi & Jokinen 2006 J Journal of Operations Management Khurana & Talbot* 1999 W University of Michigan Business School Working Paper Kim & Arnold* 1996 J International Journal of Operations & Production Management Kinkel* 2004 B Erfolgsfaktor Standortplanung Klassen & Whybark 1994 J Journal of Operations Management Kogut & Kulatilaka 1994 J Management Science Lanza & Ude* 2010 J CIRP Annals - Manufacturing Technology Lee & Lau 1999 J International Journal of Agile Management Systems Love et al.* 1988 B Facilities Location: Models and Methods Luo* 2005 J Journal of World Business MacCarthy & Atthirawong 2003 J International Journal of Operations & Production Management Maritan et al. 2004 J Journal of Operations Management Martinez & Jarillo* 1989 J Journal of International Business Studies Mascarenhas* 1984 J Journal of International Business Studies Mauri 2009 J Management International Review
x x x x
x x
x
x
x
Quantitative
Mgt. Approach
Strategic
Institutional perspective
Year
Coord.
Flows
Nr. Author
Journal (J) Conference (C) Working paper (W) Book (B) or Book section (BS)
Configuration
Int. manuf. strategy Network strategy Plant location Network structure Network spec.
Strategy
x
x x
x x x
x x x
x x
x x x x x x x x
x x x x
x x
x x
x x x
x
x x
x x x
x
x x x x
x x
x x x x
x x x x x x x
x x x x x
x x
x x x
x x x
x x
x x
x x
x x
x x x x
x
x x
x x x x
x x
x x
x x
x x x x x x
x
x
x x
x
x x
x x x
x
x
24
UNDERSTANDING MANUFACTURING NETWORKS
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111
2001 J International Business Review 1986 J Journal of International Business Studies 2003 J Tijdschrift voor Economische en Sociale Geografie 1997 J International Journal of Operations & Production Management 2004 J Journal of Purchasing & Supply Management 2008 BS In Abele et al. 2008: Global Production (Chap. 4 & 5) 2005 B Manufacturing Strategy 2008 J International Journal of Production Economics 2009 J International Journal of Production Research 1998 J International Journal of Operations & Production Management 2006 J Journal of Manufacturing Technology Management 1999 J International Journal of Production Research 1986 J California Management Review 2000 J Journal of Operations Management 2001 J International Journal of Operations & Production Management 2007 J Journal of Manufacturing Technology Management 2003 J Omega 2008 J Omega 1990 B Industrielle Standortplanung 1982 J Journal of Operations Management 2003 J Integrated Manufacturing Systems 1998 J Journal of Operations Management 2005 J Production Planning & Control 1997 J Integrated Manufacturing Systems 2001 C 8th EUROMA Conference (Bath) 2000 B Internationales strategisches Produktionsmanagement 1990 J Sloan Management Review 1994 J The International Journal of Logistics Management 2001 J Academy of Management Journal 2002 J Organization Science 1998 J Academy of Management Journal 2009 B Die strat. Rolle von Produktionsstandorten in Hochlohnländern 2010 B Entscheidungsunterstüt. für die Konfig. globaler Produktionsnw. van de Ven 1989 J Engineering Costs and Production Economics Vecchi & Brennan 2009 J Research in International Business and Finance Vereecke & De Meyer* 2009 W Vlerick Leuven Gent Working Paper Vereecke & Van Dierdonck 2002 J International Journal of Operations & Production Management Vereecke et al. 2006 J Management Science Vickery* 1991 J Decision Sciences Vokurka & Davis* 2004 J Industrial Management & Data Systems Vos 1991 J International Journal of Operations & Production Management Wheelwright* 1984 J Strategic Management Journal Wheelwright & Hayes* 1985 J Harvard Business Review Wright et al. 2009 J Asia Pacific Business Review
x x x x x
x x x
x x x
x x
x
x
x
x
x
x x
x x x x x x x x
Quantitative
x x x x
x x
Mgt. Approach
Strategic
Institutional perspective
Year
Mauri & Sambharya Mefford Meijboom & Voordijk Meijboom & Vos Meijboom & Vos Meyer & Jacob* Miltenburg Miltenburg* Miltenburg Nassimbeni* Noori & Lee Pontrandolfo & Okogbaa Porter Prasad & Babbar Prasad et al. Riis et al. Rudberg & Olhager Rudberg & West Schill* Schmenner* Shi Shi & Gregory Shi & Gregory Shi et al. Shi et al.* Stremme* Sugiura Sweeney Tsai* Tsai* Tsai & Ghoshal* Tykal* Ude*
Coord.
Flows
Nr. Author
Journal (J) Conference (C) Working paper (W) Book (B) or Book section (BS)
Configuration
Int. manuf. strategy Network strategy Plant location Network structure Network spec.
Strategy
x x
x x
x x
x x x x x x x x
x x x
x x
x x x
x x x
x
x x x x x
x
x x
x x x
x x x x x
x
x x x x
x
x x x x x
x x
* Own extensions of the literature review by De Toni and Parussini (2010)
Tab. 3: Literature overview (adapted from De Toni and Parussini (2010))
Literature Discussion Based on the literature screening and review, selected contributions are discussed in detail in the next sections. The discussion seeks to highlight research gaps and implications for the underlying study. It is organised as follows: • Section 2.2 addresses the two focal units: manufacturing site and network. • Section 2.3 is structured according to the three decision layers: manufacturing (network) strategy, network configuration, and coordination. • Section 2.4 examines frameworks and approaches for the network design and management.
UNDERSTANDING MANUFACTURING NETWORKS
25
2.2 Manufacturing Site vs. Network Perspective 2.2.1 Manufacturing Site Perspective Research on manufacturing management has originated from a single plant (Shi and Gregory, 2005). There is a rich body of literature on how to organise, optimise, plan, and run the daily business on the manufacturing site, shop floor, or even for a selected production line. With manufacturing becoming global, this scope remains important but is enhanced by considering a plant as an entity that performs value adding activities for a whole network. Two literature streams are dominant from this perspective: location decisions and strategic plant roles. Location Decisions Literature on location decisions provides methodical support for evaluating and selecting the right location to establish a new manufacturing site. Starting from pure cost considerations, qualitative but mainly quantitative approaches nowadays integrate a set of environmental and company-specific factors into mathematical programming and simulation models (e.g., Aikens, 1985; Love et al., 1988; Canel and Khumawala, 1996; Canel and Khumawala, 2001). These methods are complemented by more generic process models that shift the focus from decision support to systematic decision making (Schill, 1990; Kinkel, 2004; Meyer and Jacob, 2008). Besides these still rather technical approaches, Bartmess and Cerny (1993) were among the first pointing out a capability-driven perspective on location decisions. In order to achieve “capability driven financial returns” (Bartmess and Cerny, 1993, p. 84), they propose to break down a company’s or network’s strategy into its critical capabilities and to choose a location based on its impact on these capabilities. Therefore, they integrate external factors such as proximity to markets, customers, or suppliers, as well as internal interfaces between the different plants into the decision making process. Strategic Site Roles / Strategic Plant Roles The capability-driven perspective paved the way for the concept of strategic site / plant roles. Plant roles combine the idiosyncratic competencies of a plant with its strategic reason for or contribution to the network. One of the most popular approaches is the lead factory concept by Ferdows (1997a). He differentiates three strategic reasons for establishing and running a site; these are (1) access to low-cost production, (2) access to skills and knowledge, and (3) proximity to market. Combining these reasons with the competencies available at the site, he identifies six distinct plant roles as given in Fig.
26
UNDERSTANDING MANUFACTURING NETWORKS
11: source, offshore, contributor, server, outpost, and lead factory. According to Ferdows (1997a), plants intend to evolve in their strategic role, and it should be of central managerial interest to actively design and review their plant roles.
Site competence
High
Low
Lead
Contributor
Source Offshore Access to low-cost production
Server
Outpost Access to skills and knowledge
Strategic reason for the site
Proximity to market
Fig. 11: The roles of foreign factories (adapted from Ferdows (1997a))
Ferdows’ lead factory concept has been validated, tested, and modified quantitatively and in case studies by several other scholars (e.g., Vereecke and Van Dierdonck, 2002; Fusco and Spring, 2003; Meijboom and Voordijk, 2003). It has also served as “playground” for related studies. Meijboom und Vos (2004), for example, redefine the vertical axis of the original model, promoting a framework to track the dynamic evolution of a single plant. Feldmann and Olhager (2009a) shed some light on the development of site competencies and capabilities; studying a sample of more than 100 Swedish plants, they propose that competencies are typically assigned step-bystep as bundles to a site, starting with production-related, supply chain-related, and finally development-related competencies. Maritan et al. (2004) analyse the decision autonomy of a site regarding planning, production, and control mechanisms based on the plant role model of Ferdows (1997a). Moreover, Feldmann and Olhager (2009b) study the decision autonomy of plants depending on their strategic site reason and competencies. Based on the intra-network flows of innovation and people and the communication between sites, a different plant role typology is provided by Vereecke et al. (2006). The (1) isolated factory, (2) receiver factory, (3) hosting network player, and (4) active network player participate to a different extent in the “internal information and knowledge network”. Further, Vokurka and Davis (2004) analyse the product and process structure assigned to a plant identifying three types: (1) standardisers, (2) customisers, and (3) automators. They investigate differences in each of these types with regards to competitive priorities and performance levels.
27
UNDERSTANDING MANUFACTURING NETWORKS
2.2.2 Manufacturing Network Perspective Literature covering the network perspective widens the scope by considering the network as holistic entity or system. Shi and Gregory (1998), as well as Rudberg and Olhager (2003), provide an overview of the historical development of network theory in manufacturing. Research emerged from a pure site level, concentrating on the organisation of manufacturing operations on the shop floor, over a multiplant organisation, primarily focusing on location decisions but still treating plants as single entities, to a network organisation. But rather than being seen as an aggregation of loose plants, such an organisation has to be understood as integral system with linked nodes (plants). In line with this, Khurana and Talbot (1999) further note that plants both influence and are influenced by the whole network. Studying manufacturing networks, this system view needs to be sharpened by defining its boundaries and the prevailing management scope. Therefore, two research perspectives have been dominant: the operations management and the supply chain management perspective. In Fig. 12, Rudberg and Olhager (2003) summarise the consequences of this differentiation. “Internal networks”
“External networks”
Supply chain / Logistics management perspective
Manufacturing network / Operations management perspective Focus on the nodes
Focus on the links
Value network
Fig. 12: Perspectives on value networks (adapted from Rudberg and Olhager (2003))
While manufacturing network theory is based upon the operations management perspective focusing on the organisation of the plants, the network, and its coordination, supply chain management stems from the area of logistics focussing on the management of the material flows. Since the former is typically limited to an internal and fully owned network system, it primarily addresses the design of the nodes and their capabilities. The latter, on the other hand, widens boundaries by integrating external suppliers and customers, mainly addressing the (physical) links between the nodes (Rudberg and Olhager, 2003). 2.2.3 Discussing the Site & Network Perspectives Location decisions and strategic plant roles concentrate on a single site as focal entity. They leave aside considerations regarding a plant’s wider context and its embeddedness
28
UNDERSTANDING MANUFACTURING NETWORKS
into the network, i.e., the relations and interactions with the other entities. Research on manufacturing networks has to fill this gap; however, many aspects have been left unaddressed. Although, for instance, several scholars point out an evolution of plants changing their strategic roles (e.g., Ferdows, 1997a; Meijboom and Vos, 2004), there is no clear evidence of possible development paths or mechanisms driving this change. Further, except for some insular approaches 18, there is little research on how changing a plant role – intentionally or as a matter of evolution – affects individual sites or the network as a whole. This leads to the necessity to understand and evaluate plant roles from a superordinate network perspective. Instead of concentrating on a single role, network managers have to keep track of the collectivity of all entities in parallel, i.e., on the set-up and development of their so-called “plant role portfolio”. In this context, only the approach by Christodoulou et al. (2007) seems worth mentioning for proposing an operationalisation of such a portfolio. Four aspects of a plant role are defined by their approach: (1) its position in the internal supply chain / process stage, (2) the configuration and layout of the manufacturing processes performed at the site, (3) the geographic purpose or strategic site reason, and (4) the activities carried out by the site. These aspects are represented in a so-called “mountain model”, which basically is a graphical sketch of a pyramid which the sites can be arranged in. The model provides a sound starting point for an aggregated view on plant roles but neglects central aspects such as a plant’s contribution to the network. Moreover, the aspects on which its plant role definition is based on are not distinct, but overlapping. 19 As a consequence, clear guidance is still missing on how to best construct and balance a plant role portfolio to support manufacturing strategy, e.g., on how to decide about the optimal number and type of plant roles. Similarly, there are hardly any approaches to managing a network depending on the set-up of this portfolio. To sum up, manufacturing networks have been conceptualised as integral systems with linked nodes. Two distinct levels can be distinguished: the site level, represented by the individual plant management, and the network level, represented by the network or global operations management function, which is responsible for its development and optimisation. Both levels are inter-linked and affect each other. Thus, an integral perspective requires to take both levels into consideration and to align them. Implications from Manufacturing Site vs. Network Focus: Imp. 1.1: A manufacturing network is more than the aggregation of its individual sites. From a network level, sites cannot be managed in isolation but only in integration.
18 19
Some ideas are provided by Feldmann et al. (2010) and Cheng et al. (2011). For a visual sketch and detailed discussion of the “mountain model”, see Section 4.1.1 and Fig. 27.
UNDERSTANDING MANUFACTURING NETWORKS
29
Imp. 1.2: Network management, as superordinate authority, has to fulfil this task by integrating the site level and aligning it with the overall network targets and manufacturing strategy. Imp. 1.3: Thus, the tasks of the network management have to be defined precisely and supported systematically and methodically; tailored to its superordinate perspective.
2.3 Decision Categories & Layers of Manufacturing Networks 2.3.1 Manufacturing (Site & Network) Strategy Originating in Wheelwright and Hayes' (1985) four-stages model, the role of manufacturing has been accepted as source of strategic competitiveness. Manufacturing strategy can be studied on various levels, ranging from the company, over the business unit, division, factory, to even the shop floor. Miltenburg (2009) enhances this perspective by addressing manufacturing strategy for the network level, but he admits that a network itself can reflect a company or a subordinated unit. In line with this, a separation into manufacturing strategy, manufacturing site strategy (or capabilities), and network strategy (or capabilities) guides the following discussion. Manufacturing Strategy Manufacturing strategy can be divided into its content and its formulation process (Deflorin, 2007). Regarding the latter, Miltenburg (2009) provides a comparison of socalled process models for manufacturing strategy formulation, reviewing renowned contributions of Wheelwright (1984), Vickery (1991), Kim and Arnold (1996) and Platts et al. (1998). Briefly summarising his findings, manufacturing strategy sets the overall direction to support business strategy from the manufacturing perspective. Thereby, competitive manufacturing priorities have to be formulated to facilitate business strategy's competitive advantages. These priorities, in turn, are supported by the capabilities of the manufacturing function. In order to shape these capabilities, the decision categories of the manufacturing system have to be defined or adjusted. The content of manufacturing strategy is typically described by the characteristics of the competitive manufacturing priorities. These vary slightly according to the authors' perception, but they generally comprise the dimensions: cost, quality, delivery time and speed, reliability, and flexibility (e.g., Slack and Lewis, 2002; Miltenburg, 2005). 20 An accepted way to rank these dimensions’ importance is the “order winners / qualifiers” concept by Hill (1993). Order winners are competitive priorities for which 20
For a detailed discussion of manufacturing priorities, see Deflorin (2007).
30
UNDERSTANDING MANUFACTURING NETWORKS
an increase in performance has a positive impact on the business; hence, they enable a company to win the orders against its competitors. Qualifiers are needed to qualify for the business. They have to be met to a certain degree to gain the customers’ attention, but performance above that degree is not further appreciated. Manufacturing Site Strategy & Capabilities Manufacturing priorities are supported by manufacturing capabilities (Kim and Arnold, 1996; Miltenburg, 2009). Since research on production emerged from a single site’s perspective, designing the site capabilities was for long time considered as the main lever to influence strategy. Consequently, a plant’s capabilities are typically assessed by its performance on the manufacturing priority dimensions. Due to this, some scholars mix up terminology by not explicitly differentiating between strategic priorities and site capabilities. 21 Others come up with more concise views: Miltenburg (2008) distinguishes between manufacturing outputs, manufacturing levers, and manufacturing capabilities. Manufacturing outputs are defined by a site’s performance along the manufacturing priorities (what others have considered as its capabilities) whereas manufacturing levers are the structural and infrastructural elements of the manufacturing systems to shape the output level; manufacturing capabilities define improvement programs and best practices to adjust the levers. To conclude, typical site capabilities (in Miltenburg’s (2008) sense: the outputs) as discussed in literature are: 22 • Cost, as the site’s capability to control the financial input to manufacture the product (e.g., material, labour, overhead, and other resources). • Quality, as both the site’s capability to provide products whose features meet or exceed customers' specifications and expectations and its capability to assure on-going conformance to meet these specifications. • Delivery speed and reliability, as a site’s capability to meet or exceed the expected delivery speed and to keep delivery promises on-time and in-full. • Product range and design flexibility, as a site’s capability to produce a wide range and mix of products or to quickly implement design changes. • Order size and delivery flexibility, as a site’s capability to quickly change order sizes or delivery times. • Innovativeness, as a site’s capability to introduce novel products, processes, or products, or solutions which enable the customer to be innovative. 21
A recent example can be found in Hallgren et al. (2011) who model the competitive capabilities of plants along the classic dimensions of the manufacturing priorities (i.e., quality, delivery, cost, and flexibility). 22 Similar to competitive manufacturing priorities, there is no final agreement on the number and structure of site capabilities. For a discussion, see, for instance, Mapes et al. (1997), Colotla (2003), and Miltenburg (2008).
UNDERSTANDING MANUFACTURING NETWORKS
31
Manufacturing Network Strategy & Capabilities With rising awareness of manufacturing networks, some authors have complemented the site level by the impact of the network level on manufacturing strategy. Gulati et al. (2000) point out that “… strategic networks potentially provide a firm with access to information, resources, markets, and technologies: with advantages from learning, scale, and scope economies; and allow firms to achieve strategic objectives, such as sharing risks and outsourcing value-chain stages and organisational functions” (Gulati et al., 2000, p. 203). Shi and Gregory (1998) explicitly define four groups of strategic network capabilities, detailed by Miltenburg (2009) as “network outputs”. Built upon their work, but with some modifications of the original terminology, Tab. 4 summarises 15 capability items. Accessibility , as a network's capability to provide market proximity and access to resources of strategic importance Assure access to strategic markets and competitive factors, like …
Markets / Custom.
The network provides access / proximity to markets and customers
Competitors
The network provides access / proximity to competitors to fight them in their markets
The network enables to benefit from socio-political factors, such as to overcome trade barriers, Socio political factors to hedge exchange rate fluctuations, to exploit financial subsidies, ... Image
The network enables to benefit from image factors, such as "made in ..."
Supplier / Raw
The network provides access to suppliers, for example to assure local low cost supply, rapid delivery, high-quality raw-material, ...
material Assure access to Best cost labour resources of strategic Skilled labour importance, like …
External know-how
The network provides access to cheap work force The network provides access to high qualified work force The network provides access to external know-how, such as universities, competence clusters, engineering services, ...
Thriftiness ability , as a network's capability to achieve high economic efficiency Economies of scale
Increase efficiency Economies of scope by … Reduction of duplication
The network provides cost benefits by concentrating identical products The network provides cost benefits by concentrating products with similar manufacturing processes The network provides benefits by concentrating support functions, administrative functions, ...
Manufacturing mobility , as a network's capability to shift or transfer products, personnel, processes, or production volume to achieve flexibility and optimise resource utilisation Provide mobility of …
Products, processes, personnel
The network enables a flexible and fast transfer of products, production processes, machines, and personnel between the sites
Production volume & orders
The network provides a flexible and fast exchange of production volume between the sites or flexible order allocation
Learning ability , as a network's capability to foster external and internal learning Explore and External factors exploit know-how and innovation Internal factors about …
The network provides the possibility to unlock and share knowledge about external factors, such as local market needs and customer expectations, buying behaviour, ... The network provides the possibility to unlock and share knowledge about internal factors, such as product or technology innovations, best practices, …
Tab. 4: Strategic network capabilities (adapted from Shi and Gregory (1998) and Miltenburg (2009))
There is little research analysing how plant and network capabilities interact and how the latter impacts the manufacturing priorities. Only Colotla (2003) and Colotla et al. (2003) explore linkages between the two perspectives, concluding that both are interdependent and affect similar competitive priorities. Thereby, plant and network
32
UNDERSTANDING MANUFACTURING NETWORKS
capabilities can complement or offset each other. The authors propose a factorynetwork capability matrix as depicted in Fig. 13. Each diagonal line in the matrix visualises an isoline representing a certain degree of competitive advantage; competitive advantage (or disadvantage) gained by the factory level can be offset or even overcompensated by the network level, and vice versa. Based on these findings, they criticise that decisions are typically made independently of the network or plant level, calling for an integral process to combine both (Colotla et al., 2003).
Fig. 13: Factory-network capability matrix (adapted from Colotla et al. (2003))
Other authors link the network capability dimension with strategic management models and network taxonomies. Among these are Bartlett and Ghoshal (1989), who analyse companies' global networks regarding cost competitiveness (“global integration”), flexibility (“national responsiveness”), and innovation and know-how (“worldwide learning”). They identify three generic organisational types: (1) the global organisation striving for cost competitiveness, the (2) multinational organisation striving for flexibility, and the (3) international organisation striving for innovation and know-how. Arguing that changing environments require hybrid strategies, companies are urged to address capabilities in each dimension. Therefore, a “transnational solution” as blending of all three dimensions is introduced. Similarly, Shi and Gregory (1998) suggest companies to balance local integration and global dispersion. They propose seven generic manufacturing network configuration strategies and link them with the four groups of network capabilities. 23 Moreover, Ferdows (2009) puts forward two antithetic network strategies determined by the uniqueness of products and the standardisation of production processes. His “footloose network” is characterised by 23
For a detailed discussion of their approach, see Section 2.4.1 and Fig. 15.
33
UNDERSTANDING MANUFACTURING NETWORKS
commodity products and standard manufacturing processes, continuously searching for more efficient internal and external manufacturing alternatives. The “rooted network” has proprietary processes providing unique products. Long-term commitment and investment in competencies are necessary to maintain such competitive advantage. Finally, Miltenburg (2009) presents a framework to define a company's manufacturing network strategy based on the combination of renowned literature models. 24 2.3.2 Structural & Infrastructural Decision Categories Designing the manufacturing capabilities is realised by adjusting the decision categories of the manufacturing system (e.g., Platts et al., 1998; Miltenburg, 2009). Once more starting from the site level, operations management literature differentiates between structural and infrastructural decision categories to shape the factory manufacturing system (e.g., Hayes and Wheelwright, 1984; Slack and Lewis, 2002). Structural decisions are related to the “… physical configuration of the operation’s resources (… while infrastructural decisions comprise the …) activities that take place within the structure” (Colotla et al., 2003, p. 1187). Both categories can be broken down into distinct dimensions; in other words, designing the manufacturing system is realised by shaping its sub-systems. Tab. 5 provides an overview. Hayes & Fine & Hax Wheelwright (1985) (1984)
Structural categories Process technology Capacity Facilities Vertical integration Infrastructural categories Resource allocation and budgeting Human resources Organisation Quality Production planning and control New product development Performance measurement system
Authors
Hayes et al. (1988)
Samson (1991)
Miltenburg (1995)
Skinner (1996)
Hill (2000)
Slack & Lewis (2002)
Hayes at al. (2005)
Tab. 5: Manufacturing system’s decision categories & dimensions / sub-systems (adapted from Leong et al. (1990) and Rudberg and Olhager (2003))
The table reveals that there is no final agreement on how to divide the manufacturing system and its sub-systems. Miltenburg (2008) determines some requirements to assess such attempts: • Any division should be “comprehensive” in terms of all central decisions being covered by the sub-systems. 24
For a detailed discussion of his approach, see Section 2.4.1 and Fig. 16.
34
UNDERSTANDING MANUFACTURING NETWORKS
• Any division should be “discriminating” in terms of proposing distinct and analysable pieces with only little overlap. • Any division should be “reflective” in terms of being consistent with the manufacturing function’s own perception, meaning that the main issues of the operations management perspective are covered. Contrasting these requirements, Colotla et al. (2003) point out that most of the decisions in any sub-system have both structural and infrastructural implications, hence making any clear segmentation impossible. Though the concept of decision categories might be somewhat oversimplified, it provides a widely accepted and still useful way of categorisation (Slack and Lewis, 2002; Colotla et al., 2003). Although decisions in some of the sub-systems touch the manufacturing network, for example, the facilities and vertical integration (Rudberg and Olhager, 2003), they do not sufficiently cover that level. Consequently, shifting the focus to the manufacturing network 25, a second classification is proposed that separates between the network configuration and network coordination as central decision layers or categories (e.g., Porter, 1986; Rudberg and Olhager, 2003; De Toni and Parussini, 2010; Cheng et al., 2011). While configuration primary addresses structural decisions to physically design the network, coordination is mainly related to infrastructural linkages between the plants in the network (Colotla et al., 2003). Yet, compared to the structural and infrastructural decision categories from a plant level’s perspective, the literature shows even less agreement when it comes to defining the configuration and coordination layers and their related dimensions. 2.3.3 Manufacturing Network Configuration Layer From an operations management perspective, network configuration deals with the physical layout of the network, including decisions regarding the number and global expansion of sites, their assigned capacities, assets and technology, but also their competencies (e.g., Rudberg and Olhager, 2003; De Toni and Parussini, 2010); in other words, it deals with the network’s structure and specialisation. Network Structure Regarding the network structure, typologies have been mostly derived from the geographic dispersion of the network and the internal supply chain boundaries. As highlighted in the third column of Tab. 6, Stremme (2000), for example, distinguishes between four typical networks based on the internal material flows and the sites’ 25
Shi and Gregory (1998) call this perspective manufacturing network system.
UNDERSTANDING MANUFACTURING NETWORKS
35
production steps: monocentralistic, cross-linked, insular, and regional. In the regional network, each of the first three types can occur but is scaled down to (mostly independent) regions. Meyer and Jacob (2008), as shown in the second column, concentrate on economies of scale and scope and the necessity for local responsiveness. They propose five different types: world factory, local for local production, hub and spoke, sequential or convergent, and web-structure. Adding the product and market view, Schmenner (1982) introduces a second perspective on the network structure originating from a plant focus. Investigating 500 companies, he identifies four generic strategies for organising plants in the network: • Product plants are responsible for the manufacturing of (a set of) single products for the company's entire market; a strategy suitable for high volume products constrained to local resources or a product portfolio with very different product types. • Market area plants provide the full set of products for a dedicated market area or region; a strategy which is applicable for products requiring a high degree of local adaption or with a severe share of transportation and logistics costs. • Process plants provide a delimited set of manufacturing process steps targeting economies of scale and scope; a strategy recommended for either complex products divided into modules or for vertically integrated industries with manufacturing steps tied to local boundaries. • General purpose plants are highly flexible manufacturing sites with a comprehensive set of competencies; a strategy especially suitable for capacity balancing and products with short lifecycles. Other scholars, such as Khurana and Talbot (1999) and Hayes et al. (2005), built upon these strategies, detailing and complementing them with slightly different foci. Tab. 6 summarises and contrasts the two introduced perspectives: the (1) generic network types, primary built upon the geographic dispersion of the sites and their logistics connections, and (2) the multiplant strategies. The comparison illustrates the linkage of both perspectives. Each generic network type can be related to sites with a distinct focus following one of the introduced multiplant strategies. Only the volume strategy as proposed by Khurana and Talbot (1999) and Hayes et al. (2005) calls for clarification. In a volume strategy, sites have similar manufacturing competencies though these are tailored to a certain order size (e.g, mass production, large batches, and small batches). Customer orders can be allocated according to the most suitable production principle (Khurana and Talbot, 1999). This strategy, basically a special kind of general purpose plants, fits best to a web-structure allowing for global load balancing but with an additional focus on economic order allocation.
36
UNDERSTANDING MANUFACTURING NETWORKS Multiplant strategy / Plant focus Schmenner (1982) Khurana & Talbot (1999)
Generic network types
Hayes et al. (2005)
Product Process Product
Process Product stage / Line Techn.
Stremme (2000)
A world product is produced by a single site, markets are served by export.
World factory
Insular (concentrated)
Markets are served by local production sites tailored to the specific market requirements.
Local for local
Insular (deconcentrated)
The main share of manufacturing is conducted at a single hub. To meet market specifications, products are adapted at local spokes.
Hub & spoke
Cross linked
x
Cross linked
x
Monocentralistic
x
All sites have almost the full set of competence. Orders are balanced in the network to achieve a high degree of capacity utilisation.
Web structure
One of the generic types is scaled down and copied to (mostly independent) regions. Manufacturing site Process step
Volume
Market / Region
Prod. volume
x
x
Insular
Regional networks
General purpose
Process / Market / Techn. Geography
Meyer & Jacob (2008)
The production process is organised as a Sequential or sequential or convergent chain with production convergent steps conducted at different sites.
Market area
x
x
x
(x)
x
x
Markets served Internal flow of non-finished goods Delivery of finished goods
Tab. 6: Generic network types & multiplant strategies 26
Finally, it should be noted that the introduced types are idealistic. In practice, most networks neither have a clear multiplant strategy, nor do they exclusively match with one of the proposed network types. Instead, hybrid forms are common (Colotla, 2003). Network Specialisation The structure sketches the set-up of the network and assigns the plants’ general purpose and area of responsibility. This view is enhanced by decisions regarding the allocation of technology, assets, and equipment, all driving a site’s capabilities, or, in other words, by the network specialisation. From a network manager’s perspective, questions related to the network specialisation can be synthesised using the definition of plant roles, i.e., combining the sites’ competencies with their strategic contribution. As shown in Section 2.2.1, scholars provide different plant role typologies, mostly based on the groundwork of Ferdows (1997a). But, as also highlighted, a superordinate 26
The illustration has been adapted from Meyer and Jacob (2008).
37
UNDERSTANDING MANUFACTURING NETWORKS
network perspective is still lacking. So are approaches to design and manage a company's plant role portfolio rare. Ferdows (1997a) himself asserts that the role of a plant is of dynamic nature, striving for an evolution towards increasing competencies and a more dominating position; a progress that has to be monitored carefully otherwise it entails the risk towards an imbalanced network. Vereecke and De Meyer (2009) complement the findings of Vereecke et al. (2006) with implications for the design of a balanced portfolio. They promote that removing a plant or changing its role might affect the overall network equilibrium. In line with this, Feldmann et al. (2010) analyse the impact of changing a plant role from a network perspective in a longitudinal study, but they do not elaborate on how to track or leverage such changes. Moreover, Cheng et al. (2011) make an effort to investigate how the evolution of a manufacturing plant affects other sites and the plant role portfolio in general. The model depicted in Fig. 14 outlines their assumed interactions. Capabilities of manufacturing network & performance: efficiency and effectiveness
Situation for future transformation
Degree of manufacturing network coordination evolving
4 3
Portfolios of plants, i.e. manufacturing network configuration transforming gradually
Plant / network characteristics: 1) Dynamic of product / process / knowledge 2) Dynamic of plant capability 3) Dynamic of location advantage 4) Dynamic of network strategy
Transfer of products, processes and knowledge between plants
1 2
From manufacturing system level to plant level
Interactive plant evolutions; strategic roles of plants transforming
From plant level to network level
Fig. 14: Overview of the interactive evolutions of plants & the manufacturing network (adapted from Cheng et al. (2011))
Four types of dynamics induce a need to transfer products, processes, and knowledge between plants. These are: (1) new and modified products, processes, and knowledge, (2) emerging plant capabilities, (3) changing location advantages, and also (4) adaptations of the network strategy. Any such modification triggers an evolution of a plant’s role. This evolution impacts the overall plant role portfolio, leading to a gradual transformation of the network configuration, which, in turn, requires an adjustment of the network coordination mechanisms. Cheng et al.’s (2011) findings
38
UNDERSTANDING MANUFACTURING NETWORKS
accentuate the interplay between the configuration and coordination layer and the need of an analytical “… view and contingent thinking on the relevant factors of the plant, the network and even the company in order to make relevant decisions about the transformation of the manufacturing network” (Cheng et al., 2011, p. 1328). Yet, despite postulating an integral approach to decision making on site and network level, they fail to provide any mechanisms or tools for analysis and guidance, e.g., for the support of the systematic design and improvement of the plant role portfolio. To conclude, the network structure is close to what is commonly termed the “global manufacturing footprint”; it addresses the geographic dispersion and the sites’ general purpose mainly from the logistics perspective. The network specialisation, on the other hand, reflects the operations management’s understanding of configuration; it covers the design of the network nodes itself, i.e., their competencies and contributions. For this, an integral perspective is crucial. Based on the quantitative literature screening in Section 2.1.2 and the subsequent discussion, scholars provide a sufficient number of configuration typologies and frameworks for the network structure and the plant role definition. According to Rudberg and Olhager (2003), this is owed to the configuration layer’s origin in former location decision making and multiplant design. Nevertheless, configuration approaches handling the network as integral system – such as the design of an (optimal) plant role portfolio – leave room for enhancement. 2.3.4 Manufacturing Network Coordination Layer The model proposed by Cheng et al. (2011) points at the linkages between a network’s configuration and its coordination. Coordination, as the second decision layer, refers to the question of how to organise, link, and integrate the production facilities in order to achieve strategic business objectives (Meijboom and Vos, 1997; Cheng et al., 2011). It addresses the design and management of the physical and non-physical flows between the network's facilities, but it also encompasses the more institutional design and establishment of rules and mechanisms for the interaction between the plants themselves as well as between plants and the headquarters or central network management functions respectively. Coordination of Network Flows Coordination is often understood as the design, planning, and management of distinct flows between sites in the network (Bartlett and Ghoshal, 1989; Cheng et al., 2008; Cheng et al., 2011). Four types of flows are typically discussed: the flow of material / physical goods, information, people, and financial resources (Bartlett and Ghoshal, 1989; Vereecke et al., 2006). Literature tackles these flows to different extents. The
UNDERSTANDING MANUFACTURING NETWORKS
39
mostly tactical management of the production planning and material flows is a central part of logistics and supply chain management, and research is often based on material flow simulation and mathematical optimisation models (Pontrandolfo and Okogbaa, 1999; Vereecke et al., 2006). In this context, Bhatnagar and Chandra (1993) give an – although somewhat outdated – overview of models to solve “general coordination issues”, covering supply and production planning, production and distribution planning, inventory and distribution planning, as well as “multiplant coordination issues”, which link and align the production plans of scattered manufacturing sites. While from the operations management perspective the financial flow is often completely neglected, several authors point out the design and coordination of information and knowledge flows as key managerial task (Chew et al., 1990; Ferdows, 2006; Vereecke and De Meyer, 2009). Ferdows (2006), for instance, identifies a typology of production know-how based on the level of codification and the change rate of knowledge, and he assigns appropriate transfer mechanisms to each type. Vereecke et al. (2006) develop their typology of plant roles based on the intra-network flows of innovation and people and the communication between sites. Vereecke and De Meyer (2009) show that the more intensively a plant is embedded into such an internal “knowledge network”, the less likely its closure is. Similarly, Tsai (2001) identifies the network position of a plant – defined as its access to knowledge – and its absorptive capacity as determinants of the creation of innovation, and thus as a basis for its performance. Case studies in this context are provided by Dyer and Nobeoka (2000), who focus on the coordination of Toyota's external “knowledge network”, and by Rudberg and West (2008), who analyse the coordination mechanisms within the Ericsson network. Especially the latter convincingly argue that coordination needs a more institutional perspective, addressing the need of rules, guidelines, and standards as facilitators. Institutional Perspective on Coordination Porter (1986), as one of the first to come up with the separation of configuration and coordination, tightly connects the question of how to coordinate activities performed at dispersed facilities with the degree of autonomy each production facility carries. Several scholars have picked up on the issue of balancing decision responsibility between plants and central headquarters. Feldmann and Olhager (2009b) and (2011) recently tested for autonomy of selected decisions based on a sites competencies 27 and its strategic site reason. They identified three different structures for strategic decision
27
Their research is based on the findings of Feldmann and Olhager (2009a) who assert that competencies are assigned sequentially to sites in clusters related to production, supply chain, and finally development.
40
UNDERSTANDING MANUFACTURING NETWORKS
making in manufacturing networks: a (1) centralised, (2) integrated, and (3) decentralised one. These structures depend on a site’s competence level but not on the strategic reason for its location. Their research follows Maritan et al. (2004) who, again based on Ferdows' (1997a) plant roles, highlight decentralisation of decision making as main indicator for site autonomy and use it to evaluate differences in the autonomy for planning, production, and control decisions. In their analysis, Maritan et al. (2004) also refer to standardisation as a factor influencing autonomy; though, they do not go in detail. Likewise, Meijboom and Vos (1997) relate the technical competence of a site to its autonomy. In contrast to others, they evaluate only the degree of standardisation of production processes, quality control, and R&D in the network; hence, they reveal standardisation as a second dimension of autonomy. Summing up, compared to configuration, research on coordination is both quantitatively (as shown in Fig. 10) and qualitatively (as shown in the discussion) lagging behind. If at all, coordination is studied from isolated perspectives that lay (too much) stress on knowledge exchange; but there is no consensus on its elements and dimensions and on how they interact. There is also little research regarding general description and design models for coordination in manufacturing networks. 2.3.5 Discussing the Decision Categories & Layers Generally, supporting the manufacturing strategy is realised by addressing changes in the decision categories of the manufacturing system (e.g., Wheelwright, 1984; Vickery, 1991; Platts et al., 1998). On a plant level, scholars differentiate between structural and infrastructural decision categories, providing several underlying dimensions (e.g., Slack and Lewis, 2002; Hayes et al., 2005; Miltenburg, 2008). As to manufacturing networks, the separation into the two network layers, configuration and coordination, as superordinate decision categories is accepted. Hence, altering the network capabilities is realised by designing and modifying its configuration and coordination. Compared to research on the site level, however, there are very limited contributions clearly defining the scope and characteristics of these two layers and their underlying decision dimensions. Amongst those is Shi and Gregory’s (1998) attempt to translate the traditional structural and infrastructural categories to the network level, thus contrasting the manufacturing system with the manufacturing network system. Further, the X-GO model, as introduced in Section 1.2.1, is based upon the concept of decision layers, dimensions, and variables. Yet in both cases, the classification is of descriptive power only, lacking a deeper understanding of the layers and their operative feasibility. To go one step further, even though decisions on both layers are usually presented separately, scholars are aware that both are connected. While Pontrandolfo and
UNDERSTANDING MANUFACTURING NETWORKS
41
Okogbaa (1999) highlight that “… the two aspects of configuration and coordination are strictly related” (Pontrandolfo and Okogbaa, 1999, p. 5), Meijboom and Vos (1997) state that “… a configuration decision, in turn, leads to a certain form of coordination” (Meijboom and Vos, 1997, p. 803). Rudberg and Olhager (2003) even assert that “… coordination of activities within the network is contingent upon the configuration” (Rudberg and Olhager, 2003, p. 29) and Cheng et al. (2011) conclude that “… since the configuration of a manufacturing network changes, the coordination mechanisms need to be redeveloped” (Cheng et al., 2011, p. 1327). When focusing on configuration, literature is rich in models and typologies particularly for the network structure, i.e., the global footprint design, and the definition of single plant roles. Yet, there is still a need for an integral network perspective. When focusing on coordination, three central limitations can be identified. • First, as revealed by the literature screening and discussion, coordination is generally underrepresented in research on manufacturing networks. Whereas configuration has a long history and is frequently subject of operations management literature, less attention has been devoted to coordination issues (Pontrandolfo and Okogbaa, 1999; De Toni and Parussini, 2010), especially from an operations management perspective. • Second, although coordination of globally dispersed but interdependent plants is accepted as key managerial task (e.g., Porter, 1986; Shi, 2003; Cerrato, 2006), studies have seldom used a systematic approach to address it (Pontrandolfo and Okogbaa, 1999). Instead, most scholars provide only selected contributions focusing on limited aspects and interactions. • Third, although configuration and coordination decisions are tightly interrelated, interactions and interfaces are neither widely understood nor even investigated (Meijboom and Vos, 1997; Pontrandolfo and Okogbaa, 1999; Rudberg and Olhager, 2003). The application of conceptual models and classification schemes (e.g., Rudberg and West, 2008), but also of management tools, frameworks, and integral processes, might prove a promising attempt to fill this gap. Implications from Network Manufacturing Layers: Imp. 2.1: Manufacturing strategy is supported by site and network capabilities. Hence, shaping the network capabilities determines the key task of the network management; it provides a lever for competitive manufacturing advantage. Imp. 2.2: Shaping the network capabilities is achieved by designing and modifying the network's configuration and coordination layer. Vice versa, the design of the
42
UNDERSTANDING MANUFACTURING NETWORKS
configuration and coordination layers has to be made in accordance with the aspired manufacturing strategy. Imp. 2.3: In order to design and modify the network's configuration and coordination layers, and thus to shape the network capabilities to support the manufacturing strategy, the underlying decision dimensions of both layers need to be known and understood precisely. Imp. 2.4: Configuration and coordination cannot be addressed in isolation, but coordination needs to be considered contingent upon configuration. Hence, not only the decision dimensions but also their linkages have to be taken into consideration. Imp. 2.5: In order to support the network management in decision making, approaches are necessary operationalising and integrating the decision dimensions into a systematic and holistic management approach.
2.4 Network Design & Management Approaches The literature discussion is completed by a review of approaches to, and tools and methods for network analysis, design, and optimisation. The spectrum ranges from (1) management frameworks assisting in decision support for a single layer to (2) process models targeting a network's systematic development and improvement. Often, the distinct approaches are not independent but build upon or complement each other. 2.4.1 Network Management Frameworks Management frameworks either concentrate on a distinct decision layer or single decision dimensions, or they are integrative, combining configuration and coordination aspects with strategy. Since the former have already been touched when introducing the separate layers, the following discussion focuses on the latter targeting an integral management view. Shi et al. (1997) and Shi and Gregory (1998), as outlined in Fig. 15, introduce a map of international manufacturing network configuration. The map is based on two dimensions: the (1) geographic network dispersion, which ranges from national to worldwide and describes the expansion of operations to design, produce, and sell goods on a global scale, and the (2) degree of coordination between the scattered manufacturing sites. Starting with domestic manufacturing, seven additional generic strategies 28 are derived and linked with the four previously discussed network capabilities: accessibility, thriftiness, mobility, and learning. Each configuration is
28
For a detailed classification of each strategy, see Shi and Gregory (1998).
UNDERSTANDING MANUFACTURING NETWORKS
43
enhanced by an ideal network pattern. The patterns, termed “capability profile grid”, comprise distinct characteristics to assess the network capabilities. They are appropriate as benchmarking profile for evaluating a company's actual network capabilities. Although mentioned explicitly, the framework’s link to coordination is weak; it is only addressed in terms of a low or high degree of coordinative effort. No support is given on how to specify the degree of coordination for a certain strategy or on how to design appropriate coordination mechanisms.
Fig. 15: Framework for international manufacturing network configuration & capability profile grid (Shi et al. (1997))
Consequently, Miltenburg (2009) substitutes the coordination axis by what he calls pressure for local responsiveness, i.e., the necessity to adapt the manufacturing activities and practices to local needs and requirements. As depicted in Fig. 16, he promotes his own holistic model for manufacturing strategy setting in international manufacturing networks.
44
6
4
5
1&2
3
UNDERSTANDING MANUFACTURING NETWORKS
Fig. 16: Manufacturing strategy framework for a manufacturing network (adapted from Miltenburg (2005) and (2009))
The model links six distinct frameworks, so-called “objects”, which are well-founded in operations management literature. Most of them have been touched in the course of this chapter: (1) the seven generic manufacturing network strategies adapted from Shi
UNDERSTANDING MANUFACTURING NETWORKS
45
et al. (1997) and Shi and Gregory (1998) and (2) their related manufacturing network types (Shi and Gregory, 1998; Miltenburg, 2005), the (3) four manufacturing network capabilities / outputs (Shi and Gregory, 1998; Gulati et al., 2000), the (4) structural and infrastructural network levers / decision categories (Wheelwright and Hayes, 1985; Shi and Gregory, 1998) as well as their (5) capability level (Miltenburg, 2005; Miltenburg, 2008), and finally (6) the factory types / plant roles (Ferdows, 1997a; Miltenburg, 2005). Based on three case studies, he applies the model to map the current state of the corresponding networks and to derive implications for their future development. Although linkages between the individual objects are of conceptual and descriptive nature only, this approach gives a unique example of how to combine single frameworks to an integral management model. Nonetheless, similarly to the international manufacturing network configuration map by Shi et al. (1997) and Shi and Gregory (1998), coordination aspects are mostly neglected. Probably aware of this drawback, Miltenburg (2009) points out that “… there are other areas where more research can be done (…) New objects and other frameworks can be developed and relationships between these can be studied” (Miltenburg, 2009, p. 6200). Among the sparse work addressing this gap by covering coordination aspects, Rudberg and West (2008) have to be mentioned, although they only focus on a descriptive single case study of Ericsson's manufacturing network coordination model. To sum up, framework approaches are valuable to understand distinct network mechanisms. Nonetheless, most approaches are rather exploratory and lack a systematic and procedural character to analyse, design, and improve the network. Moreover, the network coordination layer is neglected or addressed only superficially by them. 2.4.2 Network Design, Management & Optimisation Approaches While management frameworks cover the content of manufacturing networks, design and optimisation approaches add a procedural component. Two different types shall be distinguished: (1) strategic approaches mainly based on workshops and qualitative discussion with key persons to design future network scenarios and (2) quantitative approaches based on detailed data analysis and mathematical programming to evaluate and concretise those scenarios. Strategic Approaches Strategic approaches typically integrate single management frameworks into a structured method. Colotla (2002) and (2003), for instance, proposes an approach for
46
UNDERSTANDING MANUFACTURING NETWORKS
global manufacturing strategy definition. It incorporates the bricks of a generic manufacturing strategy formulation process as put forward by Hax and Majluf (1995), breaking down the business strategy into a manufacturing strategy and aligning internal strengths and weaknesses with external opportunities and threats to support this strategy. To assure the integration of site and network capabilities, analysing internal strengths and weaknesses is recommended to be based on the factory-network capability matrix as depicted in Fig. 13. The process is purely concentrated on strategy formulation, not touching any aspects of network design and improvement. Further, it is of conceptual character only, hence lacking any proof of practical applicability. The configuration map by Shi et al. (1997) and Shi and Gregory (1998), as referred to in Fig. 15, also provides an anchor for network optimisation. Shi (2003), based on Shi et al. (2001), integrates the map as a tool for a systematic assessment of the network's configuration and capabilities into a generic global manufacturing strategy process; contrary to Colotla (2002) and (2003), here, also design aspects are addressed. The approach is based on four interrelated modules: • Identifying the requirements of globalisation, including a structured analysis of the products, markets, customers, and competitors, an evaluation of future trends, and a discussion of the current and targeted manufacturing strategy. • Assessing the current network and its capabilities, comprising the analysis of the current network configuration based on the introduced configuration map and the qualitative assessment of the network capabilities based on the “capability profile grid”. • Identifying the manufacturing mission and the network design, addressing the formulation of the future network mission, the choice of the aspired network configuration, and the design of the underlying decisions in terms of product allocation, dispersion, and coordination mechanisms. • Fostering the network transformation based on the derivation of implementation plans, the initialisation of project management initiatives, and the transformation of the factories to support the network strategy. Uniquely to this approach, a workbook serves as toolkit, with tailored worksheets underlying each module. 29 The worksheets, basically sheets for qualitative data collection and single frameworks, facilitate both the academic research and the
29
The description of the workbook approach is based on journal contributions and conference papers by Shi et al. (1997), Shi and Gregory (1998), Shi et al. (2001), and Shi (2003). Unfortunately, access to the complete workbook was restricted. Thus, it was not possible to gain insights deep enough to adequately evaluate the single frameworks and worksheets. Nonetheless, most of these constructs are assumed to be very similar to the approach by Christodoulou et al. (2007) as discussed in the following. Both concepts were developed in Cambridge while Christodoulou et al. (2007) heavily relied on the available groundwork of their colleagues.
UNDERSTANDING MANUFACTURING NETWORKS
47
practitioner's design process since they provide structure and traceability of the information and data collected, offer system and guidance for discussion, and ensure a combination of theoretical exploration and practical validity (Platts and Gregory, 1990; Shi and Gregory, 1998; Shi, 2003). In this context, the workbook is understood as action research tool (Shi, 2003). The workbook approach also lays the foundation for a second method presented by the University of Cambridge's Institute for Manufacturing (ifm). Titled “making the right things in the right place”, a step-by-step process is promoted, guiding operations managers through the challenge of (re-)designing their network. The so-called “Cambridge approach” covers four iterative phases (Christodoulou et al., 2007; Grallert et al., 2010): • The “Why” phase strives for a common understanding of the motivation and need for modifying the network. Internal and external drivers for the change are mapped and a future network vision is designed in joint workshop sessions. • The “What” phase addresses the make-or-buy decisions. It comprises the identification of the current make-or-buy strategy, a risk assessment, and the establishment of guidelines to clarify and communicate the future strategy. • The "Where" phase creates the design of the future network configuration. This most comprehensive phase starts with the definition of a "common framework" to set the preconditions for analysis and involves the creation of a joint language with regards to variables and key assumptions. Second, network options are designed by (1) evaluating the current and defining the future plant role portfolio based on the “mountain model” as mentioned in Section 2.2.3, by (2) adding coordination mechanisms, and finally by (3) deriving promising configuration options. The options’ design is aided by the plant role portfolio, in which the aspired network specialisation is sketched. Third, these options are assessed according to their impact on network capabilities and cost dimensions. Fourth, the future network configuration, which is often established for a global product line, is put into a company-wide context by discussing the interactions with other internal networks, hence striving for an integral global solution. • The “How” phase initiates the change by mobilising the stakeholders, works out the details, such as exact location decisions and the organisation of the product transfer, and finally aims at closing the loop by measuring the success of the transformation. “Cambridge's approach” enhances the workbook process by Shi (2003) and Shi et al. (2001). It is one of the most elaborated approaches combining academic concepts with practical experience. Primary targeting the design of network options with qualitative
48
UNDERSTANDING MANUFACTURING NETWORKS
tools, most of the work is carried out in workshops and interviews with the top management, striving for commitment and common understanding rather than for a detailed evaluation of these options; this is left to the company's experts (Grallert et al., 2010). The process covers a wide range of strategic considerations, which tailor it to the needs of the network management. Nonetheless, according to Grallert et al. (2010), it leaves some room for enhancement, especially with regards to measuring the success in the "How" phase. While performance indicators on the plant level are wellknown, a definition of measures on the network level is lacking. Further, although explicitly pointed out as part of the approach, the determination of coordination principles in the “Where” phase is rather superficially. 30 It is based on a simple matrix allocating selected decision responsibilities between centralised and decentralised functions (Christodoulou et al., 2007). Quantitative approaches Quantitative network optimisation approaches focus on a detailed evaluation of selected network options / scenarios. A classification is given by Jacob (2006), who differentiates between static and dynamic methods and between checklists, economic business calculations, model based simulation, and optimisation techniques: 31 • Checklists are utilised to assess network options along a set of (weighted) criteria. They can be executed with very little IT support, thus bridging the gap between quantitative and qualitative evaluation. • Economic business calculations analyse the investment decisions for selected network options concentrating on cash-flow considerations. • Simulation approaches build upon the design of a model to evaluate the behaviour of a selected target dimension. Simulation can be supported by spread sheets, system dynamic, discrete event simulation, and business games (Kleijnen, 2005). • Optimisation approaches extent simulation models by an algorithm typically maximising or minimising the target dimension. A recent evaluation of these approaches – mainly for simulation and optimisation techniques – can be found in Justus (2009) and Ude (2010); the latter compares 21 approaches that have emerged since 1995. He further proposes a novel process promoted by the Institute of Production Science at the Karlsruhe Institute of 30
Since both concepts were developed in Cambridge, and Christodoulou et al. (2007) heavily rely on the groundwork of their colleagues, the assumption that these two dimensions are not sufficiently elaborated by the workbook process of Shi (2003) and Shi et al. (2001), too, is reasonable. 31 For a description of the different approaches, see Jacob (2006), Justus (2009), and Ude (2010).
UNDERSTANDING MANUFACTURING NETWORKS
49
Technology (Lanza and Ude, 2010; Ude, 2010). A detailed positioning of his method against “Cambridge's approach” can be found in Grallert et al. (2010). As most quantitative approaches, the “Karlsruhe approach” mainly focuses on the “Where” phase of the network configuration. This phase is backed by an initial sketching of potential network scenarios and a detailed evaluation of different options based on discrete-event simulation and a quantitative and qualitative multi-criteria analysis. Comparison of the approaches Tab. 7 gives an overview of the four discussed network design and management approaches. It has been restricted to these methods for the following reasons: First, although there might be other network design processes – primary as part of the toolkit of most consultancy companies – the scope shall be limited to those approaches that gained academic attention for contributing to the scientific knowledge base. Second, there are far more quantitative processes than the Karlsruhe one, 32 but they are similar in their goals and procedure, varying mainly in the technical elaboration. Hence, it is consider sufficient to use the “Karlsruhe approach” as one of the most recent representatives for quantitative methods. Tab. 7 is split into five areas: • Area a) structures the comparison along generic process steps. It consists of the four main phases of the “Cambridge approach” (“Why”, “What”, “Where”, “How”) and the underlying generic steps as proposed by Grallert et al. (2010). • Area b) specifies and sorts the individual process steps of each of the four approaches according to the predefined structure. • Area c) evaluates to what extent a distinct step of a certain approach tackles the previously discussed elements of the manufacturing network (management). These are basically the network layers complemented by contextual factors, such as markets, environment and competitor, and the network performance. The evaluation ranges from only addressed as important but not elaborated (o), over covered and partially elaborated (+), to fully covered and explicitly elaborated by distinct tools or frameworks (++). It reflects this study’s author’s personal opinion. • Area d) evaluates and classifies the methods proposed by the single approaches. It differentiates between project management (PM), qualitative methods (QL), such as worksheets and management frameworks, as well as quantitative methods (QT) like simulation or programming. • Area e) aggregates results of the comparison. 32
For an overview and discussion, see Jacob (2006) and Justus (2009).
50
UNDERSTANDING MANUFACTURING NETWORKS
The comparison outlines the characteristics of the different approaches. Most of them start with a general analysis of the contextual environment of the company, but they differ in their priorities for the further proceeding. Colotla (2003) basically focuses on the (re-)formulation of the manufacturing (network) mission and strategy as well as on the design of the network and site capabilities. Therefore, he emphasises the need to integrate site and network capabilities but only gives conceptual guidance on how to do so. The “Karlsruhe approach” by Ude (2010) – as representative of the quantitative approaches – is restricted to configuration issues, using mathematical simulation to evaluate the best footprint options, but it mostly neglects strategic considerations. Both Shi (2003) and Christodoulou et al. (2007) provide a more comprehensive view, claiming to cover the network strategy, configuration, and coordination layer by using qualitative methods, i.e., frameworks and worksheets. Upon closer examination, however, they too focus mainly on configuration. Shi (2003), for instance, bases his process on the network configuration map by Shi et al. (1997) and Shi and Gregory (1998), adding only some general considerations about the degree of network coordination. The “Cambridge approach” by Christodoulou et al. (2007) is the only one to cover network specialisation by integrating the concept of plant roles to the configuration layer. The plant role portfolio is utilised as starting point to derive footprint options and to align coordination principles. Nonetheless, the coordination layer is restricted to the allocation of decision autonomy for some basic areas of responsibilities, leaving aside any other considerations regarding the institutionalisation and the management of the network flows. Further, none of the approaches provides sufficient guidance on how to evaluate and monitor network performance. “Cambridge’s approach” claims to assess the future footprint options based on the impact on network capabilities; actually, it gives little guidance on how to do so. Shi (2003) uses the “capability profile grid” for the assessment of the network capabilities, but all this is done on a qualitative scale only. Also, the post project evaluation of the success of the transformation has not been resolved yet, constituting the need for KPIs tailored to an evaluation of the network level. “Karlsruhe’s approach” indeed relies on a quantitative assessment of the performance of different network option by using simulation, but it is mainly concentrating on cost and operational performance, thus neglecting any further considerations regarding strategic network capabilities (like accessibility, manufacturing mobility, or learning).
7 Deriving footprint options
6 Analysing & determining coordination principles
4 Creating a common framework for analysis (variables & assumptions) 5 Analysing & designing plant roles
Colotla (2002) and (2003)
2 Formulating global mfg. strategy and scenarios
1b Analysing network and factory capabilities
1a Analysing business strategy & scanning the environment
B
o adressed but not elaborated + covered and partially elaborated ++ covered and explicitly elaborated
not adressed
14 Embedding the new process
12 Transition Implementation 13 Measuring the success
PM = Project management QL = Qualitative approach QT = Quantitative approach
4 Detailed programming and budgeting
Assess 8 Assessing footprint options based on 3a Evaluating scenarios based on options cost & network capabilities financial and non fin. impact Aggregate 9 Aggregating product or business unit solution strategies into company vision Analyse 10 Testing robustness of the options 3b Evaluating scenarios according robustness based on contextual changes to sustainability and risk 11 Mobilising for change
Identify network options
Preconsideration
buy analysis
1 Mapping the strategic and environUnderstand mental context for the network underlying 2 Embracing the change (burning motivation platform, right people, etc.) Make-or- 3 Defining make-or-buy strategy
Christodoulou et al. (2007) "Cambridge approach"
4 Transforming network config.
3 Identifying mfg. mission and designing network config.
2b Assessing current network configuration 2c Analysing network capabilities
0 Introducting international mfg. network and strategy
Ude (2010), Lanza & Ude (2010) "Karlsruhe approach"
Cambridge Approach Colotla Shi bzw. Shi & Gregory Karlsruhe Approach
A B C D
e) Aggregation of comparison
4 Testing robustness
2 Designing global mfg. network options 3 Evaluating options by simulation and multi-criteria analy.
1 Determining multidimensional target system
Shi et al. (1997), Shi & Gregory (1998), Shi et al. (2001), Shi D (2003)
1 Identifing requir. of globalis. 2a Analysing prod. and markets
C
Strategy Site capab.
Network Structure Flows
Institutional Persp.
Coordination
Network Specialisation
Configuration
c) Elements of manufacturing network (management) Markets, Business techn., Network & mfg. environ., capab. strategy compet.
+
A
++ o + o
Why
What
Context
+ o + o o
Quantitive approaches
+
o + o o
+ ++ ++
+
+
Where
How
+ + o
+ o
++ +
+
o ++ ++ + o
o o
o + o
Strategic approaches
d) Method
Network PjM, QL,QT Perform.
Perform.
o o
o +
+ + o
QL PM QL QL QL QL QL
+
QL QL QL
b) Network design, management & optimisation approaches
QT
QL
+
QL QL PM QL/QT
QL QL QL/QT QT
QL QL/QT QL
++ o + o
+ o + o
+ + ++
o + +
+ o ++ ++
+ o
o
+ o o
o + + +
QL QL QL QT
A B C D A B C D A B C D A B C D A B C D A B C D A B C D A B C D A B C D A B C D
a) Generic steps
UNDERSTANDING MANUFACTURING NETWORKS
51
Tab. 7: Overview of selected network design, management & optimisation approaches
52
UNDERSTANDING MANUFACTURING NETWORKS
2.4.3 Discussing the Design, Management & Optimisation Approaches Concluding, quantitative optimisation approaches have a long history especially within the scientific areas of operations research, mathematics, and engineering. They heavily rely on the availability and quality of data and on the expert knowledge of the persons involved in creating models. If these preconditions are met, their results provide valuable implications and a high depth of detail for the network design. Nonetheless, for the sake of quantification, these approaches have a strong cost and time focus, tending to impede the view on superordinate strategic questions. Strategic approaches can bridge this gap. They are less dependent on the availability and quality of data than on the commitment and participation of the network's key decision makers. Rather than enabling a detailed evaluation of possible scenarios or options, they can be used to map the current network state and to create potential design alternatives. This tailors them to the needs of the network management in the conceptual design phase, which calls for a high degree of information condensation and for tools rather stimulating and fostering strategic thinking instead of hindering it with too much detail. Management frameworks and worksheets, in turn, seem to be promising to underline these approaches. Moreover, a detailed quantitative analysis can be used complementary to evaluate the derived conceptual options. Again, the literature review reveals some major limitations. Generally, qualitative strategic approaches for network analysis and design are rare compared to quantitative methods, and little is known on how they succeed in practice when supporting the network management in decision making. Second, regarding the structure, there is no common understanding on how such approaches should be carried out and in what sequence the distinct steps shall be applied best; nor is it clear if a strict proceeding is realistic at all. Third, regarding the content, most of the introduced frameworks and worksheets remain oversimplified, missing any proof of academic validity and practical applicability. Further, among the discussed single frameworks and strategic approaches, the configuration layer is clearly dominating. The coordination layer is neither sufficiently addressed by frameworks or tools for analysis and improvement nor – or only basically – integrated into the existing strategic network design and management approaches. A systematic management support is missing. Implications from Manufacturing Network Management: Imp. 3.1: Network design and management approaches provide systematic decision support. They can be of both quantitative and strategic nature. Strategic approaches are rather focused on the derivation of visions and scenarios based on the creation of a common understanding and commitment; thus, they are more tailored to the network management's strategic perspective.
UNDERSTANDING MANUFACTURING NETWORKS
53
Imp. 3.2: Strategic approaches are typically based on selected management frameworks and worksheets. Currently, literature provides only a very limited number of these approaches; evidence for their operationalisation and practical applicability are scarce. Imp. 3.3: The few identified strategic approaches clearly focus on the network configuration layer. Although of key managerial and academic interest, the coordination layer is either neglected or only addressed very basically. Imp. 3.4: Thus, the need for a holistic network architecture and a strategic approach to support the network management – especially in the management of the coordination layer – is evident. Such an approach cannot be designed in isolation but is constrained by the network configuration (and strategy).
2.5 Summary & Discussion 2.5.1 Implications from Literature A comprehensive literature analysis on manufacturing networks was conducted in the course of this chapter. Selected contributions were discussed along three dimensions: • The two focal units: manufacturing site and manufacturing network. • The network decision layers: strategy, configuration, and coordination. • The practical viewpoint on frameworks and integral approaches for network design and management. Implications were given for each of the dimensions. Combining the distinct implications allows for a statement on how the manufacturing network (management) will be understood in the course of this study; it also manifests the storyline and mission of the further work in order to answer the raised research question. Combining the Implications Imp. 1.1: A manufacturing network is more than the aggregation of its individual sites. From a network level, sites cannot be managed in isolation but only in integration. Imp. 1.2: Network management, as superordinate authority, has to fulfil this task by integrating the site level and aligning it with the overall network targets and manufacturing strategy. Imp. 1.3: Thus, the tasks of the network management have to be defined precisely and supported systematically and methodically; tailored to its superordinate perspective.
54
UNDERSTANDING MANUFACTURING NETWORKS
Imp. 2.1: Manufacturing strategy is supported by site and network capabilities. Hence, shaping the network capabilities determines the key task of the network management; it provides a lever for competitive manufacturing advantage. Imp. 2.2: Shaping the network capabilities is achieved by designing and modifying the network's configuration and coordination layer. Vice versa, the design of the configuration and coordination layers has to be made in accordance with the aspired manufacturing strategy. Imp. 2.3: In order to design and modify the network's configuration and coordination layers and, thus to shape the network capabilities to support the manufacturing strategy, the underlying decision dimensions of both layers need to be known and understood precisely. Imp. 2.4: Configuration and coordination cannot be addressed in isolation, but coordination needs to be considered contingent upon configuration. Hence, not only the decision dimensions but also their linkages have to be taken into consideration. Imp. 2.5: In order to support the network management in decision making, approaches are necessary operationalising and integrating the decision dimensions into a systematic and holistic management approach. Imp. 3.1: Network design and management approaches provide systematic decision support. They can be of both quantitative and strategic nature. Strategic approaches are rather focused on the derivation of visions and scenarios based on the creation of a common understanding and commitment; thus, they are more tailored to the network management's strategic perspective. Imp. 3.2: Strategic approaches are typically based on selected management frameworks and worksheets. Currently, literature provides only a very limited number of these approaches; evidence for their operationalisation and practical applicability are scarce. Imp. 3.3: The few identified strategic approaches clearly focus on the network configuration layer. Although of key managerial and academic interest, the coordination layer is either neglected or only addressed very basically. Imp. 3.4: Thus, the need for a holistic network architecture and a strategic approach to support the network management – especially in the management of the coordination layer – is evident. Such an approach cannot be designed in isolation but is constrained by the network configuration (and strategy).
UNDERSTANDING MANUFACTURING NETWORKS
55
2.5.2 Refinement of Research Question Recalling the guiding research question as introduced at the very beginning, the detailed discussion of the literature sharpens the understanding and allows for a clarification of its single fragments: Q. 1: How can the strategic coordination of intra-company manufacturing networks be supported systematically and methodically from a network level perspective? “How can the strategic coordination …”: This study addresses manufacturing networks from a strategic operations management perspective. It concentrates on the conceptual organisation and design of the network’s nodes and their nonphysical linkages more than on its daily operations or on the logistics’ and supply chain perspective’s concentration on the physical material flows. Within that context, the present work has a strong focus on the network’s coordination. This layer comprises the network managers' decisions regarding the design of the networks flows but also regarding the establishment of rules and mechanisms shaping the interaction between the plants and between the plants and the headquarters or central network management functions respectively. Since coordination is assumed to be contingent upon configuration, the links to the network’s structure and specialisation have to be considered, too. • “… of intra-company manufacturing networks …”: Intra-company manufacturing networks define the reference object of this research. An intracompany manufacturing network is understood as network with multiple scattered sites, owned by a single company, that conduct R&D, engineering, manufacturing, and / or assembly operations. These sites are connected by organisational ties, by information and knowledge flows, and / or by material flows. Moreover, further criteria might be necessary to isolate and define intracompany networks, especially for larger companies with multiple and mainly independent operations activities. • “… be supported systematically and methodically …”: Systematic and methodical support is targeted by (1) defining the decisions to be made in order to design and manage the network, (2) understanding their linkages, (3) underlying decisions by adequate tools and frameworks, and by (4) integrating these into a holistic network management architecture and design and approach. • “… from a network management perspective …”: The level of investigation is the network management. Providing this level with support requires a high degree of information condensation and guidance for strategic decision making. Any methodical support must be tailored to this level. It needs to be designed •
56
UNDERSTANDING MANUFACTURING NETWORKS
suitable to stimulate and foster strategic thinking and to derive strategic design options. Strategic approaches can fulfil these requirements. More detailed approaches, such as quantitative modelling or optimisation, can be applied complementary to evaluate these options but go beyond the scope of this study. 2.5.3 Derivation of Heuristic Research Framework A heuristic research framework is introduced and outlined in Fig. 17 to address the raised research question. It is built upon the implications from the literature review, comprising the central elements and linkages for the following work. Basically, the framework depicts an initial sketch of an integral network architecture centred on the coordination layer. It displays the network manager’s view from an operations management perspective. Elements in dashed lines represent factors and linkages that are not within the primary scope of this work but are supposed to impact the architecture. The following assumptions are reflected by the framework: • Differentiation between network and site level: The framework differentiates between a network and site level. From the former, the network is understood as an integral system with globally scattered and linked plants. On the site level, each plant is considered having individual competencies and strategic reasons, i.e., each plant plays a certain role. Since the site level influences decision making on the network level and vice versa, it cannot be neglected completely, even from the superordinate perspective of the network management. • Reflecting the coordination layer: The linkages between the nodes can either be physical, due to the exchange of goods and material, or – from an operations management perspective – plants are interconnected by non-physical linkages, such as by information or knowledge sharing. To coordinate the network, mechanisms need to be established in order to manage these linkages. These mechanisms address the flows in the network, but they also go beyond, covering institutional coordination aspects affecting the degree of freedom of the sites. Operationalising the mechanisms, i.e., understanding and designing the decision dimensions of the coordination layer, is central. • Considering contingencies: From a contingency perspective, the design of the network’s layers is subject to internal and external FIT. As argued in Section 1.2.1, FIT can be discussed on different levels. First, the variables of a single decision dimensions have to be aligned. Second, the decision dimensions within any distinct layer have to be aligned. Third, the decision dimensions of the different layers have to be aligned among each other, and fourth, the overall system has to be aligned with contextual factors. Therefore, especially the link between the coordination and configuration layer is assumed to be tight.
UNDERSTANDING MANUFACTURING NETWORKS
57
Configuration and coordination, in turn, define the two levers for the network manager to shape the network capabilities, i.e., to define the network’s contribution to facilitate manufacturing strategy. The manufacturing strategy itself is driven by the contextual environment and challenges the company is facing, such as business strategy, mega trends, customer and competitor structure, product and process restrictions, etc. The interdependences are symbolised in the framework by the cascaded structure of the elements embracing the network layers. On the site level, shaping the structural and infrastructural decision categories of a plant predetermines its competencies, and thus its role. A plant’s role must support the aspired site capabilities derived from the manufacturing strategy, but it also has to be aligned with the requirements coming from the network level.
Fig. 17: Heuristic research framework
• Striving for the network FIT: Striving for network FIT encompasses two aspects: its achievement and its sustainment. To achieve FIT, the architecture has to be open for network design and improvement. Thus, it serves as an anchor for a strategic network (re-)design and management approach. Sustaining FIT means to react on dynamic impacts and contextual changes to keep the optimal state.
58
DESIGNING THE NETWORK COORDINATION LAYER
3 Designing the Network Coordination Layer This chapter focuses on the elaboration of the network coordination layer as central element of the heuristic research framework. Section 3.1 is dedicated to the identification of the decision dimensions constituting network coordination. In Section 3.3, each decision dimension is operationalised by a so-called management framework. Since the frameworks are developed driven by both literature and iterative discussions with industry, empirical case evidence is given for their validation. Section 3.4 completes the chapter, discussing findings and deriving implications for the transformation of the heuristic research framework into a network management architecture.
3.1 Defining the Decision Dimensions of Network Coordination 3.1.1 Defining Coordination Coordination is often referred to from an organisation theory’s perspective. Following Thompson (1967), Smith et al. (1995) relate coordination to “… the combination of parts to achieve most effective of harmonious results” (Smith et al., 1995, p. 11). Malone and Crowston (1994) simply state that “… coordination is managing the dependencies between activities” (Malone and Crowston, 1994, p. 90) while Mills and Platts (2011) give a working definition as “…the actions through which dependencies between resources (human and physical) and activities are integrated toward improved competence performance” (Mills and Platts, 2011, p. 4). The overlap between these definitions – and the very rationale behind coordination – is the interdependency between entities working on a common goal. In line with this, Rogers and Whetten (1982) point out the following general characteristics of coordination: the existence for a joint goal, an amount of resources required at higher levels in organisation, agreements and formal rules, restriction of the involved parties’ autonomy, and long-term effects on the structure and operation of these entities. These characteristics allow for a first statement about the nature of coordination decisions: They shape the pillars of interactions between entities from a strategic point of view instead of steering and controlling the daily operative business. Thompson (1967) further separates interdependencies into (1) pooled dependencies, existing when entities are basically independent but rely on a common resource (e.g., two manufacturing sites relying on a central R&D unit), (2) sequential dependencies, describing a situation where a subsequent entity depends on the activities of a predecessor (e.g., in a manufacturing network organised as process chain), and (3)
DESIGNING THE NETWORK COORDINATION LAYER
59
reciprocal dependencies, where all entities are equally dependent on each other (e.g., joint process improvement programs between several manufacturing sites). 33 In order to align the interdependent activities, coordination mechanisms need to be set in place. Designing such mechanisms is a trade-off between restricting the behaviour of the entities involved and streamlining it with superordinate strategic goals, on the one hand, and leaving autonomy to benefit from the entities’ flexibility and heterogeneous capabilities, on the other (Nassimbeni, 1998). Refining March and Simon’s (1958) two basic coordination principles, i.e., programming behaviour and communication and feedback, Mascarenhas (1984) comes up with four underlying coordination mechanisms. The first basic principle, programming behaviour, intends to make the actions of the entities predictable. It can be achieved by (1) impersonal mechanisms, like standardisation of procedures, deadlines, and schedules, by (2) system sensitivity as the result of a process of socialisation of the entities’ members, e.g., by training, knowledge exchange, or transfer of personnel, and finally, by (3) compensation systems setting targets and incentives to streamline the entities activities. The second basic principle, communication and feedback, relies on (4) formal and informal verbal interaction. Further, with regards to organisational design literature, especially Mintzberg’s (1989) introduction of coordination mechanisms still receives recognition. He distinguishes six types (Mintzberg, 1989; Nassimbeni, 1998; Mills and Platts, 2011): • Coordination through mutual adjustment builds upon mostly informal communication and information sharing between the entities in the system, leading to a synchronisation and adjustment of activities. • Coordination through direct supervision builds upon a central responsible entity defining the direction and governing the activities for the whole system. • Coordination through standardisation of work processes builds upon standard operating procedures and instructions developed and rolled-out to specify and restrict the content of the activities of each single entity. • Coordination through standardisation of inputs / outputs builds upon the specification and control of not the activity itself but its input and output. • Coordination through standardisation of skills and knowledge builds upon knowledge and resources sharing to achieve an orchestrated and align level of capabilities between the involved entities. • Coordination through standardisation of norms builds upon common beliefs and a joint mission and vision.
33
Malone et al. (1999) give a similar classification differentiating between sharing, flow, and fit dependencies.
60
DESIGNING THE NETWORK COORDINATION LAYER
Similarly, Bartlett and Ghoshal (1989) emphasise centralisation, formalisation, and socialisation as main coordination mechanisms. Martinez and Jarillo (1989) go into more detail when reviewing studies on the coordination of multinational corporations between 1953 to 1988. Their analysis is based on a working definition of coordination mechanisms as “… any administrative tool for achieving integration among different units within an organisation” (Martinez and Jarillo, 1989, p. 490); this definition is further concretise by eight common mechanisms depicted in Tab. 8. For these, the authors explicitly highlight that they are not exclusively related to MNC but can account for coordinative challenges in basically all large organisations. Structural and formal mechanisms 1 Departmentalisation or grouping of organisational units, labour division, shaping the formal structure 2 Centralisation or decentralisation of decision making through the hierarchy of formal authority 3 Formalisation and standardisation: written policies, rules, job descriptions, and standard procedures through instruments such as manuals, charts, etc. 4 Planning and harmonisation of systems and processes: strategic planning, budgeting, functional plans, scheduling, etc. 5 Output / performance control and personal control / behaviour control: financial performance, technical reports, sales and marketing data, etc., and direct supervision
Other mechanisms, more informal and subtle 6 Lateral or cross-departmental relations: direct managerial contact, temporary or permanent teams, task forces, committees, integrators, and integrative departments 7 Informal communication: personal contacts among managers, management trips, meetings, conferences, transfer of managers, etc. 8 Socialisation: building an organisational culture of known and shared strategic objectives and values by training, transfer of managers, career path management, measurement and reward systems, etc.
Tab. 8: Coordination mechanisms in multinational corporations (adapted from Martinez and Jarillo (1989))
3.1.2 Identifying the Network Coordination Decision Dimensions Shifting the scope to the operations management literature on manufacturing networks, the lack of a concise definition of coordination constitutes the need for a deeper discussion. Among the scarce contributions, Meijboom and Vos (1997) simply state that: “… co-ordination refers to the question of how to link or integrate the production and distribution facilities in order to achieve the firm’s strategic objectives” (Meijboom and Vos, 1997, p. 790). More concrete, De Toni and Parussini (2010) differentiate between the management of the internal flows (information, know-how, and best practice) and institutional and organisational aspects (like control and standardisation of processes and methods) when pointing out that:
DESIGNING THE NETWORK COORDINATION LAYER
61
“... coordination decisions concern the management of the network, in particular the management of physical and information flow integration between facilities, control practices, competence groups for integration and control, management of facilities’ roles and know-how exchange, diffusion of best practices, standard processes and methodologies” (De Toni and Parussini, 2010, p. 7). Tab. 9 provides an overview of further statements on network coordination by renowned authors. It differentiates between two perspectives: the (1) management of the network flows and the (2) institutional aspects, adding the underlying dimensions as pointed out by the respective scholars. Regarding the management of the network flows, the items addressed range from internal learning, knowledge, innovation, and technology transfer, to communication, information, people exchange, and the sharing of physical goods. Since this enumeration is partially overlapping in the single dimensions, other scholars come up with a clearer distinction between the (1) flow of material / physical goods, (2) information, (3) people, and (4) financial resources (Bartlett and Ghoshal, 1989; Vereecke et al., 2006). In any case, the chosen operations management perspective calls for some remarks. The design of the material / physical goods flows, for instance, is predetermined by the network structure, and thus by configuration rather than coordination decisions; even its daily management is mainly part of the logistics and supply chain function and not in the closer scope of the network manager’s operations responsibilities. Similarly, the design and management of the financial flows are not considered as core task from an operations perspective. Furthermore, the term “flow of information” is not concise. It covers pure information and data sharing but also the exchange of best practice and knowledge. In this context, Gupta and Govindarajan’s (1991) suggestion to separate the flow of information into a flow of administrative information as well as a flow of knowledge and innovation still remains reasonable. Moreover, the flow of people can also be understood as transfer mechanism and carrier of knowledge and / or information, thus as part of either of them. To conclude, from an operations management lens on network coordination both the design and management of the flow of (administrative) information and the flow of knowledge (and innovation) will further be considered as central decision dimensions, but their content requires a more detailed discussion.
62
DESIGNING THE NETWORK COORDINATION LAYER
Author
Definition / Description
Chew et al. (1990)
"… we have tentatively identified two core decisions that network management faces in designing an effective network; they concern the degree of cooperation versus competition among plant managers and the centralisation versus decentralisation of information flows" (Chew et al., 1990, p. 160) "In the context of manufacturing networks, coordination primarily involves tactical level decision-making aimed at planning global activities efficiently and effectively. A fundamental issue to be addressed at this level is how to design and manage the flow of goods, people, technology, and information in international networks" (Cheng et al., 2008, p. 2) "Coordination is related to the management of a network and refers to the question of how to link or integrate the facilities in order to achieve the firm’s strategic objectives. Its aim is to achieve an efficient and effective plan for global production activities, which involves primarily tactical decisions in different business areas and within several processes. In addition, coordination is also concerned with technology transfer and diffusion, as well as within-network learning" (Cheng et al., 2011, p. 1314)
Cheng et al. (2008)
Cheng et al. (2011)
De Toni & Parussini (2010)
Dimensions of network coordination Flows Institutional perspective - Flow of information - Flow of knowledge
- Flow of goods - Flow of people - Flow of technology - Flow of information
- Technology transfer - Internal learning
"... coordination decisions concern the management of the network, in - Physical flow particular the management of physical and information flow - Information flow integration between facilities, control practices, competence groups for - Know-how exchange integration and control, management of facilities’ roles and know-how exchange, diffusion of best practices, standard processes and methodologies" (De Toni & Parussini, 2010, p. 7)
Feldmann & Olhager (2009) and (2011) Jaehne et al. (2009)
Maritan et al. (2004)
Meijboom & Vos (1997) Rudberg & West (2008)
Tsai (2001)
Vereecke & De Meyer (2009)
- Incentive system - Centralisation or decentralisation of tasks related to (1) creation, (2) identification, (3) transfer, and (4) application of knowledge
- Control practices - Management of facilities' roles - Standardisation (processes, methods)
- Allocation of decision autonomy for strategic manufacturing decisions
"There are many conceptual articles that address the challenges of coordinating a multiplant network in an international context (…) Some of these frameworks are derived for the entire network in terms of material flows, or for a plant type which characterizes the network. Another approach describes the role of the plants relative to a network and allows for different plant roles within the same network (...) While Ferdows focuses on what the appropriate role for a plant would be in different circumstances (...) we focus on how to manage a plant given a particular role" (Maritan et al., 2004, pp.489-490) "… co-ordination refers to the question of how to link or integrate the production and distribution facilities in order to achieve the firm’s strategic objectives" (Meijboom and Vos, 1997, p. 790) "To facilitate the smooth coordination, companies typically develop common policies regarding manufacturing structure and infrastructure, not only in terms of manufacturing capacity and technology, but also with respect to new product introduction (NPI), production ramp-up, planning and control systems, and organizational issues" (Rudberg and West, 2008, p. 92) "Organizational units are embedded in a network coordinated through processes of knowledge transfer and resource sharing" (Tsai, 2001, p. 996 according to Galbraith, 1977) "There is a need for a coordinated evolution of the network, i.e. of both its nodes and its flows. For the senior manager sitting in headquarters and orchestrating the manufacturing network, the main message coming from our research is that the design of the manufacturing network is more than a decision of what to produce where and how to organize the logistic flows. It is also about the design and management of the flows of innovation and know-how" (Vereecke & De Meyer, 2009, p. 21)
- Coordination demand reducing strategies: outsourcing, resource provision, resource flexibility - Coordination demand fulfilling strategies: structural, technocratic, and personneloriented coordination - Decision autonomy of plants
- Standardisation of production processes
- Policies regarding manufacturing structure and infrastructure - Standardisation of processes, systems, and principles
- Knowledge transfer
- Resource allocation and sharing
- Logistics Flows - Flow of innovation - Flow of know-how
Tab. 9: Definitions & decision dimensions of manufacturing network coordination
From an institutional perspective, several authors point out the importance of shaping the degree of parental control in the network (e.g., Chew et al., 1990; Maritan et al., 2004; Feldmann and Olhager, 2009b; Feldmann, 2011). The spectrum ranges from pure hierarchical control by centralising a maximum level of authority at the
DESIGNING THE NETWORK COORDINATION LAYER
63
headquarters to decentralisation and full autonomy of each site. Thereby, autonomy is typically defined by the assignment of responsibility for making key decisions to a certain organisational level, thus by constituting the degree of centralisation in the network (Maritan et al., 2004; Feldmann and Olhager, 2009b). Other scholars refer to the degree of standardisation and control of manufacturing-related processes as coordination mechanism similarly influencing autonomy (Meijboom and Vos, 1997; Rudberg and West, 2008; De Toni and Parussini, 2010). Yet, to put it more generally, the question of coordination is not only where to assign responsibility, but – more essentially – whether to facilitate cooperation or internal competition and rivalry between sites, or to establish hierarchical guidance. Certain authors introduce a market perspective to networks, which contrasts hierarchical control. They argue that entities participate in several internal markets - for goods and services, charters, competencies, and practices - and discuss the implementation of market mechanisms instead of central authority for coordination (Birkinshaw, 2001; Cerrato, 2006). Additionally, Luo (2005) captures the idea of “coopetition”, a concept originally formulated by Brandenburger and Nalebuff (1996). He states that units in the network simultaneously compete and cooperate in selected areas, and that managers have to nurture and unleash both. As areas of cooperation, he identifies (1) technology-related know-how, such as process and product knowledge and innovation, (2) operational resources, like distribution channels, quality control programs, policies, and standards, (3) organisational capabilities, such as managerial experience, and (4) financial knowledge, i.e., transfer pricing or experience in managing cash-flows. Areas of competition are (1) parental resources, like technology, equipment, key talents, corporate support, and training, (2) a plant's system position, i.e., its position in the value chain, in the knowledge flow, its competence position, or its position to influence the headquarter in key decisions, and finally (3) market share and expansion (Luo, 2005). Accordingly, institutional levers to foster a competitive or cooperative attitude go beyond centralisation and standardisation, comprising the allocation and sharing of scarce resources, the incentive system, and the transparency of information (Chew et al., 1990; Luo, 2005). This is also in line with Jaehne et al. (2009) pointing at resource provision and flexibility as “coordination demand reducing strategies”. To conclude, the following dimensions can be isolated in order to be elaborated on as decision dimensions for the network coordination layer: • Centralisation and standardisation, as determinants influencing autonomy. • The allocation and sharing of resources, driving competition and cooperation between the sites.
64
DESIGNING THE NETWORK COORDINATION LAYER
• The design of the incentive system, again as driving forces in shaping the continuum between competition and cooperation in the network. • The design and management of the information and knowledge flows. Referring to the collection in Tab. 9, these four dimensions embrace all items identified in the literature except for the “management of facilities’ roles” as mentioned by De Toni and Parussini (2010). So far, managing a networks’ plant role portfolio has solely been considered as configuration issue on network specialisation. Yet, there is obviously a strong link to coordination that calls for further investigation. Concerning the generic coordination mechanisms introduced at the beginning of this chapter, the four dimensions are expected to match. Hence, linking the network with the organisation theory’s perspective provides validity of the decision dimensions regarding content and completeness; this will be postponed to the end of this chapter.
3.2 From Decision Dimensions to Coordination Frameworks 3.2.1 Methodical Approach The decision dimensions represent the levers to shape the coordination layer. However, with respect to the current research goal, operational feasibility of these dimensions is required. Thus, this section is dedicated to establishing first a deeper understanding of the dimensions and their underlying variables in the context of a manufacturing network; then, this understanding is utilised in providing distinct management frameworks as tools for decision support. Consequently, designing the frameworks requires both grounding in academia and demonstrated applicability in practice. For this, the methodical proceeding was based on the “process research approach” suggested by Shi and Gregory (1998): a literature-driven derivation of frameworks and their testing and refinement in practical settings. 34 Five manufacturing networks served as case studies providing the “playground” for framework design and testing. An iterative approach was employed, comprising more than 40 workshops with the different networks’ management teams to assure feasibility and generalisability of findings. Further, results were enriched by a cross-industry survey on European manufacturing companies. The purpose of the survey was two-fold: to put the frameworks into a wider context and to identify obstacles, practical approaches, methods, and “success-stories” for the network coordination in industry. Based on the survey, additional interviews were conducted with representatives of the global operations management function of three participants. 34
Similarly, Eisenhardt (1989) suggests to use a priori constructs at the early phase of a theory building process and to refine them iteratively.
DESIGNING THE NETWORK COORDINATION LAYER
65
Experiences from the case studies, findings from the survey, and impressions from the interviews will be synthesised in the next section, targeting the following goals: to (1) define, explain, and test the frameworks, to (2) understand the rationale behind the coordination mechanisms in different companies as reflected by the frameworks, and to derive implications for (3) their handling as management tools as well as for (4) the superordinate design of the network architecture. In order to cope with the ample amount of empirical data from different sources, selected “anecdotic evidence” will be given instead of a detailed presentation of all data. This allows for a condensed and target-oriented integration of the findings. Moreover, the promotion of the frameworks themselves sticks to a predefined structure: • The framework development and description, which is literature-driven and case study based. • The logic behind the framework, which is again underpinned by the case studies but also by findings from the survey and the interviews. • The discussion of findings and implications for the coordination layer in specific, for the design of the network architecture, as well as for network management in general, based on a tabular presentation. Before going into detail, a short outline of the empirical data base, i.e., the case networks, the survey, and the interview partners, is put in front. 3.2.2 Case Study Selection & Outline While the choice of case studies as research strategy for the development of the management frameworks has been justified in Section 1.3.2, the characteristics of the single cases, i.e., the individual networks as unit of analysis, and the rationale behind their selection are outlined shortly. The case networks were not selected at random but are based on theoretical considerations (theoretical sampling) (Flick, 1999). The selection was conducted to assure a high degree of generalisability of results. This can be achieved by an appropriate number of cases (e.g., Meredith, 1998; Voss et al., 2002) and by choosing polar types covering a suitable diversity (Eisenhardt, 1989). Regarding the number, Eisenhardt (1989) proposes four to ten cases as adequate for theory building while Meredith (1998) calls for two to eight; both requirements are met by the actual number of five case network in the current work. Diversity is achieved by selecting the cases based on the following criteria: • Regionally, multinational, or worldwide dispersed manufacturing network with at least five interdependent manufacturing sites. • Differences in network size, i.e., the number of employees and manufacturing sites.
66
DESIGNING THE NETWORK COORDINATION LAYER
• Differences in the configuration of the network structure according to the ideal types distinguished Section 2.3.3. • Differences in the industry type and core products manufactured. • Differences in the manufacturing processes and technology. Tab. 10 outlines the selected case networks and their characteristics. Network scope
Company Division Business Other level level unit level Excitation NW Drives NW Seals NW Edgeband NW Profile NW
Core products Electrical Excitation equipment systems Electrical Voltage equipment drives
HQ Industry CH CH
GER Mechanical products n/a Polymer products n/a Polymer products
Core processes Engineering & assembly Engineering & assembly
Network characteristics
# Sites # Employ. Global Network (operat.) (operat.) dispersion structure* 12 ca. 250 Worldw. Hub & spoke 5
ca. 400
Worldw.
Mechanical seals Edgeband
Chipping technology Extrusion
18
ca. 1700
Worldw.
12
ca. 1000
Worldw.
Frames & profiles
Extrusion
8
ca. 1500
Worldw.
Transition from world products to local for local World products / Local for local Local for local / Web structure Local for local
* According to Meyer and Jacob (2008)
Tab. 10: Outline of case study networks used in designing the coordination frameworks
The five networks manufacture different product types belonging to three industries. • The Excitation NW is a widely independent business unit of a large electrical company, and a provider of electrical devices to support and control the performance of generators in power plants. Core value adding processes are engineering and assembly. • The Drives NW similarly represents the basically autonomous business unit level of an electrical equipment manufacturer that promotes frequency drive systems for electronic motors in large industry applications, and it is also based on engineering and assembly processes. • The Seals NW is the biggest division of a mechanical company, and a provider of sealing technology for various small to large mechanical devices. It primary relies on chipping processes. • The Profile NW is a global business unit of a polymer processing manufacturer, and a provider of low cost to prime polymer frames for the building industry. Manufacturing is based on extrusion technology. • The Edgeband NW also represents a business unit in the polymer industry, focussing on the extrusion of a large variety of finishes for furniture companies. The size of the networks ranges from five to 18 manufacturing sites and from 250 to about 1700 employees in operations. In each case, worldwide network dispersion is assured with sites located in more than three world regions. The network structure covers a hub and spoke and local for local production as well as combinations with dedicated world plants or a web structure.
DESIGNING THE NETWORK COORDINATION LAYER
67
3.2.3 Survey & Interview Outline The highly iterative and explorative interaction with the case networks for developing and testing of the coordination frameworks is complemented by a quantitative survey and selected interviews. The cross-industry survey “Excellence in Global Operations (X-GO)” conducted by the University of St.Gallen 35 targeted manufacturing companies in Europe to identify management practices for the configuration and coordination of global manufacturing networks. During a period of six months (December 2010 – May 2011), approximately 550 companies with manufacturing networks were contacted via e-mail and phone, more than 250 questionnaires were placed, and 56 networks 36 of eleven industries participated. Industry type
16%
n=56
2% 4% 4% 3%
Sales volume
9%
9%
20%
12% 11%
5%
5%
9%
Machinery and equipment n.e.c.
Basic metals and metal products
3%
30%
36% n=56
Small (<100 Mio. €)
Medium (100 to 1000 Mio. €)
16%
Motor vehicles, trailers and transport equipment Repair and installation of machinery and equipment Other k.a.
Small (<200)
Medium (200 -1000 Very large (> 5000) n.a. / k.a.
43%
n=56
Large (> 1 Mrd. €)
n.a. / k.a.
Rubber, plastic products, and other non metallic mineral products Electric equipment Computer, electronic and optical products
21%
Large (1000 - 5000)
Number of sites in the network
Pharmaceutical products
4%2%
52%
Food products, beverages and tobacco Chemicals and chemical products
35
Number of employees in the network
11%
5%
n=56
5%
13%
3%
< 5 sites
5 to 10 sites 50%
Product types served*
11 to 15 sites 16 to 25 sites > 25 sites n.a. / k.a.
32%
n=59
65%
Industrial goods
Consumer goods k.a.
* Multiple answers possible
Fig. 18: General information about the X-GO survey sample
The survey was conducted by the Transfer Centre for Technology Management (TECTEM) of the University of St.Gallen. An excerpt of those questions used for this study is given in Appendix B.1. For the full set of questions and a different perspective on the X-GO survey results, see Thomas (2013). 36 52 networks participated directly in the survey; four networks were added later based on the same questionnaire.
68
DESIGNING THE NETWORK COORDINATION LAYER
Network structure*
Network organisation
World factory Local for local
Others 7%
Region 1 Region 2
14%
Hub & spoke
n.a.
Function
4% Function
Web structure
10% Other/n.a.
* According to Meyer and Jacob (2008)
Plant 2 …
39%
Matrix
24% 16%
Plant 1
23%
23%
Sequential or convergent
n=56
Product A
13%
Divisional mix Product Division
16%
Region Division
Plant 1
Plant 3
Plant 2
Plant 4
…
…
11%
n=56
Division Division
Plant 1 Plant 2 …
Fig. 19: Network characteristics of the X-GO survey sample
Fig. 18 and Fig. 19 give some general information on the survey sample and the characteristics of the participating networks. Most companies are represented with a medium-sized network embracing five to ten manufacturing sites and 1000 to 5000 employees. The products served are primary industry goods, and the dominating industries 37 are machinery and equipment, basic metals and metal products, and also pharmaceutical products. Regarding the network configuration and the organisation of the manufacturing function, all structures outlined in Section 2.3.3 are encompassed by the sample; so are most of the well-known generic organisational structures covered. As condensation of the survey results, two measures were calculated and contrasted for each participant: the (1) “overall network capability level” and the (2) conformance with the aspired capability targets. Regarding the former, the 15 items of the network capabilities introduced in Tab. 4 were assessed by distinct KPIs. 38 Based on the KPIs, for every participating network, the capability level in each item was calculated as deviation from the survey sample’s mean, aggregated, and normalised by the survey sample’s overall average level. Regarding the “network capability conformance”, the importance of the 15 items was assessed by the participants and contrasted with their absolute capability level. Distances between both were transformed to a 1 to 5 scale, and the average conformance was again calculated relatively to the survey mean. As a result, the networks could be assigned to four categories as depicted in Fig. 20.
37
The categorisation of the industry types is based on the ISCI codification, covering the classes 15 to 35. Regarding the network capabilities and the corresponding KPIs to calculate the capability and conformance level, see Appendix B.2.
38
DESIGNING THE NETWORK COORDINATION LAYER Overall network capability level & conformance
69
Overall network capability level as deviation from mean
High capability level / High conformance
II
I
IV
III
Network capability conformance as deviation from mean
Low capability level / High conformance
Low capability level / Low conformance
High capability level / Low conformance
n = 48
Overall performance level on strategic manufacturing priorities
Below average
Above average
Overall performance level on strategic manufacturing priorities as deviation from mean
n = 41
Fig. 20: Network capability level & conformance vs. performance level on strategic manufacturing priorities
• Category I “High capability level / High conformance”: The network has a higher overall capability level than the survey average and achieves a higher degree of conformance. Not only the most important network capabilities are addressed, but the capability level is above average across almost the full set. • Category II “Low capability level / High conformance”: The network has a lower overall capability level than the survey average but achieves a higher average conformance. This might be due to an efficient concentration on those capabilities considered as highly important by the network management. • Category III “High capability level / Low conformance”: The network has a higher overall capability level than the survey average but achieves a lower average conformance. This shows a mismatch between the capabilities addresses and those actually considered as important to be competitive. • Category IV “Low capability level / Low conformance”: The network has a lower overall capability level than the survey average and a lower average degree of conformance. Similarly to the evaluation in Fig. 1 of the introductory section, the bottom of Fig. 20
70
DESIGNING THE NETWORK COORDINATION LAYER
correlates the four categories with the overall performance of each participant along its strategic manufacturing priorities 39. As indicated, networks being positioned in category I achieve a much higher average “overall performance level on the strategic manufacturing priorities” than those being positioned in category IV. However, since roughly 50% of the participants are positioned in category III or IV, managers obviously struggle with shaping and realising their aspired network capabilities. Based on the four categories, interview partners were selected to verify findings from the survey and to get deeper insights into their challenges, approaches, and management practices. One representative of each of the categories I to III was chosen for a detailed on-site discussion. Category IV was left aside since the additional value of an interview was considered marginally. Semi-structured interviews were conducted with the person responsible for the network, typically the global operation manager, in some cases supplemented by his management team. The represented networks are introduced shortly; their characteristics are outlined in Tab. 11: • The Floor Care NW represents a product segment of a German manufacturer of domestic appliances, and is a premium provider of vacuum cleaners and devices. With a high real net output ratio, production activities are split into component manufacturing and final assembly. • The Dental NW is an independent business unit of a European engineering group, focussing on precious metal and special material. The business unit dental provides medical products and materials for dental prosthetics and restoration based on chemical process engineering. • The Pet Food NW represents a leading private label pet food producer in Europe. It provides the full range of complete wet and dry diets for dogs and cats, comprising food processing and packaging processes. Network scope
Company Division Business Other level level unit level Floor Care NW Dental NW
Pet Food NW
Core products GER Electrical Vacuum equipment cleaners & devices GER Medical & Dental technical prosthetics & products restoration material HUN Food Pet food products HQ Industry
Core processes Component manufacturing & assembly Chemical process engineering
Network characteristics
# Sites # Employ. Global (operat.) (operat.) dispersion 6 > 1000 Multinat. (2 regions)
Food processing & packaging
Network structure* Sequential or convergent
Network category** I
8
ca. 400
Multinat. World products (3 regions)
II
7
ca. 1000
Regional (Europe)
Web structure / Local for local
III
* According to Meyer and Jacob (2008) ** According to the survey classification of the network capability level and conformance
Tab. 11: Outline of interviewed networks & corresponding network categories 39
The calculation of the overall performance level on strategic manufacturing priorities is similar to the calculation of the overall network capability level but based on a set of competitive priorities instead of network capabilities. For details of the competitive priorities and the related KPIs, see Appendix B.3.
DESIGNING THE NETWORK COORDINATION LAYER
71
3.3 Developing the Coordination Frameworks 3.3.1 The Centralisation & Standardisation Framework Framework Development & Description Granting responsibility for making key decisions to a distinct organisational level is seen as main lever to define the degree of autonomy in the network (Maritan et al., 2004; Feldmann and Olhager, 2009b). Allocating decision authority to a central unit constitutes centralisation and restricts the sites’ degree of freedom whereas allocating authority to the single sites leads to decentralisation. The degree of standardisation is assumed to have another major impact on autonomy. Meijboom and Vos (1997), for instance, evaluate the degree of standardisation of core processes to conclude about a site’s autonomy. Intuitively, standardisation of processes gives headquarters the opportunity to retain parental control, even if their execution is decentralised. Often, previous work has been too narrow-minded, as both dimensions, centralisation and standardisation, have been investigated independently. Here, a framework is proposed, as depicted in Fig. 21, which supports operations managers in shaping autonomy in their manufacturing network based on a combination of both: standardisation (x-axis) and centralisation (y-axis). Responsibility areas System
Each site individually
Centralisation / Responsibility 1
S
System
4 11 17
3
19 22 23 10
D
20
9
Decision P Process
8 9 10 11 12 13 14 15 16 17
Standardised
2
Centralised & Standardised
Centralised 14
8
15
16 13
Production system Product data mgt. system Quality & maintenance system Management system Improvement programs (besides production) HR system Know-how exchange system …
Decision
Region
Several sites
6 7
24
Autonomous
Central unit
1 2 3 4 5
18
6
Site strategy & roles Organisational structure Manufacturing IT decisions Make-or-buy decisions Product allocation decisions Transfer pricing Production process decisions Manufact. technology decisions Long-term capa. development Short-term capacity adjustment …
Process
5
7
12 21
No / Local standardisation
Documented rules, guidelines & processes
Audited / Controlled processes & routines
Standardised (IT-) tools or methods
Individual tools / heterogeneous implementation level at each site
Individual tools / homogeneous implementation level at each site
Standardised tools / heterogeneous implementation level in the network
Standardised tools / homogeneous implementation level in the network
Fig. 21: Centralisation & standardisation framework
18 19 20 21 22 23 24
Strategic sourcing Strategic logistics Product cost calculation Long-term S&OP Intern. SC-planning / Order alloc. Short-term manuf. planning Manufacturing / Operations
P Degree of D standardisation S for the network
72
DESIGNING THE NETWORK COORDINATION LAYER
The framework allows to map the network’s AS-IS situation and to formulate targets for its TO-BE state in terms of authority allocation. To do so, it is based on the idea of evaluating and positioning so-called “responsibility areas” (as listed at the right side in Fig. 21) according to their degree of centralisation and standardisation. As argued above, centralisation depends on the assignment of responsibility to either a central unit (typically including the network management function) or to each site individually. Depending on the organisational structure of the network, other levels are conceivable to carry responsibility, too, such as a region or a subset of sites. Assessing standardisation, on the other hand, might vary according to the type of responsibility area. Most scholars restrict these areas to what they call decision categories, solely asking about the authority for decision making (e.g., Maritan et al., 2004; Christodoulou et al., 2007; Feldmann and Olhager, 2009b; Feldmann and Olhager, 2011). Feldmann and Olhager (2009b), for example, study the decision responsibility of plants based on their role; they apply 15 decision categories derived from operations management literature. Feldmann and Olhager (2011) use 14 categories, concluding that decisions in a network are made either centralised, decentralised, or integrated between local plants and central headquarters. Christodoulou et al. (2007) provide a simple matrix assigning authority for 13 decision categories between business unit, product responsibility, region, and plant level. But restricting responsibility to the authority for decision making only, is too narrowly considered. Instead, the introduced framework differentiates between (1) systems, (2) decisions, and (3) processes as three main responsibility areas. Isolated from literature and field-discussion, these areas are broken into 24 sub-categories as outlined in Tab. 12. Tackling the most important manufacturing issues, the collection of different categories is a starting point for the analysis and design of centralisation and standardisation as formal coordination mechanisms in the network Nonetheless, their completeness and terminology might vary according to a company’s specific context; this makes any adaptation and complementation permissible.
DESIGNING THE NETWORK COORDINATION LAYER
73 Authors
Responsibility areas & categories
Maritan et al. (2004)
Systems
… for primary activities
1 Production system
2 Product data mgt. system 3 Quality & maintenance system
… for support activities
4 Management system
5 Improvement programs 6 Human resource system (HR)
Hayes et al. (2005)
Vereecke et al. (2006)
Product & process development system
Quality standards Maintenance policies & practices Choice of management information system Choice of accounting system
HR policies for management HR policies for labour
Continuous improvement Methods & tooling
Quality system
Choice of standards, Maintenance goals & measures for QM operations
Measurement system Budgeting system Product & process development system HR system
Introducing a new planning & control sys.
products
Processes
strategic
proc., techn. & capa.
Decisions
organisation
7 Know-how exchange system 8 Site strategy & roles
Feldmann & Olhager (2009)* and (2011) Selection of improvement programs
Selection of quality tools
Continuous improvement
Selection of improvement programs Employee competence development
Footprint strategy
Plant focus Plant specialisation
9 Organisational structure 10 Manufacturing IT decisions 11 Make-or-buy decisions
Organisation Choice of production planning & control Raw material sourcing Component sourcing
Choice of organisational design
Process & information technology Sourcing
Make-or-buy decisions
12 Product allocation decisions 13 Transfer pricing
Product & process development system
Developing or changing a product
Product innovation
14 Production process decisions 15 Manufacturing technology decisions 16 Long-term capacity development 17 Short-term capacity adjustment 18 Strategic sourcing
Product & process development system Product & process development system Capacity
Developing or changing a production process Choice of technology
Process definition
Process choice
Transfer of technology
Manufacturing technology Timing of capacity acquisition Capacity levels relative demand Supplier selection
Equipment sourcing
Capacity Sourcing
Work allocation Selection of a new supplier
Strategic procurement Tactical procurement
19 Strategic logistics 20 Product cost calculation 21 Long-term S&OP 22 Internal SC-planning / Order allocation
operational
Christodoulou et al. (2007)
23 Short-term manuf. planning 24 Manufacturing / Operations
Not considered
*Delivery programs with customers
Long-range production Budgeting planning system Work planning & control system
Developing sales forecasts Developing a master production schedule Developing material & capacity plans Managing inventories
Production scheduling
Developing shop floor schedule
Production scheduling Production operations
Placing purchasing orders
Strategy implementation
Work planning & control system
Work allocation
Tab. 12: Responsibility areas & categories
Long-term planning & control principles Capacity levels relative demand
Short-term planning & control principles
74
DESIGNING THE NETWORK COORDINATION LAYER
Systems, as the first responsibility area, are either related to primary operations activities as proposed by Porter and Millar’s (1985) value chain, like the production system, the product data management system, or the quality and maintenance system, or they focus on support activities, such as the management and KPI system, the HR system for training and qualification, improvement programs besides manufacturing 40, or the know-how exchange system. A system’s degree of centralisation in the framework is evaluated by the allocation of responsibility for developing, maintaining, and improving the system while standardisation is about the degree of formalisation and implementation across the network; it ranges from individual tools and methods which are heterogeneously implemented at each site to standardised tool and methods homogeneously implemented throughout the network. Ex. 1.1: The discussion about the production system gives an example: While the bigger part of the survey sample (43%) pursues a centralised development and standardised implementation of its production system, a small number of networks follows a different strategy. One of these is the Floor Care NW as positioned in the framework (1). The development of an individual production system is seen as an important milestone assuring sustainable productivity increase by providing tools and methods for manufacturing, assembly, and logistics optimisation. It was designed and is improved
Centralisation / Responsibility
Autonomous
1
Standardised
14%
1 Production system
23%
continuously,
not
top-down
and
pushed centrally, but by a bottom-up approach, picking up best practice solutions of each site and setting-up
Centralised
competence
Centralised & Standardised
13%
Degree of standardisation for the network xx% = Percentage of study sample n.a. = 7%
groups
for
distinct
topics. Identified or newly designed
43% P D S
solutions are presented on a regular basis and rolled-out successively at the other sites. Thereby, standards are not copied but adapted to local
requirements. The bottom-up development of the production system creates a common platform for best practice exchange, also fostering individual sites’ readiness to actively participate in the network.
Decisions, as the second responsibility area, are focused on central strategic manufacturing issues of (1) organisation, (2) products, as well as (3) processes, technology, and capacity. Organisational decisions address plant strategies and roles, the organisational structure of the sites, which can be either centrally prescribed and copied or free of the local manager’s choice, and manufacturing-related IT decisions, 40
Like, for example, process improvement programs in administrative areas.
DESIGNING THE NETWORK COORDINATION LAYER
75
such as the selection of software, e.g., for product design or production planning. Product-related decisions cover the make-or-buy strategy, and thus determine a site’s degree of value added, the assignment of product development and manufacturing responsibility, and also the decision on transfer price setting. Finally, process, technology, and capacity decisions determine the authority for processes and technology choice and allocation, as well as the capacity development. Centralisation in this context is understood as the organisational level responsible for decision making while standardisation is operationalised by establishing rules and guidelines, routines, or even (IT-based) methods, all guiding the decision process and making it transparent and replicable. Ex. 1.2: The lack of standardisation for manufacturing-related IT decision making (10) causes severe problems in the global Seals NW – a story which could be similarly told for most of the interviewed
Centralisation / Responsibility
Autonomous
10
Centralised
2
!
Standardised
2 Product data mgt. system 10 Manuf. IT decisions
Centralised & Standardised
? !
Degree of standardisation for the network
P D S
= Cause = Problem = Target = Solution
networks. Although under central responsibility, in the past, different product
development
and
engineering solutions (such as CAD programs) were installed at the distinct sites due to the lack of network-wide
standards
for
replicable decision making. Today, the integration of the insular
software formats is one of the biggest obstacles to moving the product data management (2) to a standardised and homogeneously implemented system in the network.
Processes, as the third responsibility area, comprise both strategic as well as operative processes. The former involve strategic sourcing and logistics, addressing the selection and qualification of suppliers or logistic partners respectively, the calculation of product costs, and the long-term sales and operations planning (S&OP); the latter includes supply chain planning to balance orders in the network, short-term production planning and scheduling, and pure manufacturing / operations processes. Similar to decision making, centralisation of processes is evaluated by defining the organisational level holding the authority for process execution while standardisation begins with the documentation of processes, their auditioning and controlling, and ends with putting rigid IT systems in place to limit any process variations. Combining centralisation and standardisation, four generic network positions can be distinguished:
76
DESIGNING THE NETWORK COORDINATION LAYER
• The “centralised network” assigns main responsibility to central levels whereas the degree of standardisation is limited. This position is typical for younger or emerging networks, or those in a transfer state from decentralisation to centralisation. In this case, responsibility is often shifted formally to central units, but the establishment of network-wide standardisation is lagging behind, requiring time, financial resources, and a cultural mind-change. Further, the position is often found for decisions or processes which are considered of less importance or which occur only sporadic and irregularly. Ex. 1.3: The development of the so-called health, safety, and environment system (HSE) at the Seals NW tells the story of how a growing management perception accelerates the process of x HSE system
x Centralisation / Responsibility
Autonomous
Standardised
standardisation. Although started as an autonomous system with each site deciding what to do and how to meet
1
local safety requirements, with a
Centralised
x
2
growing number of smaller incidents
Centralised & Standardised
the system slowly evolved towards x
Degree of standardisation for the network
P D S
central control. A bad accident with a fork lifter, which caused the dead of a blue collar worker, changed the
management attention and made today’s HSE system the most standardised and strictly controlled system in the whole network within shortest time.
• The “centralised & standardised network” often emerges from the centralised network when standardisation for the responsibility areas is consequently carried forward. For processes and decisions, this means to establish formal guidelines and control structures while systems are rolled-out across the individual sites. Nonetheless, responsibility remains at the central unit. The position is aimed at for responsibility areas of highest priority, or it is often formulated as a fundamental philosophy for established networks with strong central and hierarchical power. Ex. 1.4: The Profile NW, as depicted in the framework below, provides a typical example of a strongly centralised and standardised network. All activities related to system development and improvements are dedicated to either the central business division or the superordinate corporate unit and rolled-out successively top-down across the network; so are strategic decisions made solely on these central levels. Sites are considered as extended work benches with authority restricted to scheduling and production tasks only (20, 23, 24), and even the execution of this tasks follows
DESIGNING THE NETWORK COORDINATION LAYER
77
centrally defined and controlled guidelines. Manufacturing is based on highly standardised and controlled processes and technology so that the main production resources, i.e., extrusion lines and blue-collar workers, are highly mobile. In fact, they can be transferred between sites and put in S
System
D
Decision
P
Process 23
20
Centralisation / Responsibility
Autonomous
24
Standardised
6
Centralised 7
22
20 Product cost calculation 23 Short-term manuf. plan. 24 Manufacturing / Operations
Centralised &
8
9 11 12
18 19
1
3
2 15
4
5
Degree of standardisation for the network
within
weeks.
about
Although
manufacturing
two the
sites
produce
local for local, all this is made possible
due
to
a
highly
standardised product with little
17 Standardised 13 14 16 21 10
operation
P D S
local adaption. In addition, the strongly hierarchical culture and mind-set
of
the
employees
support the stability of the process landscape.
• The “autonomous network” contrasts the centralised and standardised network by assigning maximum responsibility and freedom to the sites while showing little standardisation. This position, again, is targeted as fundamental philosophy in networks which ask for a high degree of local responsiveness, e.g., in a local for local production setting with tailored products, or in networks carrying a wide range of competencies at the single sites, as for world factories. For other structures, this position is often reduced to mainly operative processes or responsibility areas with low priority. • The “standardised network” can be understood as middle way to reduce site authority by substituting parental control by standardisation, even in a decentralised organisation. Processes and decisions are executed or made decentralised but follow rigid standards and guidelines defined and controlled centrally; a position that requires high effort to enforce process discipline. Ex. 1.5: Most companies are reluctant to take the effort to establish standardisation in a decentralised network. For the survey sample, the figure below shows how the three responsibility areas are split up between the four generic network positions. The positioning is calculated as the mean value across all categories of a distinct responsibility area. 41 Obviously, only few participants opt for the standardised network; averaged across all categories for the decision area, the percentage is only 7%, while
41To
calculate the mean value per responsibility area, first, for each of the underlying responsibility categories the distribution of the survey sample between the four centralisation and standardisation types was derived, and then the results were averaged. However, not all categories as depicted in Fig. 21 and Tab. 12 were covered by the questionnaire and included in the calculation. While all types of systems were included, decisions number 10 and 13 and processes 20 and 24 were not considered.
78
DESIGNING THE NETWORK COORDINATION LAYER
accounting for 11% with regards to the process categories. So, the idea of substituting central control by setting global standards is not fully exploited. Instead, especially processes are either both standardised and centralised, or they are directly “outsourced” to sites without even standardising. But for all that, the Dental NW takes the effort to strive for the standardised position in its processes. Coming from an insular process landscape, activities in supply chain, production planning, quality S
System
D
Decision
P
Centralisation / Responsibility
Autonomous 18%
26%
8%
30%
integrated into a global ERP system. This
Standardised
34%
Centralised
management, etc. were harmonised and
Process
15%
49%
17%
system nowadays can be rolled-out to a new
11%
7%
site within shortest time. For this, processes must
Centralised & Standardised 33%
be
strictly
standardised.
Process
discipline is kept high through the top
30% P D S
Degree of standardisation for the network xx % = Mean value per responsibility area ∅ n.a.: S = 10%; D = 7%; P = 5%
managements’
attention,
strict
parental
control by specialists at the main sites, and by occupying key positions at new sites with young executives from the central operations
function. The development and improvement of the process landscape was, and still is, carried out bottom-up, picking successful approaches from every site and promoting them as global solutions.
The Logic Behind The centralisation & standardisation framework provides managers with an aggregated perspective on the allocation of authority and the degree of autonomy in their network. This is necessary since the responsibility areas cannot be considered independent; changing one might affect the position of another. Thus, instead of limiting the scope to single processes, systems, or decisions, the holistic view of the framework enables to understand the linkages between the responsibility areas / categories. Such a view, on the one hand, helps to detect inconsistencies for the AS-IS situation, and, on the other hand, it is essential to discuss changes when sketching the TO-BE situation. Ex. 1.6: Coming back to the Profile NW, the holistic view helped to detect an upcoming problem
?
Centralisation / Responsibility
Autonomous
6 Centralised
24 Standardised
Centralised & Standardised
Degree of standardisation for the network
6 HR system 24 Manufacturing / Operations ? ! P D S
= Cause = Problem = Target = Solution
caused by the interplay of certain responsibility
categories:
The
network management was facing a
decrease
in
global
standardisation particularly for manufacturing
processes
(24).
These vary between the different regions in which the sites are
DESIGNING THE NETWORK COORDINATION LAYER
79
located. One of the root-causes was revealed by the framework as depicted below. Since the human resource system (6) including the training and development activities of the blue collar workers was given under regional responsibility, a common understanding of the processes in the global network could no longer be assured, leading to local variations in operating procedures and to a relaxation of the process discipline.
Designing the autonomy in the network along the centralisation and standardisation dimensions can hardly be considered to be isolated from other decisions. Intuitively, it has to be matched with the network’s organisational structure. This structure defines how the management of the sites and the network are integrated into the company’s organisation; thereby, it predetermines the organisational levels and formal reporting channels for manufacturing activities. Several types of organisations are known from literature, ranging from a pure functional organisation with sites attached directly to the management board or to a distinct corporate function, over more complex regional, divisional, or matrix structures, to mixed types and hybrid forms. 42 From an operations perspective, a direct functional organisation, for example, can provide value due to the utilisation of functional speciality and by clearly assigned internal responsibilities and hierarchical guidance via a dedicated instance (the network manager). This drives centralisation but also makes global standards easier to enforce. More recent structures, like a divisional or regional mix with plants separated between different organisational units or regions, allow for regional or product specialities, or, like in a matrix organisation, can reduce functional boundaries (Diederichs et al., 2008). Nonetheless, in each case, increasing the number of organisational levels involved induces organisational complexity, which, in turn, challenges central authority and the effort for global standardisation. Ex. 1.7: Decision making in the Edgeband NW clearly depicts the underlying organisational
Centralisation / Responsibility
Autonomous
Standardised
Region Centralised & 8 11 Standardised
Centralised
12 13
14 15 16 17
9
Degree of standardisation for the network
42
Central unit
8 Site strat. & roles 9 Orga. structure 11 Make-or-buy 12 Product alloc. 13 Transfer pric. 14 Prod. process 15 Manuf. techn. 16 Long-term capa. 17 Short-term capa. P D S
structure.
Coming
from
a
centralised organisation, a socalled “regional model” was introduced with sites allocated in a matrix structure to a regional unit and to a central operations function. From an operations setting
perspective, revealed
the
several
For a more detailed discussion of organisation structures, see Section 4.1.1 and, for example, Slack and Lewis (2002) or Diederichs et al. (2008)
80
DESIGNING THE NETWORK COORDINATION LAYER
problems basically caused by overlapping authorities with diverging targets. Decision making for product allocation and capacity development gives an example. Until recently, each region had to cover its own fixed costs; this led to strong competition among the different regions with no willingness to refrain from workload on the behalf of others. Hence, due to the lack of a “final instance” for decision making, an optimal capacity balancing from a network perspective was substituted by regional (sub-)optima. To ameliorate the situation, responsibility for fixed costs and capacity development was handed over to the central operations function, but it still requires an alignment with the regions. The situation improved, but it leads the sites to strive for overcapacity as they no longer have to bear the respective costs. Faced with similar problems, the Pet Food NW and the Dental NW moved towards a strictly functional structure. Coming from an autonomous and decentralised network with sites belonging to five different business units, production in the Dental NW was separated as independent function. Nowadays, the eight sites offer products to the business units but also to external markets. Extracting the sites was pushed top-down, initially with strong resistance of the business unit’s management. Retrospectively, the central and independent coordination of the production function with an assigned network manager paid off by achieving a high degree of standardisation, enabling to improve manufacturing in terms of responsiveness and delivery speed, productivity, transparency, and quality.
Finally, it is not only the organisational structure but also the network configuration that might impact the allocation of authority. In literature, often the linkages between plant roles and autonomy have been discussed. As shown, several scholars have investigated whether a site’s degree of autonomy differs depending on its plant role. Maritan et al. (2004), for example, analyse the decision autonomy of planning, production, and control mechanisms based on the plant role model of Ferdows (1997). Similarly, Feldmann and Olhager (2009b) and (2011) found that the decision autonomy of a plant and its competencies are positively related. However, all these approaches neglect the overall network perspective and its dynamics. Rather than having a circumscribed effect on an individual site’s competencies, establishing or changing a plant's role is likely to impact all other plants in the network (Cheng et al., 2011). Any such change can affect or alter the complete plant role portfolio, and thereby influences the specialisation of the network configuration, which, in turn, might induce modifications of the allocation of authority. For instance, nominating a lead factory by assigning global responsibility for process development to a certain site enhances this very site’s authority, but it restricts the autonomy of the others and puts them in a position of dependence. Similarly, establishing such a centre of gravity for process development is also assumed to drive process standardisation.
DESIGNING THE NETWORK COORDINATION LAYER
81
Ex. 1.8: For the Seals NW, the framework enabled a target-oriented discussion of the changes induced by establishing a lead factory concept. The network represents a typical example of the Centralisation / Responsibility 5
S 6 11
System
19 22
D
Decision
3
20 23 24
Lead factory
Autonomous
P
Process
20, 23, 24
!
Standardised
14, 15
2
9 10
12 15 16 17 21
Centralised & Standardised
Centralised 1
3, 7
14
4 18
7 13
Degree of standardisation for the network
3 Quality system 7 Know-how system 14 Prod. process dec. 15 Manuf. techn. dec. 20 Product cost calc. 23 Short-term manuf. planning 24 Manufacturing / Operations = Cause ? = Problem ! = Target = Solution P D S
transition phase between the autonomous
and
the
centralised and standardised position.
At
the
very
beginning, globalisation was started by setting up sales and service centres tied to local markets. Since barriers for manufacturing were low in the company's core business
and central control poor, the scattered service centres successively built up production competencies to serve local markets on their own. Site managers considered themselves as independent entrepreneurs rather than as representatives of a global company. As a consequence, the company is currently facing challenges in restricting the local authority and strengthening parental control. In a first step, core systems and decisions are being moved to central responsibility while global standardisation will follow. The definition of the German manufacturing plant as lead factory is supposed to foster this transition. The lead factory is considered as middle way between centralisation and standardisation (Tykal, 2009). Responsibility for production process (14) and technology decisions (15), as well as for the definition of tools, methods, and standards for production quality (3) will be shifted to the lead factory; so is the organisation of production-related know-how exchange and best practice transfer (7). These steps decrease local freedom and are assumed to lead to a more harmonised and stable landscape for especially the operative processes (20, 23, 24).
Discussion of Findings & Implications The first coordination framework supports network managers in modelling centralisation & standardisation. It mainly addresses Martinez and Jarillo’s (1989) understanding of structural and formal coordination mechanisms, i.e., centralisation, formalisation, and harmonisation. Its introduction is completed by summarising (1) findings for the network coordination layer in specific, (2) conclusions for the design of an integral network architecture, and (3) practical implications for the network management. Tab. 13 structures the results accordingly.
Centralisation & standardisation framework
The framework directly touches three of the basic coordination mechanisms as introduced by Martinez and Jarillo (1989): - Centralisation / decentralisation, when defining the responsibility for system development, decision making, and process execution. - Formalisation and standardisation, when defining the degree of standardisation for systems, for decision making, and processes execution. - Harmonisation of systems and processes, when defining the degree of the systems' implementation in the network.
(1) … network coordination in specific The degree of site autonomy within a network can be modelled as a function of (1) the centralisation and (2) standardisation of (3) distinct responsibility areas (i.e., systems, decisions, and processes) and their underlying categories. Centralisation and standardisation in this context can act as substitutes in order to maintain parental control.
Findings and implications for … (2) … the design of a network architecture Centralisation and standardisation as a decision dimension of the coordination layer is not independent. Instead, its design is affected by other categories. An alignment is necessary, especially with the organisational structure of the network and the plant role portfolio as part of the network configuration.
(3) … the network management in general Twenty-four categories in three different responsibility areas were isolated to analyse and design the autonomy and authority in the network by evaluating their degree of centralisation and standardisation. These comprise: (1) systems for primary and (2) support activities, strategic decisions regarding (3) organisation, (4) products, and (5) processes, technology, and capacity, as well as (6) strategic and (7) operative processes. The completeness and terminology of the categories might vary between companies, making their adaptation and complementation permissible. The categories of the responsibility areas are not independent. Instead, an integrated approach reflecting their linkages and dependencies is valuable in order to analyse and design the degree of autonomy from the superordinate perspective of the network management properly.
82 DESIGNING THE NETWORK COORDINATION LAYER
Tab. 13: Findings on the centralisation & standardisation framework
DESIGNING THE NETWORK COORDINATION LAYER
83
3.3.2 The Resource Sharing Framework Framework Development & Description Resources enable an entity to operate in its competitive environment. Their provision and availability is often reason for competition in the network. According to Luo (2005), this “… competition arises because (…) resources are limited in quantity, and allocating and deploying them have to depend on the parent firm’s global strategy” (Luo, 2005, p. 75). The statement sheds some light on the nature of resources. First, since resources are tied capital and per se limited in quantity, defining their scarcity or their degree of availability for the network is of central managerial interest; already from an economic imperative. Second, their provision and allocation is crucial since both influence the interdependence among sites and their relationship to the central headquarters. Hence, resource allocation becomes a coordination mechanism for the network management to shape the continuum between competition and cooperation among the sites. The dedication of scarce resources to selected sites, for instance, strengthens their position and puts others into dependency; an effect that can be attenuated if resource sharing is intensified. Accordingly, Jaehne et al. (2009) point out resource availability and resource flexibility as two basic coordination strategies. The resource allocation & sharing framework as depicted in Fig. 22 builds upon these considerations. It introduces the intensity of resource sharing (x-axis), which is an indicator of resource flexibility, and the scarcity of the resource in the network (y-axis) as the two main dimensions to define the resource strategy. Scarcity in the network Limited amount
Competition
Rather limited amount
Cooperation 5
6
Balanced amount Rather sufficient amount
4
Sufficient amount
3
Dedicated Resources
2
Resource Pool 1
No exchange / Seldom and to sharing a small extent Extent of provision as proportion of possessing Resource categories: sites vs. requiring sites: 1 R&D capacity (structural and / or specialists) Possessing sites > requiring sites Possessing sites = requiring sites Possessing sites < requiring sites
2 3 4 5 6
Intensity of sharing / exchange Frequently and to a large extent
Engineering capacity (structural and / or specialists) Supply chain capacity (specialists) Basic manufacturing capacity (structural) Special manufacturing capacity (structural and / or specialists) Support functions (structural and / or specialists)
Fig. 22: Resource allocation & sharing framework
84
DESIGNING THE NETWORK COORDINATION LAYER
Analogous to centralisation and standardisation, the framework allows sketching the current and developing the future resource strategy from an integral perspective. For this, so-called resource categories have to be evaluated according to their degree of scarcity and their degree of sharing between sites. Sharing can occur either as physical exchange, e.g., by transferring machines and tools or by moving workers, or nonphysically by granting access to permanently installed resource. Additionally, the extent of provision, as the proportion of sites possessing and requiring the resources, needs to be defined, giving an indication of the demand and supply ratio; it is reflected by the size of the resource categories’ bubbles. The categories themselves are derived from literature and field-discussion. Tsai and Ghoshal (1998), for instance, propose information, products and services, personnel, and support as resource types. Likewise, Luo (2005) introduces technology, equipment, and talents, but also capital, know-how, and supplies as resources that are subject to competition between units. Both do not tailor their selection to manufacturing networks but to multinational enterprises in general. They also mix up resource, information, and knowledge exchange, which will be treated as separate coordinative issues in this study. In turn, the framework promoted here is based on an adoption of their categories, understanding resources either as (1) structural capacity, e.g., assets, machines and machine hours, and equipment, or as (2) personnel capacity / specialists that are required by the sites to compete in their environment. Further, both capacity types are separated functionally. This follows Feldmann and Olhager’s (2009a) findings that competencies are typically allocated as bundles to a site: starting with production-, supply chain-, and finally development-related competencies. Accordingly, the framework separates between (1) R&D and (2) engineering resources, which are mostly represented by manpower but might also comprise infrastructural equipment, (3) supply chain resources / specialists, (4) basic manufacturing resources, like production capacity (machine hours) and assets, (5) special manufacturing resources, as certain equipment, tools, or trained production specialists, and (6) internal support functions. Again, these categories provide a starting point for discussion, but they may need some company-specific refinement or complementation; especially the separation between structural and personnel capacity / specialists can sometimes be superficial. Integrating the discussed dimensions reveals four generic types of resource sharing strategies as reflected in the framework: • For the “dedicated strategy”, resources are allocated at almost each requiring site while availability is sufficiently high throughout the network. The strategy often accounts for autonomous sites acting responsively to meet fast changing customer requirements, like, for example, in a local for local structure, or for
DESIGNING THE NETWORK COORDINATION LAYER
85
world factories being capable of serving global markets with a short and unstable planning horizon. In both situations, providing the sites with a capacity cushion for production resources empowers them to cope with lead time expectations and to quickly react in terms of order size and delivery flexibility; similarly, the dedication of R&D and engineering specialists allows them to meet product range and design flexibility. In these cases, resources are often tailored to the facilities’ specific requirements, hindering any sharing. This, for instance, accounts for resources underlying structural restrictions, e.g., for machines with long changeover times preventing capacity exchange between different sites. Summing up, for this strategy, the traditional trade-off between cost and customer focus is made in favour of customer orientation while resources are often idiosyncratic, thus making any exchange difficult to realise. • For the “competition strategy”, resources are also dedicated to the distinct sites and not shared, but their insufficient amount leads to scarcity. This could be forced in favour of economic aspects or caused by limited availability. The position can induce competition especially between structurally similar sites asking for the same type of resource. Thereby, competition either arises for the initial allocation of the resources, e.g., for granting the financial budget for investment, or for their usage. Some network managers explicitly make use of such situation, igniting competition by taking resource allocation or access as measure to incentivise the site management. Ex. 2.1: The Edgeband NW shows how to transform the resource strategy into a coordinative mechanism. Located centrally and with free access, sharing the scarce R&D capacity (1) in the network until recently depended on a cooperative culture between the sites. With the internal demand
Scarcity in the network
Competition
Cooperation
1
1 R&D capacity
1
growing
unpredictable,
Dedicated Resources
No exchange / Seldom and to sharing a small extent
Resource Pool
Intensity of sharing / exchange
Frequently and to a large extent
= Cause = Problem = Target = Solution
and the
more
resource
became a bottleneck, making a fair
? !
bigger
allocation
difficult.
and
sharing
Consequently,
the
network management decided to change its strategy; by assigning a fixed R&D budget to every site at the beginning of the year, they
now aspire to better control for the overall demand. The height of this budget will be tied to the incentive system of the local site management, which can be influenced by the sites’ annual performance. Thus, the resource will be – although not physically – switched from a cooperative to a competitive position in the framework with each site fighting for its individual share.
86
DESIGNING THE NETWORK COORDINATION LAYER
• The “cooperation strategy” provides a relaxation of the competition strategy: It attenuates competition by fostering resource sharing. Moreover, the scarcity of the resources calls for cooperation between the sites when demand exceeds availability. In many cases, the responsibility to organise resource sharing and to balance demand and provision is assigned to a central entity. A typical example is manufacturing capacity in a web structure: Sites have similar competencies allowing for balancing manufacturing orders in case of local bottlenecks. Load levelling is carried out by a central supply chain function linking sales and production. This function consolidates and assigns orders in the network optimising global utilisation. Yet, resources also might get trapped unintentionally in this position, for instance, when demand for specialists or support functions increases in a growth period or due to seasonal peaks. Ex. 2.2: The Pet Food NW gives an example of how to benefit from the cooperation strategy. The network acts in a highly volatile and demanding business in terms of delivery speed and flexibility – failing to meet the challenging time fences is penalised directly by losing orders. In the past, this flexibility was met by the provision of dedicated production resources as capacity cushions at each site. Caused by the volatility, sites continuously asked for more capacity to react even to short demand
Scarcity in the network
Competition
Cooperation
4
?
4
Dedicated Resources
Resource Pool
3
3
No exchange / Seldom and to sharing a small extent
Intensity of sharing / exchange
Frequently and to a large extent
3 Supply chain specialists 4 Manufacturing capacity ? !
= Cause = Problem = Target = Solution
peaks, which – from a network perspective – summed up to a decreasing
overall
capacity
utilisation and higher fixed costs. To stop this vicious circle,
the
“swing-item
concept” was introduced. At each site, a defined amount of production capacity (4) is now
reserved for the manufacturing of so-called swing items; basically orders that can be allocated freely in the network. The coordination of the load levelling is done by a central supply chain manager (3) who aggregates and allocates the incoming demand. Although there is still some overcapacity needed, its overall amount can be better controlled. So, moving parts of the production capacity from a dedicated to a coordination strategy paid off by reducing fixed costs. Finally, the sites’ capability in terms of flexibility is substituted by the network’s capability in terms of volume mobility.
• The “resource pool strategy” is based on bundled resources which are accessible to all the requiring manufacturing sites in the network, causing a high intensity of sharing either by actually moving the resource or by granting free access to fixed assets. Since resources are sufficiently available, little
DESIGNING THE NETWORK COORDINATION LAYER
87
competition but also little cooperation is necessary to coordinate the sharing. This position is often aspired for important specialists or support functions that can be moved freely and of which a shortage impacts competitiveness, such as for project engineers or task forces. It is also found for physically tied but flexible resources, such as R&D or IT specialists located in a centralised department. In any case, a systematic approach is needed to split the costs for the provision of the pooled capacity. Ex. 2.3: The global production strategy team (GPS) at the Seals NW provides specialists for the support and guidance of the manufacturing sites. The team was formed in order to align the mostly independent and autonomous local production and service centres with the vision of a global
Scarcity in the network
Competition
Cooperation 6
6
!
1
Dedicated Resources
No exchange / Seldom and to sharing a small extent
2 Resource Pool
Intensity of sharing / exchange
6
? !
Frequently and to a large extent
6 Support specialists (GPS) ? !
= Cause = Problem = Target = Solution
manufacturing
network.
But
instead of working proactively on innovative solutions to strengthen the network’s competitiveness, the specialists got more and more stuck in the role as reactive “firefighters”. The framework led the discussion into two directions for relaxing
the
situation,
both
related to the creation of some slack resources. The first solution was to dedicate some local specialists to the individual sites to support the central GPS team. But these local positions would have had to be filled with freshly hired external candidates since the parent headquarters could not release internal specialists. For this, the fear emerged that these candidates could literally get “swallowed” by the local organisation, losing their connection to the central goal and vision. This is similar to the phenomenon explained by Ghoshal and Bartlett (1988) that local slack resources foster the development of local innovations but hinder the adoption of central innovations. Instead, the management decided for the second solution by taking external candidates to increase the manpower of the central team; i.e., moving it from a cooperation strategy into a resource pool and distracting it from the operational business.
The Logic Behind There is more to designing the resource strategy than just deciding about the physical allocation of assets or capacity. It reflects a coordinative mechanism to shape the sites’ interactions. Analogous to the centralisation & standardisation framework, the introduced approach provides network managers with an integral perspective on resource categories. These, again, might be interrelated, making a holistic perspective
88
DESIGNING THE NETWORK COORDINATION LAYER
valuable. But in addition to mapping the AS-IS situation, the framework can also be applied to anticipate dynamics and formulate proactive measures to respond. Ex. 2.4: In the context of the Profile NW, the framework was applied to map and redesign the pillars of the future tooling strategy. Tooling comprises three types of resources; first, engineers for tool construction (5a), second, capacity for tool testing and running-in at the local site (5b), i.e., the industrialisation process comprising the installation and ramp-up of a tool at the local extrusion line, and third, experienced running-in specialists (5c, 5c’) to master the ramp-up. Today, tool construction is pooled at a global tool shop providing its capacity to the network. The running-in process is
Scarcity in the network
5c
?
4& 5b
?
Competition
5c’
Cooperation
?
!
5b
Dedicated Resources
No exchange / Seldom and to sharing a small extent
5a
Resource Pool
Intensity of sharing / exchange
Frequently and to a large extent
4 Manuf. capacity 5a Tooling specialists 5b Running-in capacity 5c Running-in specialists (local) 5c´ Running-in specialists (central) ? !
= Cause = Problem = Target = Solution
conducted locally at the sites on the same extrusion lines which cover the daily business; thus, production (4) and running-in capacity (5b) stress the same resources. Running-in specialists are
highly
trained
local
operators (5c) which can be supported by a task force at the central tool shop (5c’). Adding
the anticipated dynamics to the set-up reveals several problems. Although still sufficient, the capacity for tool construction is expected to run short in the near future due to increasing internal demand. Further, seasonal peaks in customer demand use up local production resources to the debit of the running-in capacity, putting local and global running-in specialists under pressure. The future tooling strategy must be able to handle these changes: For the tool shop, IT support and process improvement is considered to increase the efficiency to cope with the steadily increasing demand for new tools. At the local sites, production capacity and running-in capacity will be strictly separated, reserving distinct extrusion lines to handle the running-in even during seasonal peaks, hence changing its position from competition to dedicated resources. Finally, the unstable request for local running-in specialists will be balanced by increasing the amount of the central task force employees and facilitating its sharing.
Moreover, the resource strategy is contingent upon the network structure. While especially operative resources of product or market area plants are mostly independent and tailored to their idiosyncratic challenges, which hinders any sharing, the same type of resources in general purpose plants is per definition flexible and interchangeable. The network specialisation, too, impacts resource allocation and sharing. A plant role portfolio, for instance, reflects the bandwidth of competencies of the distinct plants in the network. These competencies are built upon a site’s availability of, or access to,
DESIGNING THE NETWORK COORDINATION LAYER
89
underlying resources. While factories with limited competencies, such as outpost or offshore factories in Ferdow’s (1997a) terminology, require solely production capacity to fulfil their strategic purpose, factories with a higher strategic role are more demanding. If these sites are further intended to actively support the network, a need for resource sharing is also given. Again, establishing a lead factory approach illustrates this reasoning. A lead factory is a typical example of a central entity with a supportive character and a high degree of embeddedness in the network. It often bundles specialists for production, engineering, or R&D, which are moved into a pooling or coordination strategy. From a network perspective, such a step might be beneficial since resource bundling might release synergy effects, thus decreasing scarcity (or increasing an efficient exploitation of the resources). From a site managements’ perspective, upgrading the role of the designated lead factory corresponds to a downgrade of the other plants’ roles. Ex. 2.5: Coming back to the Seals NW, the aspired standardisation in operative processes is sought to be achieved by assigning global responsibility to a lead factory. This step is underlined by a
Scarcity in the network
Competition 2a
4
2b
5
Cooperation
2b
Dedicated Resources
No exchange / Seldom and to sharing a small extent
5
Resource Pool
Intensity of sharing / exchange
Frequently and to a large extent
2a Engineering capacity 2b Engineering specialists 4 Manufacturing capacity 5 Manufacturing specialists (processes)
concentration of the scattered specialists for engineering and production
processes.
As
depicted in the framework on the left: While the pure engineering and manufacturing capacity (2a, 4) will remain at the local sites, processes will be developed, rolled-out,
supported,
and
controlled by specialists at the lead factory (2b, 5). Besides improving standardisation, the targeted resource pooling strategy is also expected to have two other major impacts. First, the resource availability is assumed to increase due to the exploitation of synergy effects when putting specialists closely together. Second, the sharing of the specialists is considered to positively influence the network culture by ending the sites’ insular thinking and establishing personal relationships in the network.
Discussion of Findings & Implications Resource allocation & sharing have been introduced as second coordination framework. Examples have been given on how to apply the framework to map and develop a network’s resource strategy. Although such a strategy is tightly related to the network configuration, it basically shapes the coordination mechanism which Martinez and Jarillo (1989) refer to as departmentalisation, i.e., the (de-)grouping of structural or personal resources, and cross-departmental relations, e.g., when setting up
Resource allocation & sharing framework
The framework further addresses the degree of competition and cooperation between the sites by deciding about the resources to compete for or cooperate with.
The framework directly touches two of the basic coordination mechanisms as introduced by Martinez and Jarillo (1989): - Departmentalisation, when (de)grouping structural and / or personnel capacity (resources). - Cross-departmental relations, especially when setting up specialists in integrative departments (e.g., task forces).
(1) … network coordination in specific The resource strategy decides about the availability, allocation, and sharing of resources in the network. It can be modelled as a function of (1) the scarcity, (2) the degree of resource sharing, and (3) the extent of provision of (4) distinct resource categories.
Findings and implications for … (2) … the design of a network architecture Resource allocation and sharing is tightly connected to the network structure and specialisation as two main decision categories of the configuration layer. Both directly impact the resource strategy formulation.
(3) … the network management in general Six resource categories were isolated to analyse and design the resource strategy in the network; they are separated functionally and according to structural and personnel capacity / specialists. Functionally, they comprise: (1) R&D, (2) engineering, (3) supply chain management, (4) basic and (5) special manufacturing, and (6) support functions. The completeness and terminology of the categories might vary between companies, making their adaptation and complementation permissible.
90 DESIGNING THE NETWORK COORDINATION LAYER
specialists in integrative departments. Further, it can be utilised to drive competition and cooperation. The findings are summarised in Tab. 14.
Tab. 14: Findings on the resource allocation & sharing framework
DESIGNING THE NETWORK COORDINATION LAYER
91
3.3.3 The Incentive System Framework Framework Development & Description Incentive systems provide mechanisms to motivate an intended behaviour by facilitating desirable or restricting unwanted actions. Targets are defined and rewards held in prospect for meeting these targets. In a network, incentives are crucial for coordinating the behaviour of the site management, and it’s the challenge of the network manager to position the sites’ interplay between cooperation and competition (Chew et al., 1990; Luo, 2005). This position is influenced by the organisational level on which the targets are set. Referring to Salter (1973), Gupta and Govindarajan (1991) already stated in the context of a multinational enterprise that “… the incentive bonus for a division general manager need not always be a function of the focal division's performance; in fact, the incentive bonus may be tied partly (or even totally) to the performance of a cluster of divisions” (Gupta and Govindarajan, 1991, p. 781). For a manufacturing network, this raises the question whether to agree targets on the individual sites’ level, or whether targets should be collectively agreed on for a cluster of sites (or even the network in total). Targets on site level support individual behaviour or can foster competition; targets on the network level can create a basis for cooperation by addressing a common goal (Bartol and Srivastava, 2002). The position is also affected by the way rewards are allocated. Tying them to a site’s individual contribution, e.g., by connecting its performance with the height of the management’s bonus payment, is likely to ignite competition while allocating them to equal parts between sites can create a culture of mutuality. The incentive system framework as depicted in Fig. 23 builds upon these considerations. Targets are agreed …
Based on individual achievement / contribution of site
… above network level
2
… identically for all sites … individually for each site No targets set
1
Coopetition
… for the network / a group of sites
To equal parts
5
4
Individualism
3
Collectivism
Rewards are
Reward types: Autonomy & responsibility
Collaboration
x
Financial payments
Reputation & awards Reward mechanism not used
Performance categories: allocated ... 1 Overall financial performance 2 Market & sales performance 3 Operational performance 4 Contribution to learning 5 Conformance with strategic goals
Fig. 23: Incentive system framework
92
DESIGNING THE NETWORK COORDINATION LAYER
The framework supports network managers in the design of their incentive system. They have to decide about the performance categories to incentivise and the types of rewards applied. Performance categories can be separated into outcome-dependent and behaviour-dependent ones (Gupta and Govindarajan, 1991). Outcome-dependent categories have the advantage of basing any performance evaluation on “hard facts” underlying the target formulation and monitoring with tailored KPIs to measure the output. Such categories typically address the overall financial performance or the market & sales performance, but they can also be related to operational performance, such as lead times, inventory levels, or the overall equipment efficiency. They require transparency and create comparability, which makes them suitable to drive competition among the sites, but their definition can be hard to account for local specialities. Behavioural categories are more difficult to measure. In the framework, they are divided into a site’s contribution to learning in the network (Luo, 2005), e.g., by sharing its knowledge and experience with others, and its conformance with strategic goals, such as mastering special projects or fulfilling a designated plant role. Network managers have to align the selection of performance categories with adequate reward types. While financial payments / monetary boni are probably the most popular, others might be especially beneficial in combination with behavioural performance categories or as complementation. Following Luo (2005), an incentive system to foster knowledge sharing and best practice exchange “… may (also) entail such rewards as increased percentage of retained earnings, name recognition as an excellence center or global champion, higher autonomy dedicated by corporate headquarters, and greater resource support for future operations” (Luo, 2005, p. 86). The framework encompasses these dimensions, differentiating between financial payments, autonomy & responsibility, as well as reputation & awards. The combination of performance categories and appropriate reward types has to be positioned in the framework along the (1) organisational level that is subject to target agreement and (2) the way rewards are allocated. The organisation level ranges from single sites, which might be incentivised individually or by identical targets, to a larger group of sites, or even the overall network. In addition, organisational levels above the network are conceivable, such as a superordinate division or company (Gupta and Govindarajan, 1991; Bartol and Srivastava, 2002). Regarding their allocation, rewards are often tied to a site’s individual achievement or contribution, directly compensating for its performance, but in some cases, they might be shared to equal parts between sites independent of their individual performance. Combining the single dimensions reveals the four positions in the framework:
DESIGNING THE NETWORK COORDINATION LAYER
93
• For the “individualism position”, targets are agreed on site level and rewards are allocated according to the sites’ individual achievements. This position can be aspired for basically all network types but is found for different reasons. Concentrating on outcome-related performance dimensions, as long as the products and underlying processes of the plants are different, such as in a product or a process structure, targets are formulated individually on site level to give a clear direction to each plant manager. As soon as the plants are structurally similar and the outcome comparable, e.g., for market area or general purpose plants, common targets can be appropriate to ignite internal competition; this is often the case when establishing an internal performance benchmarking. Moreover, network managers can also make use of this position to foster cooperation between sites, e.g., by explicitly forcing a centre of excellence to fulfil its dedicated responsibility to share its expertise. Ex. 3.1: The Edgeband NW illustrates how the individualism position is occupied to ignite competition in the network – an example that is to some extent representative for nearly all the other case networks as well. The network composes a set of structurally very similar sites serving local markets with highly standardised products and processes. This context makes the operational performance (3) well comparable. Performance, in turn, is evaluated along four main dimensions: Targets are agreed …
Based on individual achievement / contribution of site
… above network level
Coopetition
… for the network / a group of sites
… identically for all sites
… individually for each site
3
To equal parts
Collaboration
3 Operational performance
process productivity energy
general,
consumption
and
and
wastage.
Each
dimension is broken down
Collectivism
No targets set
in
costs, as well as material costs
Individualism
performance,
into a set of centrally defined Rewards are allocated ...
KPIs. Targets for the site managers
are
formulated
based on this common set of KPIs defining the reference point for their financial rewarding. But, as part of an on-going internal benchmarking approach, the performance along each dimension is also regularly made transparent in the network, and – to ignite competition – sites are ranked based on their results. Though comparability is generally given due to structural similarity, there were debates in the past with some site managers questioning their benchmarking results particularly for the process performance dimension. Hence, a standardised “benchmarking product” was defined and the performance is now calculated using this product.
• For the “collectivism position”, targets are also agreed on site level but rewarded equally and independently of the sites’ contribution. The position is
94
DESIGNING THE NETWORK COORDINATION LAYER
often found to finance a central support function which all sites can benefit from but which cannot be afforded by a single site alone. Instead, all entities pay a fee that is calculated as percentage of their financial or strategic performance to fund a common service or support function that, then, is shared between all sites, i.e., as sort of “common property”. Thereby, the service’s perceived value has to justify the sites’ expenditures. The position gives priority to the collectivity, allowing balancing performance differences between sites. Such differences can be driven by external factors, e.g., when facing volatile markets, but also by internal reasoning, e.g., when burdening selected sites with time and resource consuming special projects that go beyond daily business. Ex. 3.2: The Excitation NW, as a business unit, is part of a larger division embedded into the organisation of a technology-driven multinational enterprise. The division comprises several networks acting independently but relying on a central R&D function. The R&D function is pooled at the European headquarters assuring access to local know-how and skills. It provides its technical competence to all the distinct networks. This support is crucial in constituting the technical supremacy of the products. Aware of its value, there is a common agreement on a joint funding of the function based on a fee tied to each networks’ annual financial performance. The volatility of the different businesses accounts for a balanced charging of the single business units over time. The Seals NW faces similar considerations with regards to the financial founding of the network’s lead factory support. Since the lead factory remains a manufacturing entity itself serving local markets, not allocating the expenses for its support would directly charge its own manufacturing costs. Hence, different approaches are considered, ranging from a general fee paid by all sites, which is similar to the collectivism solution in the Excitation NW, over a direct payment of the requested support service, to mixed approaches covering basic support by a low fee but charging a more intensive engagement of the lead factory separately.
• For the “collaborator position”, targets are agreed above a single site’s level, e.g., for a group of sites or for the whole network, or even above the network level. But similar to the collectivism position, they are compensated equally and independently of the sites’ contribution. Targets above network level are connected to a company’s overall performance, as in profit or gain sharing programs when tying the employees to a company’s success by holding out a certain proportion of the annual earning (Henderson, 2005). Any of such group targets are supposed to be beneficial in supporting an atmosphere of cooperation and in fostering knowledge and innovation sharing (Bartol and Srivastava, 2002). But generally, allocating rewards to equal parts might induce
DESIGNING THE NETWORK COORDINATION LAYER
95
the risk of free-riding since smaller contributions by one site can be substituted by higher contributions of others. However, this risk can be mitigated when keeping the group of actors involved small and making the contribution of each traceable and transparent. Ex. 3.3: The Floor Care NW and the Dental NW give some examples of how to utilise and control the collaboration position to actively foster cooperation in the network. In case a distinct plant or several sites in the network should perform very poorly, the operational (3) or financial performance targets Targets are agreed … … above network level … for the network / a group of sites
… identically for all sites
… individually for each site
Based on individual achievement / contribution of site
Coopetition
To equal parts
Collaboration 1
Individualism
No targets set
3
1 Overall financial performance 3 Operational performance
(1) for these low-performing sites are defined as “group targets”, i.e., targets which are copied identically into the bonus systems of the managers at both the well-
Collectivism
and Rewards are allocated ...
the
low-performing
sites. Thereby, the reward level
of
the
“high-
performer(s)” is tied to the progress of the ”low-performer(s)”. Setting these types of group targets, cooperation is triggered by encouraging strong site(s) to actively provide support to and share knowledge with weaker sites and by encouraging weak site(s) to accept this support. Free-riding is avoided since the group of sites involved is kept small, and all actors have to contribute actively to the realisation of the common goal.
• The “coopetitor 43 position” enables a second approach to reducing free-riding. Even if incentives are set on a network level, it might be appropriate to connect a site's rewards – at least partially – to its individual contribution to the overall network target. Hence, while cooperation is required to achieve the common target, yet, competition evolves regarding the absolute height of individual rewards. The position is applicable especially for operative performance categories, which can often strongly benefit from cooperation. A typical example is the definition of a network-wide target for inventory level reduction that cannot be achieved individually by any single site. Meeting the target is the prerequisite for the payment of compensation, but in the end, each plant manager is rewarded individually based on his personal reductions.
43
Just as a reminder: “Coopetition”, as originally formulated by Brandenburger and Nalebuff (1996), states that actors can simultaneously compete and cooperate in selected areas.
96
DESIGNING THE NETWORK COORDINATION LAYER
Ex. 3.4: Results of the survey sample emphasise the reluctance of most networks to explicitly incentivise knowledge sharing. Only 39% of the networks participating in the survey agreed to set particular targets for “contribution to learning” while 34% stated not to use any such targets; 27% gave no answer. One reason for this might be the existence of indirect mechanisms facilitating bilateral support and know-how sharing, as in the example of the Floor Care NW or the Dental NW; both rely on group targets on operational performance. Another reason could be the lack of adequate approaches for the implementation of direct incentives for knowledge sharing. The example of the operational performance benchmarking at the Edgeband NW, as introduced above, illustrates how the incentive system was utilised as main method to actively foster knowledge Targets are agreed …
Based on individual achievement / contribution of site
… above network level
Coopetition
… for the network / a group of sites
… identically for all sites
… individually for each site
2 3
1 Individualism
To equal parts
Collaboration
4
3 Operational performance 4 Contribution to learning
and best practice exchange. Based on the initial KPIdriven
internal
site
benchmarking, a knowledge data base was established, forcing the sites not only to
Collectivism
No targets set
make their performance but Rewards are allocated ...
also along
their the
improvements four
distinct
performance dimensions transparent to the network. To do so, the monetary potential of each generated idea is evaluated by an independent expert group. Further, the sites’ incentives on operational performance (3) are complemented by rewarding their level of cost-saving potential promoted, i.e., their contribution to learning (4). Currently, also efforts are made to move the system from the individualism to the coopetition position. Parts of a site’s individual rewards on cost-savings will only be granted if the network in total achieves an average performance progress above a certain percentage. Hence, not only the passive knowledge contribution of each actor but also the active implementation of the ideas in the network is fostered.
The Logic Behind The initial "gut feeling" that most networks still follow an individualism strategy, wherein they basically limit their incentive system to outcome-based performance categories linked with financial rewards, has been backed by both the case networks discussions and the survey results. Instead, the framework offers an integral perspective with multiple and more subtle approaches for managers in charge of designing their incentive system. It is built upon the fundamental understanding of such system as primary coordination mechanism to influence the degree of competition and cooperation between the sites. In this context, the combination between performance categories and reward types can be fundamental. It is not only
DESIGNING THE NETWORK COORDINATION LAYER
97
important “what” to incentivise but also “how”, assuming that different performance categories ask for different rewards. Fig. 24 outlines the survey results regarding the utilisation of performance categories and applied reward types. Utilisation of performance categories 1
Overall fin. performance
0%
91%
3
Operational performance
4%
84%
Conformance with strategic goals
16%
2 4 5
Market & sales performance Contribution to learning
16 %
n=56
Reward type: Autonomy & responsibility
Reputation & awards n.a.
75 %
x
Overall financial performance
Financial payments
16 %
2 22 %
48 %
n=50
Market & sales performance
16%
39%
27%
68%
no
n=56 32 %
1
16 %
73%
34%
Reward types applied per category* 16 %
11%
yes
17 % 33 %
3 31 %
54 %
n.a.
3%
43 %
4
13%
16%
32 %
n=54
n=37
Contribution to learning
* Restricted to networks answering the question regarding the utilisation of the performance category with “yes” or “n.a.”
28 %
34 %
5 32 %
35 %
Operational performance
9%
23 % n=47
Conformance with strategic goals
Fig. 24: Performance categories & reward types
Most networks focus on output-related performance categories: in fact, 91% of the study’s participants on overall financial performance, 73% on market & sales performance, and 84% on operational performance. While “money making”, i.e., targeting the financial and market performance, is clearly dominated by monetary compensations, in some cases, different approaches are applied. This is true especially for behavioural performance categories but partially also for rewarding operational performance. Actually, the benefit of behavioural performance dimensions and nonmonetary rewards is nothing new to research; nonetheless, managers have obviously been reluctant to consider it in the context of a manufacturing network. Reasons for this might be difficulties in measuring performance and the fact-based, engineeringdriven mind-set among European manufacturers. Yet, holding out reputation and awards – like the well-known "lean awards" or the nomination of a "centre of excellence" – have proven to be successful measures in facilitating individual sites' participation in knowledge sharing and learning. Similarly, motivating wellperforming sites with free access to centrally organised training or education programs for key employees might be valuable. Another approach to foster cooperation is to sensitise for incentives above network level. In particular, strongly linked or structurally similar sites offer potential for being streamlined to cooperate on a common goal; moreover, group targets for a subset of
98
DESIGNING THE NETWORK COORDINATION LAYER
sites can be applied, forcing internal support among smaller groups. Such cooperative culture is valuable as foundation for a joint “network thinking”, but it lacks any guidance for poorly performing plants by negatively affecting their motivation for continuous improvement. Internal benchmarks between sites are often established to adjust for this, but they put high requirements on the comparability of the sites as well as on the data transparency and quality of the measurement system. Regarding this, Chew et al. (1990) already stated that “… managers will need to understand the system in some detail. For example, lagging managers will need to understand why they are lagging. If the quantitative measures of managerial performance are not true to the managers’ performance, the system can go badly wrong” (Chew et al., 1990, p. 157). Although information technology has undergone a major progress, defining and comparing KPIs still led to debates and leaves room for improvement in most of the case networks. This shifts the discussion to configuration decisions driving the comparability between the sites, thus influencing the design of the incentive system. As indicated above, it is particularly the structural network configuration which determines the multiplant strategy, thereby setting the basic pillars for the plants’ characteristics. The operational performance of process plants, for instance, is hard to compare, which hinders initiation of competition. In contrast, the high degree of interdependence of such plants working on a common product makes cooperation easier to incentivise. Product plants, in turn, similarly have different characteristics but lack the goal to jointly work on a product; a fact that reduces the common denominator for both cooperation and competition, thus calling for individual targets on site level. On the other hand, market area or general purpose plants that are basically built upon identical processes and technologies allow for a benchmarking of the operational performance; their structural similarities increase the impact of shared knowledge and best practice exchange, too. Moreover, it’s not only the network structure but also the network specialisation which has to be taken into account when designing the incentive system. The nomination of plant roles can be applied as reward mechanism motivating the site management, but it also reflects the network managements’ perception and expectations regarding a site’s competencies and performance. This perception has to be accounted for when formulating incentives. The performance of a pure manufacturing site in terms of productivity, for instance, is hard to compare with the performance of a lead factory being distracted from large scale production due to process development or industrialisation obligations. On the contrary, expecting the same amount of costsavings due to newly generated ideas from both the lead factory and the production plant would either waste the potential of the former, or it could bring the latter into
DESIGNING THE NETWORK COORDINATION LAYER
99
trouble. Hence, especially in the case of internal competition, the interpretation of KPIs for target setting needs adjustment based on the intended plant roles. Similarly, the specific capabilities of sites have to be considered when going for network goals, being aware that the sites’ contribution can vary enormously. Ex. 3.5: The management of the Profile NW recently decided to establish three different types of sites roles: a lead factory, standard manufacturing sites in established markets, and start-up sites for Operational performance KPIs dimensions Costs
low
med.
high
Quality conformance
incentivising
the
individual
sites’
performance is based on a joint set of
COGS* per ton
Overhead costs (absolute) Investments (absolute)
Quality specification
expansion in new markets. Measuring and
CIP** savings p.a.
Customer compl. p.a. (#) Waste per ton (%)
Delivery speed Finished goods availability (%)
operational performance dimensions and underlying
KPIs.
Yet,
the
different
expectations of the network management regarding the contributions of the certain plant roles ask for adjustments on the individual
target
profiles.
Overall,
specification quality and product range flexibility are critical order qualifiers not allowing for any differences between the
Product range flexibility
Target profiles:
Changeover time (h)
* COGS = Cost of goods sold Standard manuf. sites ** CIP = Continuous improvement Lead factory Start-up sites
sites. The overhead cost and investments, on the other hand, reflect the differing tasks of the sites; they are particularly high for the
lead factory, which locates the pool of manufacturing and tooling specialists, and low for start-up sites. Likewise, the requirements regarding the cost-saving potential due to continuous improvement activities vary. While they are demanding for standard manufacturing sites and the lead factory, startup sites are only expected to make little contributions, focussing on market penetration instead of process optimisation; similarly, delivery speed is of less importance for these start-up sites.
Discussion of Findings & Implications The incentive system complements the set of coordination frameworks by a second lever to shape the degree of cooperation and competition in the network. It is linked to Martinez and Jarillo’s (1989) socialisation process as basic coordination mechanism, teaching a corporate culture by rewarding the do’s and sanctioning the don’ts. Its implementation also addresses what the authors refer to as formal output and personal control mechanisms, e.g., when incentivising output-related performance tying rewards to a site’s achieved performance level, or when incentivising personal behaviour tying rewards to the fulfilment of strategic goals. Analogously, the incentive
Incentive system framework
The framework touches three of the basic coordination mechanisms as introduced by Martinez and Jarillo (1989): - Output and personal control mechanisms, when incentivising output-related performance tying rewards to its performance level or behaviourrelated performance tying rewards to its fulfilment. - Lateral or cross departmental relations, when fostering cooperation in the network, e.g., by setting targets above network level or by incentivising contribution to learning. - Socialisation, when teaching a competitive or cooperative corporate culture and shared objectives by rewarding the do's and restricting the don'ts.
(1) … network coordination in specific The incentive system decides about the degree of competition and cooperation between the sites in the network. It can be modelled as a function of the (1) organisational level of target setting and the (2) mechanism of reward allocation for a combination of (3) performance categories and (4) rewards types.
Findings and implications for … (2) … the design of a network architecture Designing the incentive system is dependent on the network structure and specialisation. Both influence the fundamental potential for cooperation and / or competition as well as the comparability between the sites.
(3) … the network management in general Five different performance categories were isolated to analyse and design the incentive system in the network; linked with three types of reward mechanisms. They comprise: (1) overall financial performance, (2) market & sales performance, (3) operational performance, (4) contribution to learning, and (5) conformance with strategic goals. The completeness and terminology of the categories might vary between companies, making their adaptation and complementation permissible.
100 DESIGNING THE NETWORK COORDINATION LAYER
system also addresses lateral or cross-departmental relations when fostering cooperation. Findings are summarised in Tab. 15.
Tab. 15: Findings on the incentive system framework
DESIGNING THE NETWORK COORDINATION LAYER
101
3.3.4 The Information & Knowledge Sharing Framework Framework Development & Description According to Doz et al. (2001), the success of multinationals often stems from their ability to “sense” information, to translate it into knowledge, and to share it rapidly in the network. For manufacturing networks, Chew et al. (1990) pointed out at information sharing as main coordination mechanism already in the early 1990s by noting that “… network managers need to recognize their obligation to manage information flows with at least as much attention as they manage physical flows” (Chew et al., 1990, p. 156). Decisions have to be made about the “What, Where, When and How to share” questions (Cheng et al., 2008). “Where to share”, respectively the question of which sites to involve, might require a case-by-case decision for a distinct piece of information or a certain type of knowledge, but generally, it is determined by defining the exchange structure and transparency. The exchange structure predetermines the channels how to collect, process, and distribute information and knowledge among the sites. Transparency, in turn, defines the sites’ accessibility to the available information and knowledge. The construction of the information & knowledge sharing framework as displayed below relies on these two dimensions. Information & knowledge categories I
Exchange structure
External Info.
I
Internal Info.
K
External information
Knowledge
1 2 3
Centrally provided
Internal information
Transparency
Limitation Centrally coordinated
6
7
8
Medium Low
13
9
15
I: Availability of information K: Intensity of sharing
4
5
3
No access
Access to limited data / info.
Access to most data / info.
Access to full data / info.
No access
Access limited to selected sites
Access for all requiring sites
Access for all sites
Exchange mechanisms information Informal channels, e.g.: • ad hoc telephone calls & e-mails, meetings • social activities
I
4 5 6 7 8 9
Site strategy & roles Financial site performance Market & sales performance Operational site performance Sales & operations planning Manufacturing planning data …
Knowledge 11
Networking
Isolation
No exchange
12
14
Decentralised
10
2
1
Centralised & decentralised
High
Markets / Customers Competitors Suppliers …
10 11 12 13 14 15
Product innovations Product changes / improvements Technology / process innovations Manufacturing best practice Management experience & practice Business & supp. proc. improv. …
I Degree of transparency K
Exchange mechanisms knowledge
Formal channels, e.g.: • databases & sharepoints, intranet • regular & formal meetings
Moving people / job rotation
Customised projects / project support
K Competence Training & groups qualification Manuals, systems, databases Mechanism not used
Fig. 25: Information & knowledge sharing framework
102
DESIGNING THE NETWORK COORDINATION LAYER
The framework supports managers in mapping and designing their “information and knowledge network”. Distinct information & knowledge categories can be positioned based on the evaluation of their exchange structure and degree of transparency. For this, the exchange structure follows Chew et al.’s (1990) differentiation between a decentralised “plant to plant” flow with little and only indirect central guidance and a centralised “plant to network” flow with central coordination. As a consequence of the field-discussions, the exchange structures are complemented by splitting the central structure into a “centrally coordinated structure”, reflecting the original idea of a designated hub which is responsible for the collection and distribution of information and knowledge, and a “centrally provided structure” with only a single source for creation and promotion. Further, a mixture between the centralised and decentralised structure is possible, either in a transition period or also intentionally evoked by relying on a central flow while allowing for some decentralised exchange. The degree of transparency, as second dimension, gives managers the opportunity to control the flows by granting or restricting access to knowledge and information to certain sites or to a certain piece of data. Deliberately restricting access, for example, provides a form of parental control and gives headquarters a mechanism to manipulate site activities. The information & knowledge categories define the “What to share”. Tab. 16 outlines classifications of such categories in the context of multinational enterprises and manufacturing networks by several scholars. Fundamentally, some of them separate information (and data sharing) from knowledge (and innovation) exchange. Gupta and Govindarajan (1991), for instance, define knowledge as the “… transfer of either expertise (e.g. skills and capabilities) or external market data of strategic value” (Gupta and Govindarajan, 1991, p. 773) and distinguish it from administrative data. What both term “expertise”, is similar to other authors’ understanding of knowledge. It typically covers product, process, and management know-how and innovation (Ghoshal and Bartlett, 1988; Luo, 2005; Cheng et al., 2008). Luo (2005) complements this selection by adding knowledge about operational capabilities in support activities. What Gupta and Govindarajan (1991) term “external market data” comprises information about customers, competitors, and suppliers. Their idea of administrative data, in turn, was originally focused on planning and financial information, but Vereecke et al. (2006) refined it in the context of a manufacturing network, adding purchasing requirements, forecast data, inventory levels, and production plans as examples.
DESIGNING THE NETWORK COORDINATION LAYER
103 Authors
Information & knowledge categories 1 2 3 4 5 6 7 8 9
Ghoshal & Bartlett Gupta & Govindarajan (1988) (1991)
Information about markets / customers Information about competitors Information about suppliers Site strategy & roles Financial site performance Market & sales performance Operational site performance Sales & operations planning Manufacturing planning data
10 Product innovations
External market data (customers) External market data (competitors) External market data (suppliers)
Cheng et al. (2008)
Information about market trends Information about sources of supply
New product devel. Throughput processes Ideas for product (product design) development & introd.
New process devel. Throughput processes & introd. (process design)
New mgt. system devel. & introd.
15 Business & support process improvements Not considered
Vereecke et al. (2006)
Luo (2005)
Admin. information (financial information)
11 Product changes / improvements
12 Technology / process innovations 13 Manufacturing best practice 14 Management experience & practice
Tsai & Ghoshal (1998)
Input processes (purchasing skills) Throughput processes (packaging design) Output processes (marketing know-how) Output processes (distribution expertise)
Admin. info. (forecasts) Admin. info. (inventory) Admin. info. (product. plans) Techn. knowledge Product Product innovations specifications (products) Product specific processes Product specific mgt. knowledge Techn. knowledge Process Production know-how innovations (technology) (processes) Production know-how (operations) Mgt. experience & Managerial Mgt. knowledge knowledge innovations (production system) Mgt. knowledge (cooperation) Operational knowledge Admin. info. (purchasing)
Tab. 16: Information & knowledge categories
The framework builds upon the literature-driven isolation of the categories with some modifications as a result from discussion with the case networks. Especially “external market data” will be treated as information instead of knowledge, requiring similar approaches for collection, storage, analysis, and sharing. Therefore, the information categories are split into (1) external information comprising markets, customers, and suppliers and (2) internal information covering strategic, financial, and operational performance, and also planning data. Regarding the (3) knowledge categories, the separation into products, technology / processes, as well as management experience has proven to be stable in discussions; it is additionally enhanced by dividing each into innovations and (smaller) improvements (or best practices). Again, it should be noted that these categories give a comprehensive overview but might require some companyspecific renaming or modification.
104
DESIGNING THE NETWORK COORDINATION LAYER
The “How to share” defines the exchange mechanisms applied for diffusion. For information categories, the framework focuses on two basic mechanisms only: formal channels (such as regular and formal meetings or databases) and informal channels (such as ad-hoc telephone calls, informal ties, and social activities), reflecting either hierarchical guidance or informal lateral relations (Tsai, 2002). Mechanisms used are indicated by the coloured segments embracing the information category bubbles. When it comes to knowledge sharing, the differentiation is more subtle: A wider range of tools and methods from which the network manager can select is embraced. In any case, their selection should be aligned with the type of knowledge to be transferred (Ferdows, 2006). Building upon Polanyi (1958) and Nonaka (1991), Ernst and Kim (2001) and (2002), for example, describe knowledge diffusion between entities as a process of conversion between tacit and / or explicit knowledge. Explicit knowledge is formally codified and can be “… combined, stored, retrieved, and transmitted with relative ease and through various mechanisms (while) tacit knowledge refers to knowledge that is so deeply rooted in the human body and mind that it is hard to codify and communicate. It is knowledge that can only be expressed through action, commitment, and involvement in a specific context and locality” (Ernst and Kim, 2001, p. 12). Four types of conversion are conceivable: (1) socialisation as a tacit-totacit conversion, (2) externalisation as a tacit-to-explicit conversion, (3) combination as an explicit-to-explicit conversion, and (4) internalisation as an explicit-to-tacit conversion. Examples are given on how each conversion process can be underlined by a suitable combination of transfer mechanisms, ranging from hand outs, blueprints and manuals, over on-site observations, to training or the temporary dispatching of specialists. Ferdows (2006) complements the level of codification with the change rate of knowledge. He introduces a typology of production know-how depicted in Fig. 26. Tacit
Form of production know-how
Codified
1 4
Moving people Manuals and systems
Slow
2 3
Projects Joint development
Fast Speed of change of production know-how
Fig. 26: Production know-how & primary transfer mechanisms (adapted from Ferdows (2006))
DESIGNING THE NETWORK COORDINATION LAYER
105
For each type, again, appropriate transfer mechanisms are suggested, covering (1) moving and rotating people, (2) temporary project support, as often granted by inhouse consultants or support functions, (3) long-term joint development activities and projects, and (4) manuals or systems supported by tailored trainings. For knowledge sharing, the framework refers to this classification, introducing moving people & job rotation, customised projects & project support, training & qualification, manuals, systems & data bases, as well as competence groups as concrete mechanisms. Finally, the “When to share”, i.e., the state and maturity that information and knowledge have to reach before being transferred (Cheng et al., 2008), is considered only implicitly. With regards to information sharing, this question is related to the availability of information; therefore, the size of the bubbles indicates whether the information required by the network manager could be provided adequately by the sites. For knowledge, this size reflects the intensity of sharing. In both cases, the measure primary assesses the degree of information & knowledge in-flow to the network. To fully cover the “When to share” question, appropriate measures need to be set in place to explicitly assure the quality. Summing up, the framework reveals the following four positions for information & knowledge sharing: • In the “isolation position”, information & knowledge sharing between sites occurs only sporadically and mostly decentralised while transparency is limited. A network in this position can be described by a set of what Vereecke et al. (2006) refer to as “isolated factories”. This position typically lacks formal exchange channels and hierarchical pressure; lateral ties are weak. Sites are often either organisationally or structurally independent, reducing the site manager’s perceived benefit of any exchange, or they are highly competitive, causing a reluctance to expose any potential competitive advantage. Besides intentional protection of the own knowledge base, absence of central pressure for sharing can also foster the site managers’ myopia. Since incentives are missing, they tend to get stuck in daily business, not even being aware of the information asymmetry in the network and the potential value their specific knowledge might have for others. Ex. 4.1: The lack of a common “network thinking” in the Seals NW is also reflected by the design of the information & knowledge flows, revealing a typical example of the isolation position. Besides the centrally coordinated collection of some financial performance data (5) and the provision of product innovation by the R&D function (10), there is only little exchange, neither of information nor of knowledge. As discussed, the reason for this lies in the historical development of the network which evolves only slowly from independent and autonomous market area plants. Especially the isolation
106
DESIGNING THE NETWORK COORDINATION LAYER
position for product changes / improvements (11) and innovations in technology / processes (12) Exchange structure
I
I
External Info.
Internal Info.
K
Knowledge 10
Transparency
Limitation 5
!
13
Networking
Isolation
4
6
11
12
4 5 6
9 10 11 12 13 I
9
K
Degree of transparency
= Cause
? = Problem ! = Target
= Solution
Site strategy & roles Financial site perf. Market & sales performance Admin. manuf. data Product innovation Product changes / improvements Techn. / process innovations Manufacturing best practice
turned out to be a central barrier as local adaptations dilute any attempt
to
achieve
global product quality and
process
standardisation. network
The
management
struggles to move both categories
into
a
centrally coordinated transparency position. In a first step, it was intended to centralise the exchange of product and process knowledge. But especially for product changes, the scattered IT landscape turned out to be a main obstacle to harmonise the product data management, and thus to make product changes transparent and accessible in the network.
• In the “networking position”, sites evolve towards Vereecke et al.’s (2006) “active or hosting network players”. The exchange, again, is anchored in primary decentralised structures, but a higher degree of transparency and intensity of sharing is achieved. Therefore, lateral ties between the management and functional staff of the distinct sites have to be strong. Two different transformations have been investigated to approach this position. Starting from an isolated position, high effort might be necessary to motivate any sharing, e.g., by setting appropriate incentives or by creating a joint “network thinking” to establish social ties. For a network manager, such movement requires strong integrative and mediating skills. Coming from the coordinated transparency position, decentralised and bilateral exchange might occur as peripheral matter, e.g., when a low-performing site tries to get deeper into the “success-story” of a high-performing site and attempts to bypass the central and formal structures. In this case, the network management faces a trade-off between the benefit of informal ties and the risk of diluting defined structures, which, in turn, might impact process standardisation. Ex. 4.2: Mastering the transition from the isolation to the networking position at the Dental NW can be ascribed to three “success factors”. First, shifting sites previously formally tied to independent business units to an independent manufacturing function was accompanied by the nomination of a dedicated network manager. To provide guidance for this step, the function was originally considered as superordinate position centralised at the headquarters. However, it was decided in the end that it
DESIGNING THE NETWORK COORDINATION LAYER
107
should be hold by a site manager in personal union to create a better standing of the network manager against the other site managers. Second, the coordination of the network is not based on a strict hierarchical and decentralised structure, but it is built on close communication between its actors, ranging from monthly cross-divisional strategy meeting, over monthly site meeting with the network management and the site heads for operative issues, to weekly telephone calls. Additionally, every other month a site information meeting is scheduled to keep the employees up-to-date about the business development. Furthermore, monthly jour fixes between the network manager and his direct employees are held to discuss new ideas and future trends. Except for the jour fixes, each meeting is formally organised, based on a jointly agreed agenda, strict rules, and a distinctive meeting culture. The strong formalisation of communication provides a suitable method for information & knowledge exchange. As the third factor, moving people is consequently applied as transfer mechanism. Management positions at new or acquired sites are occupied by young executives trained at the headquarters or mature sites; a step assuring a common leadership and management understanding and strong lateral ties at the executive level. Overall, the network coordination is dominated by a very present and active network manager with a clear vision, an integrative character, and strong communications skills. Nonetheless, since this approach entails the risk of a strong dependence on a certain person, more formalisation of the exchange is targeted in the next future.
• In the “transparency position”, a central entity, typically the headquarters, but possibly also a site with a dominating position, steers information & knowledge sharing, either by being responsible for its creation and provision, or its collection, processing, and transfer. The sites’ access is not restricted, leading to high transparency. This position can be aspired to ignite competition, especially when revealing performance-related information, such as for internal benchmarking. When it comes to knowledge, it also constitutes a fair cooperation by reducing information asymmetry (Luo, 2005). Generally, data quality should be high for this position to assure the comparability and increase the acceptance of performance data or to avoid a multiplication of misleading knowledge. Ex. 4.3: The establishment of the internal performance benchmarking forced the Edgeband NW to shift the operational site performance (7) from the isolation position. A central coordination of the data provision, a set of jointly agreed KPIs among the site managers, and central pressure by setting incentives on the sites’ performance increased the information availability and moved operational site performance to a transparency position. As discussed previously, the benchmarking also marked a starting point for internal knowledge exchange. Tying the incentive level not only to a site’s performance but also to its promoted improvement potential, moved the manufacturing best practice exchange (13) towards the transparency position: Any suggestion for improvement is evaluated by an
108
DESIGNING THE NETWORK COORDINATION LAYER
expert group, stored in a central and freely accessible database and, depending on its potential, Exchange structure
I
I
External Info.
Internal Info.
Knowledge
K
7
Transparency
Limitation
7
1
Isolation 7
Networking
13
2
Operational site performance 13 Manufacturing best practice
K
Degree of transparency
= Solution
formation
competence
of
groups
or
trainings. The transparency the
13
13
? = Problem ! = Target
the
of performance data and
I
= Cause
spread in the network by
centralisation
formalisation
of
and best
practice exchange currently induce a movement towards the
networking
position,
strengthening ties between site managers and specialists and allowing them to challenge the best practice solutions on informal channels.
• In the “limitation position”, the exchange structure is again centralised but with low transparency. Low transparency can either stem from strategic considerations, e.g., when denying or limiting the sites’ access to information or when allocating knowledge only selectively. Information and knowledge are considered as scare resources, and granting access is exploited as coordination measures. This accounts less for comparative performance information than for information of strategic value, such as aspired business targets or external information about customers and markets, or for critical know-how. Moreover, access can also be actively restricted to prevent a drain of sensitive data or knowledge, e.g., when operating in areas with low intellectual property rights’ protection, or to prevent the sites from information overload. In both cases, sites can be characterised by the “receiver factories” of Vereecke et al. (2006) since they depend on the favour of a parental entity. Yet, this position might also be occupied less intentionally in case of insufficient or mismatching exchange mechanisms that are structurally restricting any sharing. Ex. 4.4: How the linkage between type of knowledge and exchange mechanism can affect the position in the framework could be observed in the Excitation NW. Here, the quality and speed of the value creation process are heavily relying on the customer-specific engineering process at the beginning of an order. Engineers negotiate with the customer the specifications of the product. The more complete and detailed the definition of the specifications, the shorter the lead time will be. Therefore, a project was initiated targeting a codification of these engineering best practices (13) and their translation into a central expert system – which would have meant a strong competitive advantage. However, the tacitness and variance of the engineering knowledge, which heavily relies on the expertise of the local
DESIGNING THE NETWORK COORDINATION LAYER
Exchange structure
I
I
External Info.
Internal Info.
K
109
Knowledge
Transparency
Limitation
!
13
?
Isolation
13 Manufacturing best practice (engineering)
13
I
? = Problem ! = Target
K
= Solution
was underestimated. So far, the process of moving the a transparency position got
13
= Cause
factors and idiosyncrasies,
engineering knowledge into
Networking
Degree of transparency
engineers about contextual
stuck
in
the
limitation
position, hence failing to achieve a higher degree of standardisation
for
the
engineering process.
The Logic Behind Ideally, all information and knowledge promoting the network’s objectives should be accessible by the sites requiring it. However, this often remains illusory (Tsai, 2001). In fact, managers face the challenge of asymmetric information availability in the network, underlined by individual and potentially conflicting targets between the site (and network) managers. To structurally support any exchange, decentralised or centralised approaches can be suitable. Decentralised approaches make it hard to assign responsibilities to foster exchange, and they rely strongly on the sites' motivation and offered incentives to participate in sharing. At the same time, however, they can create a form of "network thinking" based on informal relations. For centralised approaches, responsibility to foster the sharing is clearly allocated, which induces hierarchical pressure, but it requires such function to get deeply involved into the plants’ operative business to be up-to-date with regards to knowledge creation (Chew et al., 1990). Among the case networks, especially the younger ones and those evolving from the isolation position initially assigned this responsibility to a meditative and integrative network manager in order to build up lateral and informal ties. There was, however, a general agreement on the necessity to institutionalise more formal exchange structures and mechanisms in the near future to make personal relationships less dominant. Selecting the right exchange mechanisms, in turn, fundamentally drives the willingness and finally the intensity of sharing. Choosing the appropriate tool can foster sharing while the wrong one can hamper it. Therefore, network managers should consider various factors including the type of knowledge and information to be shared, cost considerations for detection, storage and transfer (which can be substantial especially when converting tacit into explicit knowledge), and also cultural aspects.
110
DESIGNING THE NETWORK COORDINATION LAYER
Ex. 4.5: In the Floor Care NW, establishing lateral ties with the Chinese plant and fostering the intercultural cooperation between China and Europe was supported by an iterative choice and development of appropriate exchange mechanisms. Since conventional mechanisms failed at the very beginning primary due to cultural differences, weak ties were established in a first step by regular video-conference meetings displacing more anonymous communication channels, such as e-mail and telephone calls. In a second step, temporary employee exchange was initialised to create mutual understanding, not only at the management level but also on functional layers. Today, the conventional channels work well across the two plants as employees draw on a closer relationship based on common projects, discussions, and meetings.
Again, the information & knowledge flows are also linked to the network configuration. First, the availability of knowledge in the network is highly influenced by the competencies of the single sites and their strategic reason, i.e., the plant role. The more distinct the competencies of a site, the higher the expectations regarding its contribution to knowledge creation and sharing should be. A lead factory, for example, is expected to play a central role in the “knowledge network”, especially for manufacturing processes and technology, while less dominating plants are primary in a receiver position benefitting from an coordinated access. Following Ernst and Kim’s (2002) idea of a network flagship that mediates and supports knowledge diffusion in an external supplier network, the lead factory in an internal network can also influence the exchange structure when lending itself to a central information and knowledge hub besides the headquarters. The bandwidth and level of competencies of the sites also affect their ability to absorb information and knowledge and to benefit from the access: The higher a site’s absorptive capacity and learning ability, the faster and more effectively knowledge can be exploited and transformed into better performance (Tsai, 2001). Intuitively, this also depends on the type of knowledge and the network structure. Market area or general purpose plants with high overlap in products and processes can better utilise best practice sharing in process and technology than process or product plants. On the other hand, such structurally different sites might be less reluctant to sharing in other categories, e.g., when it comes to customers, competitors, or best practices in business processes. Further, the interdependence between process plants linked by material flows might per se constitute a higher degree of information sharing, e.g., for planning data, inventory, etc.
DESIGNING THE NETWORK COORDINATION LAYER
111
Ex. 4.6: How a lead factory can affect the exchange structure is seen at the Seals NW. Moving the insular sites from an isolated position and creating a joint “network thinking” is supported by centralising the knowledge and information flows. But centralisation in this case is divided between Exchange structure
I
External Info.
I
Internal Info.
K
4 5 6
Knowledge 10
Transparency
Limitation 9
5
7
Centralised via Isolation headquarters
4
13
12
Centralised via lead factory Networking
11
7
9 10 11 12 13
I
6
K
Degree of transparency
= Cause
? = Problem ! = Target
= Solution
Site strategy & roles Financial site perf. Market & sales performance Operational site performance Admin. manuf. data Product innovation Product changes / improvements Techn. / process innovations Manufacturing best practice
two
entities:
headquarters
the
and
a
designated lead factory. While any information sharing (5, 7, and 9) shall be coordinated by the
central
function,
network knowledge
creation, collection, and diffusion
shall
be
coordinated by a global lead factory. For this, any responsibility to facilitate transparency for product changes / improvements (11), technology / process innovations (12), as well as for the organisation and support of manufacturing best practice exchange (13) is allocated to this most dominating site in the network; a step which is reflected in the framework by a shift in the current exchange structure from decentralised to centrally coordinated. Finally, this development is underlined by the nomination of the so far only implicitly assumed but not officially communicated role of a lead factory (4).
Discussion of Findings & Implications Information & knowledge sharing can be a prerequisite for other coordination mechanisms, e.g., when revealing performance data to underpin a benchmarking or when sharing inventory data to enable load levelling. As such, especially information transparency is the backbone for Martinez and Jarillo’s (1989) formal output / performance control mechanisms. But the design of knowledge and information flows can become a coordination measure itself. It can be part of formalisation and standardisation when promoting formal and central exchange structures and mechanisms. It might support planning and harmonisation by increasing the data quality for planning-related information and creating formal reporting structures. It can also promote lateral relations through formal platforms for knowledge sharing or informal communication by decentralised exchange structures fostering personal “networks”. Thus, it underlines a socialisation process, creating an organisational culture by sharing and teaching a common knowledge base, procedures, and values. Again, findings are summarised in Tab. 17.
Information & knowledge sharing framework
The framework touches six of the basic coordination mechanisms as introduced by Martinez and Jarillo (1989): - Formalisation and standardisation, by promoting formal and central exchange structures and mechanisms for codified information and knowledge (e.g., manuals, systems, data bases). - Planning and harmonisation, by increasing the data quality for planning-related information and creating formal reporting structures. - Output and personal control mechanisms, by increasing the data quality for output- / performance-related information and creating formal reporting structures. - Support for the building of lateral relations through formal platforms for knowledge sharing to establish lateral ties (e.g., custom. projects, trainings, moving people, competence groups). - Informal communication, by promoting informal and decentralised exchange structures fostering personal "networks". - Socialisation, by promoting a common knowledge base, procedures, and norms via appropriate mechanisms (e.g., trainings, moving people, customised projects) or by facilitating the exchange of mgt. experience and practice.
(1) … network coordination in specific The design of the information & knowledge flows in the network can be modelled as a function of (1) the exchange structure and (2) transparency of (3) distinct information and knowledge categories linked with appropriate (4) exchange mechanisms and affected by (5) the information availability or the intensity of sharing.
Findings and implications for … (2) … the design of a network architecture The design of the information & knowledge flows is dependent of the network structure, which can affect the readiness for and the potential of the exchange. It is also influenced by the network specification since plant roles can affect the plants position in the knowledge and information network, the exchange structure in general, and the sites' absorptive capabilities. Additionally, the organisational structure of the network defines the formal reporting and supervision channels, hence, particularly affecting the information sharing.
(3) … the network management in general Fifteen information and knowledge categories were isolated to analyse and design the networks' information and knowledge flows; linked with distinct exchange mechanisms. They comprise: (1) external information about markets / customers, competitors, and suppliers, (2) internal information about strategic, financial, and operational perf. and planning data, and (3) knowledge about products, technology / processes and mgt. experience; each is separated into innovations and improvements. The completeness and terminology of the categories might vary between companies, making an adaptation and complementation permissible.
112 DESIGNING THE NETWORK COORDINATION LAYER
Tab. 17: Findings on the information & knowledge sharing framework
DESIGNING THE NETWORK COORDINATION LAYER
113
3.4 Summary & Discussion 3.4.1 Findings from the Design of the Coordination Layer This chapter has isolated the main decision dimensions of the coordination layer. These comprise: (1) centralisation & standardisation, (2) resource allocation & sharing, (3) the incentive system, as well as (4) information & knowledge sharing. Literature-driven and in close discussion with practice, these dimensions were elaborated and transformed into distinct management frameworks. Tab. 18 summarises the elements of these frameworks with reference to the corresponding literature; input from the case networks is not explicitly outlined. Coord. decision dimensions
Variables
Centralisations & standardisation
Responsibility areas & categories
Degree of centralisation Degree of standardisation
Resource allocation & sharing
Incentive system
Information & knowledge sharing
Resource categories
Characteristics systems (primary or support activities); decisions (organisation, products, processes & technology, capacity); processes (strategic or operational) central unit - each site individually systems (indiv. tools / heterogeneous implement. standard. tools / homogeneous implement.); processes & decisions (no / local standardisation standardised (IT-) tools or methods) functional (R&D, engineering, scm, basic & special manufacturing, support functions); type of capacity (structural or personnel / specialists) limited - sufficient
Selected Authors (alphabetical order)* Chri s todoul ou et a l . (2007), Fel dma nn & Ol ha ger (2009) & (2011), Ha yes et a l . (2005), Ma ri ta n et a l . (2004), Rudberg & Wes t (2008), Vereecke et a l . (2006) Fel dma nn & Ol ha ger (2009), Gupta & Govi nda ra ja n (1991), Ha yes et a l . (2005), Ma ri ta n et a l . (2004) Ma ri ta n et a l . (2004), Mei jboom & Vos (1997), Rudberg & Wes t (2008)
Fel dma nn & Ol ha ger (2008), Luo (2005), Ts a i & Ghos ha l (1998)
Scarcity in the network Intensity of sharing / no exchange / sharing exchange frequently and to a large extent Extent of provision proportion of possessing sites vs. requiring sites (>, =, <) Performance outcome-dependent performance (financial, market & categories sales, operational); behaviour-dependent performance (contribution to learning, conformance with strategic goals) Reward types monetary (financial payments); non-monetary (reputation & awards, autonomy & responsibility) Organisational level agreed individually for each site of target agreement agreed above network level
Coe et a l . (2008), Ghos ha l & Ba rtl ett (1988), Ja ehne et a l . (2009), Luo (2005)
Allocation mechanisms Information & knowledge categories
Ba rtol & Sri va s ta va (2002), Henders on (2005)
Exchange mechanisms
Exchange structure Transparency
based on individual achievement / contribution to equal parts external inform. (market, competitors, supplier); internal inform. (strategic, financial, operational); knowledge (products, processes & technology, management experience, business & support processes) information (formal or informal) knowledge (customised projects / project support, training & qualification, manuals, systems & databases, competence groups, moving people & job rotation) decentralised - centrally provided information (no access - access to full data / info.); knowledge (no access - access for all sites) low - high
Luo (2005), Ja ehne et a l . (2009)
Chew et a l . (1990), Gupta & Govi nda ra ja n (1991), Luo (2005)
Ba rtol & Sri va s ta va (2002), Chew et a l . (1990), Luo (2005) Ba rtol & Sri va s ta va (2002), Chew et a l . (1990), Gupta & Govi nda ra ja n (1991), Sa l ter (1973)
Cheng et a l . (2008), Ghos ha l & Ba rtl ett (1988), Gupta & Govi nda ra ja n (1991), Luo (2005), Ts a i & Ghos ha l (1998), Vereecke et a l . (2006) Ba rtol & Sri va s ta va (2002), Erns t & Ki m (2001), Erns t & Ki m (2002), Ferdows (2006), Ts a i (2002) Chew et a l . (1990) Gupta & Govi nda ra ja n (1991), Luo (2005)
Availability of information Gos ha l & Ba rtl ett (1988), Gupta & Intensity of low - high Govi nda ra ja n (1991), Vereecke et a l . (2006) knowledge sharing * Authors addressing or operationalising the variables. Additional input was gained from the discussion with the case networks.
Tab. 18: Operationalisation of the coordination decision dimensions
114
DESIGNING THE NETWORK COORDINATION LAYER
Applicability and feasibility of the frameworks were demonstrated, using them in practical settings to map the AS-IS state, outline central weaknesses, and to discuss the design of future TO-BE mechanisms for the coordination of several case networks. Moreover, the differences in these case networks regarding global dispersion, size, configuration, industry type, products, and technologies account for generalisability of the frameworks. As indicator for the completeness of the decision dimensions and the accuracy of the logic behind the frameworks, a match with the generic coordination mechanisms derived from the organisation theory’s perspective in Section 3.1.1 is expected. Tab. 19 outlines a comparison with Martinez and Jarillo’s (1989) eight basic coordination mechanisms in multinational corporations. These were chosen since they are considered the most detailed and complete collection among the literature contributions reviewed. Although the comparison is reduced to the direct and most obvious linkages only, literally all the generic coordination mechanisms are touched. The frameworks themselves embrace the central coordination levers from a superordinate network management perspective and provide operational feasibility for this layer. They focus on supporting the network manager in making core decisions, thus underlining a conceptual strategic network design and management approach. Concrete measures and actions have to be derived and initiated subsequently. Nonetheless, recalling contingency theory as anchor for the research process makes some remarks necessary. First, the decision space reflected by a single framework can be understood as system itself with interdependencies between its underlying variables, hence requiring internal FIT in terms of a contingent set-up of the variables already on this level. The Profile NW’s progress towards decentralised but globally standardised manufacturing processes serves as example. Process standardisation was hampered by a mismatching degree of centralisation for the HR system; its assignment to regional authority caused variances in the training and diluted process discipline of the operators. Second, the distinct decision dimensions can be interrelated, meaning that the position in one framework has to be aligned with the position in another. Thereby, the reflected coordination mechanisms can be complementary, substitutional, contrary, or independent. Hence, internal FIT of the coordination layer in general has to be sought for, too. The example of a global performance benchmarking requiring changes in the incentive system, but also regarding the quality, exchange structure, and transparency of operational performance information shows this. Third, it has become obvious that the coordination decision dimensions are affected by contextual factors. Most importantly, evidence for the generally assumed contingency of the coordination layer on the network configuration has been looked at. From the coordination lens, this shifts the perspective to external FIT.
Coordination mech. (organisation theory)
Information & knowledge sharing
Incentive system
Resource allocation & sharing
Centralisation & standardisation
Network coordination decision dimension
Departmentalisation
(De)grouping of structural and / or personnel capacity (resources)
Defining the degree of standardisation for systems, for decision making, and processes execution
Centralisation or decentralisation
Promoting formal and central exchange structures and mechanisms for codified information and knowledge (e.g., manuals, systems, data bases)
Formalisation and standardisation
(De-)centralising responsibility for system development, decision making, and process execution
Planning & harmonisation of systems & processes Increasing the data quality for planning-related information and creating formal reporting structures
Defining the degree of the systems' implementation in the network
Incentivising behaviour-related performance and tying rewards to its fulfilment Increasing the data quality for output- / performance-related information and creating formal reporting structures
Incentivising outputrelated performance and tying rewards to its performance level
Output / performance and behaviour control
Martinez and Jarillo (1989) Lateral or crossdepartmental relations Creating formal platforms for knowledge sharing to establish lateral ties (e.g., customised projects, trainings, moving people, competence groups)
Fostering cooperation in the network, e.g., by setting targets above network level or by incentivising contribution to learning
Setting up specialists in integrative departments (e.g., task forces)
Informal communication Promoting informal and decentralised exchange structures fostering personal "networks"
Socialisation Facilitating the exchange of mgt. experience and practice
Promoting a common knowledge base, procedures, and norms using appropriate mechanisms (e.g., trainings, moving people, customised projects)
Teaching a competitive or cooperative corporate culture and shared objectives by rewarding the do's and restricting the don'ts
DESIGNING THE NETWORK COORDINATION LAYER 115
Tab. 19: Network coordination vs. coordination from an organisation theory’s perspective
116
DESIGNING THE NETWORK COORDINATION LAYER
Modelling all the direct and indirect linkages within and between the decision dimensions and with the contextual factors is illusory. Instead, the frameworks have to be considered as bricks for a discursive, comprehensive management support, guiding the analysis and conceptual development of the network. Doing so properly, however, requires a holistic perspective and makes the frameworks' integration into a network management architecture and approach necessary. 3.4.2 Implications for a Network Management Architecture With respect to the literature analysis, network configuration can be split into network structure and specialisation. In the discussion with the case networks, both of the decision dimensions of the configuration layer turned out to be influencing the design of the coordination mechanisms. First, the structure defines the geographic dispersion and physical linkages of the sites and sets the multiplant strategy which determines the sites’ general purpose and area of responsibility. The discussions revealed several examples for how this can affect the positions in the coordination frameworks. To recapture only some intuitive ones: Comparing a network with general purpose or market area plants to a network with process plants, first of all, this structural decision certainly impacts the resource allocation and sharing. While special manufacturing capacity, machines, and experts might be necessarily dedicated for process plants to perform their idiosyncratic manufacturing steps, pooling or cooperation – particularly of specialists – might be adequate for the other two configurations. Moreover, the decision about the multiplant strategy affects the comparability of the operations processes and their performance, which, in turn, can influence the readiness and potential of knowledge and best practice sharing and also the degree of competition or cooperation between sites in general. This has to be taken into account when designing the incentive system. Igniting competition by making operational performance transparent and embedding it into a formal site benchmarking could be useful for general purpose or market area plants with similar processes, but it is supposed to come to nothing for process plants. Similarly, network targets fostering knowledge exchange might be more valuable if the nature of operations is similar across sites. On the other hand, the willingness to share knowledge and information in other categories, e.g., for business processes, might be higher if sites are less comparable or competing. As to the network specialisation, several scholars have previously pointed out its closeness to coordination. While De Toni and Parussini (2010) consider the “management of facilities’ roles” (De Toni and Parussini, 2010, p. 7) as coordination objective itself, Christodoulou (2007) state that “… co-ordination choices must also fit
DESIGNING THE NETWORK COORDINATION LAYER
117
with decisions regarding plant roles” (Christodoulou et al., 2007, p. 27). Accordingly, various examples were found in the discussion with the case networks for this relation. To recall only some: The degree of responsibility and autonomy in the network is not independent of the site’s competencies. Especially very dominating sites with distinctive plant roles strive for more authority. But plant roles also reflect the management’s expectations towards a site’s contribution to the network; hence, they influence the design of the site’s incentives. Moreover, they might impact the exchange structures for information & knowledge sharing as in the case of a lead factory coordinating any process and technology improvements, but they also influence resource allocation and sharing when pooling process specialists at such site. Furthermore, the two configuration decision dimensions have to be complemented by the organisational structure of the manufacturing function. Defining the sites’ integration into the company’s organisation, it predetermines the organisational levels involved in decision making and the reporting channels for manufacturing activities, thereby influencing not only standardisation and centralisation decisions but also the formal information channels in the network. The organisational structure is somewhat “stuck in the middle” between the configuration and coordination layer. On the one hand, it is related to Martinez and Jarillo’s (1989) departmentalisation as coordination mechanisms, yet, on the other hand, the nature of the underlying decisions is clearly long-term-oriented, interpreting the network structure from an organisational perspective. Summing up, the network coordination layer is not only “somehow” contingent upon, but deeply linked with the network configuration. This makes isolated coordination decisions deceptive. Therefore, a combined approach is essential. So far, however, decision support by management frameworks has been restricted to the coordination layer only, limiting their usability and scope. Hence, to assure a comprehensive understanding of the network’s architecture, the three derived perspectives (decision dimensions) of the configuration layer have to be operationalised analogously.
118
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
4 From the Coordination Layer to a Management Architecture This chapter aims at linking the coordination dimensions with the configuration layer, thus complementing the network management architecture. Therefore, Section 4.1 is dedicated to operationalising the configuration decision dimensions and, thereafter, to transforming the initial heuristic research framework into the final management architecture. In Section 4.2, the applicability of the architecture is demonstrated by two comprehensive case studies. Section 4.3 finishes the chapter by discussing findings with respect to a strategic network design and management approach.
4.1 Introducing an Integral Architecture for Network Management 4.1.1 Elaborating the Network Configuration Layer Elaborating the configuration layer concentrates on the operationalisation of the network structure, the network specialisation, and the network organisation. Indicated by findings from the literature review, scholars have addressed these three dimensions to different extents. While there is little necessity to enhance and refine existing approaches and models with regards to network structure and organisation, network specialisation so far has only been tackled from the site level, thus lacking a network perspective on plant roles. As a consequence, the distinct dimensions require a different depth of attention in the following. Thereby, the discussion is based on the case of the Profile NW to exemplarily elucidate the linkages between the configuration decision dimensions and the coordination layer. Network Structure Close to what is commonly known as “manufacturing footprint”, the network structure comprises the geographic dispersion and physical linkages of the sites. In research, it has been described by generic network types and multiplant strategies. Both approaches have been proven to be theoretically and practically stable conceptual frames. Since they are basically derived from the internal supply chain connections between sites, the product and market allocation, and the product flows into the markets, more details on them can be obtained by getting deeper into the network’s internal and external material flows. There are several methods and software tools available for modelling material flows on different organisational levels, ranging from external supply chains, over intra-company networks, to single production sites and
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
119
even to the shop floor. However, with the primary focus on stimulating general strategic considerations, most of these tools tend to be somewhat over-engineered and detailed. As the aim of this work is not to draw an as accurate as possible picture of material flows but to derive a basic understanding of them, it seems appropriate to start off with a high degree of aggregation. From a practical point of view, this can be obtained by analysing the network’s main product groups or types (and their allocation to sites and markets) and by roughly sketching the internal and external volume streams between plants (semi-finished goods and intercompany business) and from plants to markets (third-party business). Ex. 5.1: The figure below outlines the aggregated projection of the Profile NW’s structure on the world map. It reflects the network’s geographic dispersion with the sites’ location, each site’s overall production volume for a given year, the product allocation for the main product types indicated by the volume mix per site, as well as the finished goods’ flows into the markets. Because manufacturing is based on a single-step process, no internal flows of semi-finished goods occur. Site 1
Site 3
Site 4
16 12
Site 6
30
26
11
Site 7
2
2 1
Site 5 Site 2
Volume mix (product types) Prime
Standard Basic
Site 8
Flow of finished goods (in tons p.a.) Intensity
x
Production site with x% of the network‘s overall production volume
The network encompasses three different product types: (1) prime products, which are of leading edge quality, satisfying mainly the demanding European customers, (2) standard products covering the mid-price segment as commodities, and (3) basic products allowing coping with price pressure in new markets. As depicted by the material flows, the multiplant strategy varies according to the product types. While prime products follow a product strategy and are concentrated to site 1 only, standard and basic products are supposed to follow a local for local strategy with market or region area plants.
120
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
But as shown, this allocation is not selective, especially for Europe and the market coverage in Russia, which is partially served by the local site 3 and by exports from Europe. Moreover, the analysis highlights some general issues. Although so far understood as global network, in fact 95% of the overall production volume is manufactured in Western and Eastern Europe and Russia, serving mainly these regions; sites 6 to 8 account for the remaining 5%. Hence, the strong position in the established and saturated European markets is contrasted with a very poor competitive position in North America and in the fast growing emerging markets; a situation that creates pressure on the company’s demanding business targets.
Network Specialisation The network specialisation aims at enhancing the physical network structure, taking into account the sites’ individual competencies and their strategic purpose. As argued, such understanding can be condensed by a site’s distinct role. But from a network manager’s point of view, instead of concentrating on single roles, the design of a socalled plant role portfolio is required; it aggregates and transforms the plant role concept from the site level to the superordinate network level. Scholars are highly aware that such an approach is needed, seeing its benefit in the development of “… a ‘language’ which makes it easier to describe the network and how plants contribute to its objectives” (Christodoulou et al., 2007, p. 25). But except for Christodoulou et al.’s (2007) “mountain model”, no useful solutions have been provided yet.
Fig. 27: “Mountain model” (adapted from Christodoulou et al. (2007))
The “mountain model” is based on four aspects to define plant roles; not all of them are explicitly detailed in Fig. 27: • A site’s position in the internal process stage / supply chain, e.g., whether it is a production facility, a pure assembly plant, etc.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
121
• A site’s geographic purpose, based on Ferdows’ (1997a) collection of strategic site reasons. • The activities performed by the site, basically reflecting the competence dimension in Ferdows’ (1997a) model. • The configuration or layout of the manufacturing processes performed by the site, i.e., the spectrum from low automated job shop production to highly automated production lines. Although conscious of the fact that “… not all (of the aspects) will be relevant in every case, but formal consideration will prevent issues being overlooked” (Christodoulou et al., 2007, p. 25), the model turned out to be somewhat “half-baked” with the four single aspects not fully differentiable – not from each other and not from the other configuration dimensions. To give examples: First, a site’s position in the internal supply chain is considered as physical decision, and as such it is part of the network structure rather than of the network specialisation. Similarly, this aspect is overlapping with the site activities. Taking an assembly process as example, it might be hard to differentiate whether it should be considered as process stage or activity. Further, although the configuration and layout of the processes performed influence a site’s competencies, they are of inferior importance compared to the strategic site reason and its range of competencies themselves; indeed, they are more a consequence of a plant’s role than a determining aspect. Finally, the “mountain model” still considers sites as isolated entities, neglecting their embeddedness in and contribution to the network. Using the “mountain model” and Ferdows’ (1997a) plant role approach as conceptual backbone, Fig. 28 promotes a modified management framework for the construction of the plant role portfolio. This framework basically transforms Ferdows’ (1997a) two axes model to a hexagonal game board. The first axis, “strategic site reason”, is reflected by the game board’s edges; their arrangement enable to distinguish between access to low-cost, access to skills and knowledge, and proximity to markets, as the most relevant reasons for establishing and running a site 44. A combination of two reasons, too, is possible in the respective transient areas. “Site competencies”, Ferdows’ (1997a) second axis, are represented by the concentrically layers of the portfolio; these indicate different site types. The evaluation of site types, in turn, is formalised by the so-called “site type matrix”, as shown in the upper right corner of Fig. 28. It builds upon the (1) bandwidth of competencies performed by a site and its (2) strategic importance for the network.
44
In fact, Ferdows identified two more strategic reasons, i.e., pre-emption of competition and the control and amortisation of technological assets, but he considers these as less important. Their lesser importance has also been empirically validated by Vereecke and Van Dierdonck (2002).
122
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
• The (1) bandwidth of competencies mirrors the “mountain model’s” understanding of “activities performed”. It also reflects Feldmann and Olhager’s (2009a) findings that competencies are allocated to sites as sequent process bundles. Thus, the more processes are performed by a site and the more specific these are, the higher a site’s bandwidth of competencies is assumed to be. In practice, a tailored refinement is necessary to identify a network’s idiosyncratic processes and to bring them into a logical order that represents an evolvement in its sites’ competencies. • The (2) strategic importance is assessed along the network’s dependence on this very site. It can be related to the transaction costs for phasing-out the site, including considerations about its production volume or product range, which makes it difficult to find an alternative source of supply, its singularity of processes performed or products delivered, or its contribution to the network. 45 Again, a network-specific adjustment of the factors is recommended.
Fig. 28: Plant role portfolio & “site type matrix”
When combining the two dimensions of the matrix, a site can be assigned to one of four site types; for the present termed: (1) basic, (2) critical, (3) leverage, and (4) 45
For the assessment procedure of a certain site, each of the chosen factors is separately evaluated on a scale. An overall value is calculated by multiplying the individual scores. Based on the overall value and / or the individual scores for the most critical factors, sites can be either ranked relatively, or numerical limits can be defined to position the site on the y-axis of the matrix.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
123
prime. Linking a site’s type with its dominant strategic reasons, it can be placed as “token” in the plant role portfolio; the combination of both dimensions constitutes its role. The size of the token’s outer ring symbolises the installed capacity at the site while the inner ring stands for the capacity utilised. Further, the processes actually performed by the site are outlined by the token’s coloured segments. Processes performed not only for the site itself but also for the network are additionally highlighted (by a star). Generally, not all possible combinations of site types and strategic reasons are feasibly or desired for a certain plant role; for instance, it would be misleading to nominate a prime site with low cost access as lead factory. Therefore, the plant role portfolio is complemented by the definition of target zones; these are highlighted areas that constitute a certain plant role. The target zone exemplarily defined in Fig. 28 prescribes that a lead factory plant role must be a leverage or prime site with access to skills and knowledge (besides other possible site reasons). Ex. 5.2: The design of the plant role portfolio for the Profile NW started with the evaluation of the sites along the introduced “site type matrix”. Eight different processes were selected to describe the bandwidth of competencies. These comprise four technology levels underlying the manufacturing processes: level 4 as the (not yet elaborated) low-cost technology level for basic products, level 3 as the standard technology level for commodity products, level 2 as the improved future technology for a Strategic importance of site Site 3
Site 4
III. Critical
Site 5
small
level for prime products. The scope
I. Prime Site 7
Site 8
and level 1 as the high technology
Site 1
Site 6
IV. Basic
cost-sensitive standard production,
II. Leverage wide
Tech. + Running-in + Logistic level 3/4* + Tech. standard + Running-in center level 2* + Tooling + Tech. * Not yet defined complex level 1
was Bandwidth of competencies performed
low
high
Production volume (p.a.) x Singularity of products / technology x Contribution to the network
Site 2
“Site type matrix”
enhanced
by
additional
processes such as logistics, the competence for tool construction, and the ability to perform the running-in phase for standard and complex tools respectively. The evolution in the site competencies is depicted by the x-axis of the matrix,
which brings the processes into a logical rank. While sites are always equipped with the technology level 3, with the running-in competence for standard tools, as well as with a logistics function, other processes are added successively and context-dependent. Moreover, the strategic importance of sites (y-axis) was assessed along their production volume, the singularity of their products manufactured and technology applied, as well as by their contribution to the network in terms of processes wherefrom other sites can benefit. The results of the evaluation are complemented by the sites’ primary strategic reasons and translated to the plant role portfolio below; enhanced by target zones for three different plant roles: (1) “lead factories, (2) “standard sites”, and (3) “start-up sites”.
124
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
The portfolio emphasises the multiplant strategies indicated by the evaluation of the material flows. Sites 3, 4, and 5 provide proximity to markets, following a local for local strategy with a reduced set of competencies. Site 2 is a low-cost site serving the Eastern European markets. Each of these sites reached a critical production volume, benefiting from economies of scale for standard products; hence, they serve as “standard sites”. High-end technology for prime products is solely tied to the dominating site 1. This plant assures access to the “German engineering know-how” acting, although not yet officially nominated, as “lead factory” that provides tooling and running-in skills to the network. In order to complement the strong presence in Europe, site 6 was established some time ago with limited manufacturing competencies to break into the North American market. However, it got stuck on its evolution from a “start-up site” to a “standard site” by not being able to reach a critical volume so far. One reason for this is the missing technology to compete in the mainly price-driven local markets outside Europe. Since the designated technology level 4 has not yet been brought to stability, today’s “start-up sites” are equipped with the standard technology level 3, causing them to inefficiently manufacture basic products in small lot sizes on the large scale technology for standard products. Site 7 and 8 are about to face a similar fate unless technology level 4 is launched; a fact that is also reflected by the analysis of the material flows. To cope with that, site 7 shall be empowered with its own tool construction competence to promote and support the development of the level 4 technology, thus acting as second “lead factory” for the company’s presence outside Europe; this intention justifies its transition phase as “leverage site type”.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
125
Network Organisation Finally, the network organisation defines the formal integration of the operations’ functions, i.e., the network and site management, into the company’s organisational structure. Thereby, it prescribes the supervision and reporting channels. It enhances the physical structure and the network specialisation by defining organisational ties, thus bridging configuration and coordination. Analogous to the overview in Fig. 29, literature is rich in providing ideal organisational types, often supplemented with a presentation of their historical evolvement and a discussion about advantages and drawbacks. 46 As shown in the figure, such typologies mostly start with a pure functional organisation, with sites tied directly to the management board or another functions (e.g., manufacturing or sales), over complex divisional or matrix structures, to mixed types and hybrids.
Fig. 29: History & principles of organisation structures (adapted from Diederichs et al. (2008))
Complementary to the organisational focus is the centre organisation; the sites themselves can be managed as cost or profit centres. While for the former, a predefined budget is granted, often in combination with the incentive to increase operational performance by a reduction in manufacturing costs, profit centres have individual profit and loss responsibility and compete with internal plants and external competitors. Again, hybrid forms are possible. As almost every network manager should be (and is) aware of the organisation chart of the operations function, a further elaboration on this configuration dimension is unnecessary for the purpose of this work. 46
See, for example, Slack and Lewis (2002), Miltenburg (2005), or Diederichs et al. (2008).
126
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
Ex. 5.3: The Profile NW represents a business unit (BU) embedded into a larger division of a multinational company. The manufacturing organisation can best be described by a matrix structure with product region mix. The company’s sites are attached as cost centres to the distinct regions they are located in and are supervised by regional operations. In parallel, there is a divisional operations …
Region 1 Operations region 1
Region n Operations region n
…
…
…
Site 1
…
Site 2 …
…
Divisional function 1
Divisional operations
…
Operations BU 1*
manufacturing networks; among those the Profile NW. The
Operations BU m*
…
BU = Business unit * Positions can be hold by identical persons
Manuf. scope
Manuf. scope
Site n
function coordinating the BU’s
Division …
Division 1
function is split into a team of technical
and
operational
specialists; team members are not exclusively assigned to one business unit. As depicted by the
centralisation & standardisation framework, when it comes to manufacturing-related decisions for the network or the sites, the corporate operations function formally overrules the regions’ authority. With the findings from the network structure and specialisation at hand, the organisational form is currently challenged. First, the regional matrix model seems by far too overloaded – at least for the S
System
D
Decision
P
23
20
Autonomous
Centralisation / Responsibility
Profile NW. Due to the dominance in
Process 24
Standardised
Each site indiv. Several sites
Europe, there is basically only one “real”
world
Centralised 7
22
Centralised &
Region
17 standardised 13 14 16 21 10
8
9 11 12
18 19
1
3
2 15
4
5
Degree of standardisation for the network
Division P D S
Hence,
the
current approach is prone to create bureaucracy
6
region. and
“organisational
waste”. For example, each region, and even the sites, are based on identical, internally copied structures. Recalling the regional differences in market size and in the sites’ production volume, this
approach causes costly duplications in administrative functions and (superficial) need for coordination. Condensing the findings with respect to the future design of the Profile NW, the following implications were derived: Generally, a separation into only two regions will be sufficient: “Europe”, including Central Europe, Eastern Europe, and Russia, as well as a “Rest of the World” region. The organisational structure in these regions has to be adapted to contextual requirements; especially manufacturing in Europe will be put under functional control. Each region will be technologically supported by a “lead factory”: the European region and the world-wide manufacturing of prime and standard products by the dominating site 1, the start-up sites for the rest of the world by the leverage site 7. Without doubt, these changes induce an alignment of the network coordination. Therefore, the coordination frameworks have to be passed and adapted iteratively.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
127
4.1.2 Transforming the Research Framework into a Management Architecture Taking the initial heuristic research framework from the literature review as reference point, the elaboration of the coordination and configuration layers reveals the necessity for some adaptations to transform it into a holistic network management architecture. First, the linkage between configuration and coordination has to be accentuated. Precisely, the coordination decision dimensions, i.e., centralisation & standardisation, resource allocation & sharing, the incentive system, and information & knowledge sharing, have to be anchored in the configuration layer. This layer, in turn, is split into the different configuration dimensions: the physical network structure, the organisational network structure (network organisation), and the network specialisation. The final architecture is outlined in Fig. 30.
Fig. 30: Network management architecture
For the purpose of practical usage mainly, the architecture is linked with and presented along the idea of the so-called PARTS analysis. Based on “Porter’s five forces” (Porter, 1979), PARTS is a qualitative strategic management frame developed by Brandenburger and Nalebuff (1995) to support the analysis of a company in its strategic environment (Friedli, 2000). It understands a company’s business activities as a game played by different actors, and the capital letters refer to the game’s five
128
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
elements: Players, Added value, Rules, Tactics, and Scope. Players describe the actors from the company’s point of view. These can be customers, suppliers, substitutors, and complementors; each of them is able to provide a distinct added value. Rules define the game’s general structure. They might arise from customs, law, contracts, or are set by regulating institutions. Tactics address the players’ behaviour: their individual targets, assumptions, and perceptions. The scope finally constitutes the borders and limitations of the game. With these five elements being applied to describe a company’s business environment, its boundaries, and actors, the PARTS analysis provides a holistic framework not only to understand a complex social system but also to discuss possibilities to change it in order to maximise the game's output by systematically shaping or altering its bricks. Tailoring the PARTS analysis to a manufacturing network requires a modification of its elements and a linkage to the management architecture’s single dimensions: PLAYERS – From a network manager’s focus, “players” are determined by the underlying plants. This raises the general question about which plants actually constitute the network, or vice versa, about which plants to encompass by the network architecture; an issue that emerges especially for large and less focused companies covering a range of different products, markets, etc. In such case, criteria for a segmentation of potentially more than one intra-company network become necessary. Getting an overview of the sites’ geographical dispersion, their product portfolio manufactured, processes performed, and their physical linkages and material flows can answer this question. Typically, a manufacturing network can be isolated – and the players assigned to – based on the similarity or interdependence of the products manufactured and / or processes performed and linked. In some cases, also the geographical dispersion might be an appropriate characteristic to differentiate between networks, for example, when restricting the focus to an organisation’s European activities if these are independent of other regions’ business. Hence, understanding the network structure is crucial for selecting the “players”. ADDED VALUE – As argued above, each site in the network has individual competencies. Combining these competencies with its strategic reasons defines its potential contribution to the network, i.e., the “added value” of this very player. On the site level, the added value is synthesised by a single plant role, on the network level, it was illustrated how plant roles can be aggregated to a plant role portfolio determining the network specialisation. RULES – To manage the players’ linkages and interactions, “rules” need to be established. Rules comprise both the decision about the network organisation, which determines the formal integration of the operations functions and prescribes the
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
129
supervision and reporting channels, as well as the design of the coordination mechanisms reflected by the single coordination dimensions. TACTICS – On a network level, tactics comprise the targets (or capabilities) aspired by the network manager to support the manufacturing strategy. On the site level, this reasoning corresponds to the site targets (capabilities). Site capabilities are similarly influenced by the manufacturing strategy, but they also have to fit with the network targets. Due to this work’s focus on the network level, the procedure of breaking down the manufacturing strategy into site capabilities, aligning the site capabilities with the network capabilities, and finally shaping the structural and infrastructural decision dimensions to define the plant roles will not be detailed. Moreover, the potentially hidden intentions of the players – which might match with, but could also contradict the formal targets – are neglected. According to this restriction, tactics address primary the aspired network capabilities. SCOPE – The “scope” finally sets the borders and limitations for the network management. It manifests itself in the manufacturing strategy, which influences the capability definition on network and site level. The manufacturing strategy has to align the existing or to be established capabilities with the internal and external contextual environment and the challenges the company is facing, such as mega trends, the business strategy, the customer and competitor structure, and also the boundaries and limitations stemming from the products, processes, and technology. Summing up, the derived architecture promotes a methodical anchor for a structured and methodical analysis and design of manufacturing networks. Its single elements reflect the central decision dimensions of the network’s configuration and coordination layer, surrounded by the strategic and contextual forces. Similar to the workbook approach promoted by Shi and Gregory (1998) and Shi (2003) and elaborated by the “Cambridge approach” of Christodoulou et al. (2007), both layers are detailed by management frameworks providing support to the network manager for strategic decision making on the distinct dimensions.
4.2 Working with the Network Management Architecture 4.2.1 Case Study Selection & Outline In the following, in-depth case studies will demonstrate the work with, the applicability, and the utility of the management architecture. Although the architecture has actually been applied in various industry projects, 47 the subsequent discussion is 47
Including the five networks for the development of the management frameworks, the architecture has actually been applied in ten industry projects so far.
130
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
restricted to two case studies only. Since the single management frameworks have already been validated in practical usage, solely the proof of validity for the whole architecture and its integration into a strategic network design and management approach are missing. Consequently, the additional contribution of more than two bipolar cases is considered marginally; hence, theoretical saturation is assumed, allowing to stop the research process (Eisenhardt, 1989). Again, the selection of the cases is grounded in theoretical sampling and not at random (Flick, 1999). To be more precise, both case networks were isolated from the industry survey sample with respect to their position in the network capability level and conformance evaluation. As depicted in Fig. 31, case 1, the Elevator NW, is positioned in category I, reflecting an (slightly) above average network capability level and a high conformance. Case 2, the Chocolate NW, is positioned in category IV, mirroring a (significantly) lower overall capability level and conformance compared with the survey average. Overall network capability level & conformance
Network capability conformance as deviation from mean
Overall network capability level as deviation from mean
Low capability level / High conformance
High capability level / High conformance
II
I
IV
III
Case 1: Elevator NW
Case 2: Chocolate NW
Low capability level / Low conformance
High capability level / Low conformance
Fig. 31: Selection of the networks for the in-depth case studies
The two positions constitute the bipolarity of the cases by contradicting challenges for the network management. The position of the Elevator NW indicates a well-defined and stable state, but contextual changes are questioning that position and force the management to induce some modifications on the network’s capabilities. These, in turn, call for a realignment of the configuration and coordination decision dimensions. Their conceptual redesign was supported by the network architecture. The Chocolate NW fundamentally lacked an initial “network thinking”, thus requiring a step by step approach to design the network’s TO-BE state from scratch; again based on the network architecture.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
131
The different situations are accounted for by the presentation of the cases. The Elevator NW case is described in detail, covering the AS-IS analysis and the design of the TO-BE network state as motivated by the strategic directives of the top management. The discussion is led along the single management frameworks and structured by the PARTS architecture. The analysis of the Chocolate NW was conducted similarly, but due to the less distinct “network thinking”, it turned out to be difficult to switch directly from the AS-IS to the TO-BE state without constituting a common understanding of the network’s aspired directions for evolvement before. Hence, for the presentation, the details of the AS-IS analysis will only be summarised shortly. Instead, the scope is concentrated on the creation and evaluation of potential TO-BE scenarios setting the direction for the future network development. Overall, the illustration of the cases follows a narrative style with the two networks / companies made anonymous. Although, to the best of the own knowledge, all substantial information is presented correctly, some of the key figures are modified slightly to protect the companies identity and competitive advantages. 4.2.2 The “Elevator NW” Case The Elevator NW is part of a leading global manufacturing and service Group. The competence spectrum ranges from production, installation, maintenance, and modernisation of “low”, “medium”, and “high segment” passenger elevators to freight solutions, escalators, and moving walks. The Group is headquartered in Western Europe and employs over 40’000 people in more than 100 countries. The annual sales volume regularly outrages €6 billion, generating an operating profit (EBIT) of more than €500 million. The service business accounts for the biggest share in profit, the contribution of manufacturing is of minor importance. The Group is running business in three world regions: Europe, Americas, and Asia. Manufacturing activities are divided accordingly and split into three largely independent regional supply chain organisations. The supply chain function holds the responsibility for the Group’s own manufacturing plants and the management of its broad 1st tier supplier base. 4.2.2.1 Network Analysis & Target Setting Players Striving for a reduction of complexity, it was decided at the very beginning to limit the focus of the case study to the Group’s independent elevator business and the European supply chain organisation. With about 50% of the new elevator installations in the European region and about €1 billion manufacturing-related turnover, the largely autonomous “SC-Europe” represents the Group’s biggest manufacturing organisation,
132
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
also running most of the own elevator manufacturing sites. The underlying network structure is sketched in Fig. 32. The European supplier base comprises a mix of four internal suppliers (SP), i.e., the own manufacturing sites SP. 3, SP. 4, SP. 7, and SP. 16, as well as about 60 external 1st and 2nd tier suppliers. Both parties embrace about 1000 employees working for “SC-Europe”. 1 Order center
Headquarters Commissioning
Purchasing order
~ 60 external 1st tier SPs. 49
SP. 4 SP. 16 SC-Europe (50% NI)
9
5
2 35
4 own manuf. sites (SP. 3, 4, 7, 16)
SP. 3
SP. 7
SC-Americas (24% NI)
x x
~60 external 1st tier SPs.
SC-Asia (26% NI)
Material flows Technology lines: 1) Sheet Metal 2) Mech. & Comp. 3) Electronics 4) Drives
Financial flows Customer order
Material flows
Customer
5 Consolidation hubs
Internal / external 1st tier supplier with x% of the overall network’s monetary purchase order volume p.a. Consolidation hub Order center
NI = New installations
SP(s) = 1st tier supplier(s)
Fig. 32: Elevator NW: Network structure
The upper right part of Fig. 32 outlines the logic of the customer to customer order cycle, which reveals insights into the network’s manufacturing model. Customer orders are placed at the headquarters first, where orders are configured using standardised and non-standardised subassemblies. Subassemblies are transformed into purchasing orders which are allocated by an order centre to the respective internal and external 1st tier suppliers. The relation between purchasing orders and suppliers is bijective; their selection, nomination, and qualification are part of any new product development process. This allows for an automated order placement once the links are defined. The group-owned manufacturing sites and the external 1st tier suppliers represent the same supply level. According to the subassembly types manufactured, suppliers can be grouped into four technology lines: (1) Sheet Metal, (2) Mechanics & Components, (3) Electronics, and (4) Drives. A supplier might cover more than one technology line, and also the product portfolio among the suppliers of one single line can be redundant. For Europe, the monetary volume split of the purchasing orders between internal and external suppliers is about 1:1 with the 21 biggest suppliers
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
133
(intern and extern) accounting for more than 95% of the total volume. On the subassembly level, the multiplant strategy is best described by regional plants with a partially overlapping and competing product portfolio. Closing the customer order cycle, the finished subassemblies are transported to one of five European logistics hubs (C-hubs), where the single material flows are consolidated and the orders forwarded to the customer’s construction site for installation. The optimisation of the internal material flows from the consolidation hubs to the customers was conducted separately, taking their location as given. From a marketing view, products / customer orders can be separated into a “low”, “medium”, and “high segment”. From a manufacturing perspective, this separation describes the degree of engineering activities, which, in turn, determines the production volume mix. While the “low segment” relies on commodity parts and a high volume / low variety production mix, the “high segment” requires customised solutions and a low volume / high variety mix; the “medium segment” is in between. In summary, the following considerations framed the identification of the manufacturing network’s players: • Based on the sites’ geographic dispersion, it was decided to build upon the Group’s prevailing structure, restricting the scope to the European supply chain. • Based on the Group’s product portfolio, it was decided to concentrate on the elevator business only (since there is little overlap with the escalator operations) but to cover all three product segments; from “low” to “mid” and “high”. • Again based on the network structure, it was decided to enhance the intercompany network by taking the 21 biggest 1st tier suppliers for the European region into account – independent of their ownership; 48 these are four internal manufacturing sites and 17 external suppliers. • Based on the players’ product portfolio, it was also decided to subdivide the supplier base according to the four different technology lines: Sheet Metal, Mech. & Comp., Electronics, and Drives, taking each as a separate but dependent unit of analysis. Tab. 20 summarises the characteristics of the Elevator NW.
48
At a first glance, including internal and external players is contradicting to this study’s focus on intracompany manufacturing networks as stated by the main research question. This restriction is somewhat relaxed by the fact that all players represent the same supply level. Although the ties to the own sites are closer, from the Group’s perspective, all suppliers are considered as individual profit centres competing for purchasing orders. Hence, they do not represent the traditional understanding of a supply chain with the suppliers conducting sequential manufacturing steps. Further, integrating the external supplier perspective will be understood as complementation of the initial scope but not as a fundamental shift.
134
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
Network scope
Company Division Business Other HQ Industry Core level level unit level products Elevator n.a. Mechanical Elevators NW*** & electrical products
Core processes 1) Sheet Metal 2) Mech. & Comp. 3) Electronics 4) Drives
# Sites (operat.) 4 int. / 17 ext.
Network characteristics
# Employ. (operat.) 1000 int. / 1000 ext.
Global dispersion Regional (Europe)
Network structure* Sequential or convergent
Network category** I
* According to Meyer and Jacob (2008) ** According to the survey classification of the network capability level and conformance *** Supply Chain Europe organisation
Tab. 20: Elevator NW: Network characteristics
Added Value The isolation of the players was complemented by the analysis of their “added value”. To elaborate the network specialisation, a distinct plant (or supplier) role portfolio was constructed for each of the four technology lines. Therefore, suppliers were evaluated on the dimensions of the introduced “site type matrix”. First, the bandwidth of competencies was defined based on the network’s manufacturing-related value creation process; for this, the main activities, i.e., (1) product design, (2) industrialisation, and (3) operations, were broken into level 1 processes. Since process competencies might vary according to the technology lines and the production volume mix, both dimensions were further used to detail the processes. Tab. 21 shows the process landscape. It comprises 14 processes per technology line and 44 in total. Processes Main activity Product design Industrialisation
Operations
Technology line
Volume mix
Level 1 process
x
x
x
x
x
x
x
x
Strategic SCM
x
x
x
x
Process design (processes & methods)
x
x
x
x
Production processes
x
x
x
x
Quality
x
x
x
x
x
x
x
x
Prototyping
Design to cost / manufacture Production design (layout, infra. & log.)
Service & maintenance concepts Engineering
Low vol. / high variety
High vol. / low variety
Tab. 21: Elevator NW: Process landscape
The level 1 processes were grouped and ranked hierarchically to reflect the path of a supplier’s evolution in competencies; the rank defines the x-axis of the “site type matrix” in Fig. 34. For each technology line, suppliers were assigned to either a small or a high bandwidth of competencies performed for the Elevator NW. The evaluation was part of a survey based assessment validated by a subsequent discussion with the network management team. Thereby, not just the bandwidth but also the suppliers’
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
135
actual process competence and efficiency were considered. Fig. 33 gives an example of a selected supplier profile condensing the results of the assessment.
Fig. 33: Elevator NW: Example of a supplier profile
136
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
The upper left part of the profile mirrors general information about the supplier, including the average delivery costs to the five consolidation hubs, the supplier’s primary strategic (site) reason, and the information used to evaluate its strategic importance, i.e., the y-axis of the “site type matrix” in Fig. 34. This evaluation is grounded in a supplier’s singularity of products, its share of purchasing order volume, and the number of product types provided; each criterion evaluated with respect to a single technology line. An overall score was calculated based on these criteria, allowing assigning the supplier to either a low or high strategic importance. Moreover, the bottom of the profile displays the supplier’s competence and efficiency for the 14 processes per technology line. The upper right part, in turn, highlights the supplier’s capability profile for the AS-IS and TO-BE state as rated by the network management and the supplier itself; below it, the supplier’s position in the “site type matrix” is indicated. The overall “site type matrix” for all suppliers and technology lines is shown in Fig. 34. Suppliers are numbered from 1 to 21. Some occur several times (e.g., SP. 3, SP. 9, or SP. 16), serving more than one technology line.
Fig. 34: Elevator NW: “Site type matrix” per technology line (AS-IS)
Combining the suppliers’ site types and strategic reasons led to a distinct plant role portfolio per technology line as highlighted in Fig. 35. The size of the players show their purchasing order volume in the respective technology line, ranked from low (1) to high (5); the coloured segments reflect the processes performed for the Elevator NW as indicated by the “site type matrix”. Further, it is marked whether a supplier is internal and group-owned or external.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
Fig. 35: Elevator NW: Plant role portfolio per technology line (AS-IS)
137
138
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
The derived plant role portfolios underpin the network specialisation: • Starting with Sheet Metal, the plant role portfolio reveals a broad and unfocused supplier base. Most of the external suppliers are leverage suppliers providing small- to mid-scale volumes for standard products only, but they perform almost the full bandwidth of processes for the Elevator NW – from production to product design. Even though the competencies needed for standard Sheet Metal manufacturing are rather basic, the big volume suppliers SP. 4, SP. 9, SP. 16, and SP. 18 do not allow benefiting from low cost access. Nonetheless, their proximity to the consolidation hubs allows controlling for the high transportation costs. Internally, particularly the prime supplier SP. 3 provides a high degree of competencies; it is the only site promoting low volume / high variety manufacturing of customised Sheet Metal solutions for the “high segment”. SP. 4 and SP. 16 serve the “low to mid segment” mainly. • For Mechanics & Components, the portfolio appears similarly unbalanced with a severe subset of basic small scale suppliers; but here, they perform production processes mainly. The picture reflects the large and variant number of different mechanical components so far supplied by scattered sources. But since most of the suppliers are neither unique in the products nor salient in their production volume, there is still potential for consolidation. Such considerations are underpinned by the sites’ limited low cost access – a fact carrying even more weight than for Sheet Metal products because of the Mechanics’ & Components’ lower transportation costs stemming from a higher packaging density. Internally, SP. 16 is the only group-owned manufacturer. As a provider of standard components, it competes with external suppliers, but as a provider of safety relevant mechanical devices, it is the network’s single source. • For Electronics, the portfolio is clearly dominated by the internal SP. 7 accounting for more than 95% of the overall purchasing order volume in the high volume / low variety business. Only for the “high segment”, selected customer-specific devices are supplied by SP. 3, justifying its wide level of competencies grounded in mastering the low volume / high variety processes. Again, the limited access to low cost sources is evident, especially for the large scale production at SP. 7. • For Drives, most of the components are sourced externally by two systems suppliers that cover more than 60% of the overall purchasing order volume. Nonetheless, most of the competencies are redundant and also available at the two internal sites. Comparing these, SP. 16 is a large scale source for high volume / low variety standard solutions while SP. 3 again provides engineered products mainly for the “high segment”; it also promotes the most promising
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
139
future technology for Drives. Historically, the high volume production of SP. 16 originated as copy of the Drives production at SP. 3. Rules In terms of the management architecture, “rules” are understood as the mechanisms driving the linkages and interactions between the single network players. They comprise the network organisation and the decision dimensions of the coordination layer underlined by the introduced frameworks. For the Elevator NW, the frameworks for centralisation & standardisation, resource allocation & sharing, as well as information & knowledge sharing were applied to support the network analysis and (re-)design methodically; the incentive system was omitted. The Elevator NW is coordinated by a functional manufacturing organisation with the four internal sites assigned to the European supply chain management; the sites themselves are run as profit centres. Besides heading the internal manufacturing activities, the supply chain function holds the responsibility for the management of the external 1st tier supplier base, for outbound logistics, and for purchasing. The function itself is tied to the superordinate Group management. Having the organisational entities specifying the network’s authority levels identified, the centralisation & standardisation framework was derived as outlined in Fig. 36. Responsibility areas
Each site individually
Centralisation / Responsibility
S
System
1a 1b 2b 3b 3c 10 18 23
24 a
11 b 24 24 d e
D
Decision P Process
24 f
Several sites
Standardised
24 c
6 7
Production system (layout) Production system (processes) Product data mgt. system (dev.) Product data mgt. system (prod.) Quality Mgt. (system & stand.) Quality Mgt. (tools & methods) Maintenance system Management system Improvement programs (besides production) HR system Know-how exchange system
Region Europe (SC-Europe)
Decision
8 17
19 22 19 21 22
9
Centralised & Standardised21
Centralised
Group level
Degree of standardisation for the network
1a 1b 2a 2b 3a 3b 3c 4 5
7
Autonomous 24 c
System
5 3a 3a 20
24 b
12 13 14 15 16 6
No / Local standardisation
Documented rules, guidelines & processes
Audited / Controlled processes & routines
Individual tools / heterogeneous implementation level at each site
Individual tools / homogeneous implementation level at each site
Standardised tools / heterogeneous implementation level in the network
Internal view External view
1-24 from external view excl. 3a, 19, 21, 22, and 24c
2a
4
11 a
Standardised (IT-) tools or methods Standardised tools / homogeneous implementation level in the network
P D S
8 Site strategy & roles 9 Organisational structure 10 Manufacturing IT decisions 11a Make-or-buy (new product) 11b Make-or-buy (existing product) 12 Product allocation decisions 13 Transfer pricing 14 Production process decisions 15 Manufacturing techn. decisions 16 Long-term capa. development 17 Short-term capacity adjust.
Process
18 Strategic sourcing (2nd tier) 19 Strategic logistics 20 Product cost calculation 21 Long-term S&OP 22 Intern. SC-planning / Order alloc. 23 Short-term manuf. planning 24a Prototyping 24b Design for cost / manuf. 24c Custom design engineering 24d Process design 24e Production design 24f Manufacturing
Fig. 36: Elevator NW: Centralisation & standardisation framework (AS-IS)
140
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
The framework differentiates between an internal view, limiting the system’s scope to the group-owned players (bubbles), and an external view, widening it by integrating the external suppliers (squares). Internally, the analysis reveals a two-sided picture: Manufacturing-related areas are predominantly decentralised and conduced autonomously with little network-wide formalisation and standardisation. Support systems, strategic decisions, and processes, however, are standardised with the central responsibility hold by either the European supply chain function or the Group level. More detail can be obtained when getting deeper into the distinct responsibility areas and categories as introduced in Section 3.3.1: • Systems for primary activities, such as the production system (1a, 1b) and tools & methods for quality and maintenance (3b, 3c), are designed and maintained autonomously. They mirror the sites’ broad authority on manufacturing operations, lacking any network-wide standardisation. The same accounts for the management of production relevant data. Although product data (2a) is created centrally as part of the development process, this process is conducted independently of the manufacturing function, in turn, forcing each site to create its own production master data (2b). Nonetheless, quality standards and guidelines for development, manufacturing, and supply chain (3a) are defined centrally, aiming at a network-wide harmonisation. • Systems for support activities, including the management and KPI system (4), the human resource system (6), as well as improvement programs besides production (5), are designed and initiated centrally, seeking for global harmonisation. The local management of the know-how system (7), on the other hand, indicates the site’s lack of motivation and the missing parental pressure for best practice sharing and exchange. • Organisational decisions are made on a regional level, except for manufacturing-related IT decisions (10) comprising the selection of information technology for the design and execution of production processes. For these, the decentralisation of authority in the past led to a scattered and isolated IT landscape, which, today, makes any production relevant data on manufacturing procedures and machine times hard to standardise and share. • Product-related decisions, such as make-or-buy considerations for new products (11a), are highly standardised and defined on Group level. This level also drives the actual product allocation (12) and the corresponding selection of 1st tier suppliers (19); both are part of any new product development process. • Analogously, strategic decision making about process and technology choice and allocation (14, 15) and about the long-term strategic capacity development
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
141
(16) require the Group’s approval for investments; operational decisions like the short-term capacity adjustment (17) are under regional authority. • Similarly, strategic processes are mostly centralised and standardised. The selection and management of third-party logistics providers (19) connecting the suppliers with the consolidation hubs, for instance, require a centralised coordination. The same accounts for sales & operations planning (21). • On the contrary, most of the operative processes introduced by the process landscape are executed autonomously with little standardisation in the network, ranging from prototyping (24a), process and production design (24d, 24e), over pure manufacturing (24f), to the management of the 2nd tier suppliers (18). Custom design engineering (24c) is restricted to these sites engaged in the “mid to high segment”. Moreover, the design for cost / manufacturing (24b), i.e., the involvement of the manufacturing function in the new product design process, has recently been centralised and partially formalised, making use of manufacturing sites to support a cost-sensitive product development. From the external view, most responsibility areas are decentralised and autonomously conducted by each supplier. Besides exceptions, only general quality standards (3a) are prescriptive for all network players, the management of logistics providers (19) and the order allocation (22) are under the region’s control, and long-term sales & operations planning data (21) is passed to external suppliers. The findings obtained can be ascribed to the Group’s decision to run the own manufacturing sites as independent profit centres competing both internally and with external suppliers; such decision fosters local optimisation. For the Elevator NW, the decision is realised by a high degree of site autonomy for manufacturing operations contrasted with strong parental control for responsibility areas of strategic importance. The attitude towards local optimisation is also underlined by the resource allocation and sharing. The framework in Fig. 37 again combines the internal and external view, but here, it also distinguishes between the four technology lines (by colour coding). The overall view outlines little effort on departmentalisation and limited possibilities for lateral relationships; instead, a focus on dedicated resources and competition prevails. Internally, R&D capacity (1) is centralised at the headquarters and separated from the supply chain organisation. Access is granted to all internal sites but mainly restricted to design support since product development is typically not pushed by the manufacturing function. Except for standard manufacturing equipment and machines (4.2), all other resources are kept short with little sharing. The situation ignites competition for the initial resource allocation but creates pressure on the resource availability, too. Only the two internal electronic manufacturers (Sp. 3 and Sp. 7) have
142
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
recently established some exchange of custom design engineers (2c) and production design specialists (5.2c). Both resources are represented by highly qualified engineers who support either the “mid” and “high segment” customers in the design of tailored solutions or are responsible for the development and optimisation of manufacturing steps and layouts. For the custom design engineers, a certain degree of local dedication is necessary in order to meet the customer requirements on product range and design flexibility. In contrast, production design specialists (5.2), but also the process design specialists (5.1) carrying out the industrialisation of new products, could probably be shared; particularly between sites serving similar technology lines. The same accounts for the local supply chain specialists (3) managing the 2nd tier suppliers. These resources offer potential for departmentalisation and cooperation. Externally, except for selective joint R&D projects, resources are assigned locally with no cooperation. Scarcity in the network Limited amount
1
Rather limited amount
2a 2b 2d Competition 2,3,5
Balanced amount
4.1 4.1 a b 4.1 4.1 4.1 c d
Rather sufficient amount
4.2 4.2 a b 4.2 4.2 4.2 c d
Sufficient amount
5.1 a
5.1 b
5.2 a
5.1 c
5.2 b
5.2 d
5.1 d
3
1
Cooperation
2c 5.2 c
Dedicated Resources
Resource Pool Intensity of sharing / exchange Frequently and to
No exchange / Seldom and to sharing a small extent Extent of provision as proportion of possessing Resource categories: sites vs. requiring sites: 1 R&D capacity Possessing sites > requiring sites Possessing sites = requiring sites Possessing sites < requiring sites Internal view
External view
2 3 4
a
a large extent
5 Special manufacturing capacity Custom design engineers (CDE) 5.1 Process design specialists (PE) Strategic supply chain specialists (SCM) 5.2 Production design specialists (PS) Basic manufacturing capacity 4.1 Manufacturing capacity 4.2 Standard manuf. equipment & machines
Sheet Metal
b
Mech. & Comp.
c
Electronics d Drives
All tech. lines
Fig. 37: Elevator NW: Resource allocation & sharing framework (AS-IS)
The information & knowledge sharing framework completes the discussion on the rules. It again pursues the general competition strategy. As highlighted by the framework in Fig. 38, the design of the “information network” is targeting output and behavioural control by relying on centrally coordinated and formalised structures. Only selected information is made transparent to the plants if considered supportive for competition, such as the financial and operational site performance. Correspondingly, plants in the “knowledge network” act mainly as isolated players not being in regular touch for innovation and best practice exchange.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
143 Information & knowledge categories
I
Exchange structure
External Info.
Centrally provided
I
8
1
Internal Info.
9
Knowledge
External information
6
Internal information
2
Transparency 11
Limitation Centrally coordinated
K
10
8
14
11
3
Isolation 12
Decentralised
13b
Medium Low
13a
Networking
4
15
No access
Access to limited data / info.
Access to most data / info.
Access to full data / info.
No access
Access limited to selected sites
Access for all requiring sites
Access for all sites
I: Availability of information K: Intensity of sharing High
Markets / Customers Competitors Suppliers
4 5 6 7 8 9
Site strategy & roles Financial site performance Market & sales performance Operational site performance Sales & operations planning Manufacturing planning data
Knowledge
Centralised & decentralised
No exchange
7
5
1 2 3
Internal view External view
Exchange mechanisms information Informal channels, e.g.: • ad hoc telephone calls & e-mails, meetings • social activities
I
10 Product innovations 11 Product changes / improvements 12 Technology / process innovations 13a Manufacturing best practice 13b Engineering best practice 14 Management experience & practice 15 Business & supp. proc. improv.
I Degree of transparency K
Exchange mechanisms knowledge
Formal channels, e.g.: • databases & sharepoints, intranet • regular & formal meetings
Moving people / job rotation
Customised projects / project support
K Competence Training & groups qualification Manuals, systems, databases Mechanism not used
Fig. 38: Elevator NW: Information & knowledge sharing framework (AS-IS)
A detailed look at the single information & knowledge categories complements the picture. It starts with the focus on information sharing: • External information about markets, customer, or competitors (1, 2) is raised centrally by the Group’s market intelligence functions. Transparency is limited, feeding the own sites with only selected data on a monthly basis. 2nd tier supplier data (3), however, is shared frequently and openly. It is based on formal monthly exchange calls and informal lateral ties, both constituting the networking position in the framework. • Any discussion on site strategies and roles (4) was oppressed for a long time, and formal plants roles were not assigned. This isolation is currently being relaxed by establishing elementary informal communication. • In line with the competition strategy, financial performance data (5) is coordinated centrally but made fully transparent to the network. The same accounts for information on operational site performance (7); however, due to a missing coordinative instance, operational performance has not yet been embedded into a formal benchmarking process for internal learning and best practice sharing.
144
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
• Market & sales data (6) is not generated by the single manufacturing sites; these are not actively competing for market share. Instead, limited data is again formally raised and distributed by the Group’s function. • The sales & operations planning process (8) is highly standardised and iterative. It contains the provision of order forecasts to the sites which are broken into an operations forecast and reported back. The combination of less monthly demand variations and a quite stable planning forecast leads to a sufficiently high information availability; nonetheless, access to other sites’ data is limited. This process is among the only interfaces to external suppliers when forwarding selected sales & operations planning data as input for their operations planning. • Administrative manufacturing data (9), comprising production controlling data, inventory levels, absenteeism rate, etc., are reported monthly to the headquarters, but again, access is restricted. The knowledge sharing and exchange reinforces the impression of insular sites: • Product innovations (10) are basically provided by the central R&D function to selected sites. However, since internal sites are getting closer involved into the new product development process to assure manufacturing suitability and cost sensitivity, some bilateral cooperation between R&D and manufacturing is emergent. External suppliers are excluded from the process with solely changes in product parameters being communicated to them. • Product changes and improvements (11) are primary related with bug fixes. They are updated regularly, informing all requiring internal and external sites. • Long time there has been no sharing at all for innovation and best practices in processes and technology (12), manufacturing (13a), engineering (13b), and in business and support processes (15). Currently, with the recent organisation of quarterly functional operations meetings, the sites’ mindset is slightly changing, but it is still far away from a networking position. Only management experience and practice (14) is shared more frequently, facilitating socialisation on the management level; not least because of the Group’s promotion of a common training and education platform. Tactics & Scope The main manufacturing-related challenges the Group’s elevator business is confronted with in Europe can be captured by two headings: (1) a rising cost competition and (2) the need to provide a high external product variety while mastering internal process complexity. Product variety is driven by the network’s presence in three market segments and intensified by facing a broad, dispersed
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
145
customer base. So far, the scattered customer demand is met by offering a high degree of local product adaptation; this causes the need for a wide and predefined product range in the high volume business and for customised solutions in the low volume business. Since manufacturing authority is assigned to the single sites with little standardisation, product variety directly impacts the internal process complexity in the network; a problem additionally amplified by the long product life cycle requiring a ten years guarantee for spare part provision. As to the cost competition, the Elevator NW faces an increasing pressure particularly in the “low and mid segment” which account for more than 95% of the new installations. Newer market analyses revealed the price as most important selling argument in more than 80% of the customer orders; a situation that is even exacerbated by a rising price fight in the saturated European markets. There, the expected growth rate of the construction industry is marginally and the competitor structure consolidated and balanced; the five biggest players account for about 85% market share. Moreover, the competitors’ provision of overlapping product portfolios, offering a broad bandwidth of alternative options, has sharpened the customers’ expectations and led to a downgrading of the traditional strategic priorities (product range, quality, speed, and delivery dependability) to order qualifiers. At the same time, since orders are negotiated with prime contractors responsible for new installations but not for the subsequent operations, innovations concerning the product lifecycle performance are hard to convert into a competitive advantage. Hence, with the price becoming the most important order winner, the Group is forced to adjust its European manufacturing strategy. As a reaction, a cost leadership program has recently been launched in order to strengthen the competitive position in the elevator business; a position that shall be solidified by retaining order size and delivery flexibility as the Group’s traditional order winner. The changes in the manufacturing strategy require an alignment of the Elevator NW’s capabilities. Based on the selection from Shi and Gregory (1998) and Miltenburg (2009), Fig. 39 outlines the aspired capability profile summarising the network management’s tactics. On the one hand, the capability targets strive for cost superiority, expecting the network to better exploit economies of scale and scope and to realise a reduction of duplications in administrative areas. Analogous, access to markets / consolidations hubs has to be retained to control delivery costs. On the other hand, a consolidation and exploitation of the scattered knowledge base is targeted to foster internal learning and best practice exchange. Creating distinct competence hubs and concentrating the knowledge management at these is supposed to drive standardisation and reduce internal complexity, also allowing for a higher degree of process mobility.
146
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
low Assure access to strategic markets and competitive factors, like …
Network capability level
high
Markets / Consolidation hubs Competitors Socio political factors Image
Accessibility
Supplier / Raw material
Assure access to Best cost resources of labour strategic importance, Skilled Labour like …
Thriftiness ability
Increase efficiency by …
Manufact. mobility
Provide mobility of …
Learning ability
Explore and exploit know-how and innovation about …
External know-how Economies of scale Economies of scope Reduction of duplication Products, processes, personnel Production volume & orders External factors Internal factors
TO-BE state
Fig. 39: Elevator NW: Network capability profile (TO-BE)
4.2.2.2 Scenario Development Besides the results from the AS-IS analysis, the Group management made a clear statement regarding the network’s future direction in the course of the cost leadership initiative, thereby defining the basic pillars for the Elevator NW’s TO-BE design: • Streamline the component portfolio! • Streamline the supplier base! • Establish a lead factory approach! Recalling the current network architecture and the network management’s tactics, these pillars are reasonable. The AS-IS analysis of the configuration and coordination layer revealed a broad supplier base with partially overlapping competencies, resources, and limited low cost access. Both internal and external suppliers act with a highly competitive mindset that lacks any kind of collaboration. With regards to the challenges the Elevator NW faces in Europe – the provision of a high external product variety to or below competitive prices – and the appropriate adjustment of the capabilities, configuration and coordination do no longer match with strategy. In other words, the architecture’s FIT is thrown out of balance, calling for realignment of its elements. Thereby, the three pillars stated by the top management shall guide the way. First, the realisation of economies of scale and scope, a reduction of duplications, and
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
147
better access to best cost labour are crucial to achieve price competitiveness; they constitute the need for a streamlined supplier base and product portfolio. Second, the promotion of high external product variety with manageable internal production complexity calls for stable, efficient, and standardised manufacturing processes. For this, third, a lead factory approach shall canalise the responsibilities for process and technology development and improvement, and also for manufacturing and engineering best practice sharing. Subsequently, insights are given how the network management architecture was applied to systematically support the conceptual design of the TO-BE network. It starts with the adjustment of the players and added value, i.e., the configuration decision dimensions, and outlines the impact on the coordination layer by aligning the rules. 4.2.2.3 Conceptual Network (Re-)Design Adjusting the Network Players & the Added Value The process of adjusting the network players and their added value is shown exemplarily for the Sheet Metal technology line. For this, Fig. 40 anticipates the findings by highlighting the aspired ideal plant role portfolio. It marks the reference point for the following discussion. The complete results, which also take the other three technology lines into account, are given thereafter.
Fig. 40: Elevator NW: Plant role portfolio for the Sheet Metal technology line (TO-BE)
148
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
Using the plant role portfolio as “playground”, first, a rough picture of the network’s TO-BE configuration was drawn by sketching the aspired competence spectrum and authority of the lead factory, discussing its impact on the supplier base, and by screening the players for potential lead factory candidates. This picture was detailed successively. In this course, a “lead factory target zone” was defined for the plant role portfolio and complemented by two additional targeted zones. The intended characteristics of all three corresponding plant roles can be outlined as follows: • A lead factory supplier is a leverage or prime supplier type which is closely embedded into the network and masters the full set of processes for Sheet Metal production. To avoid knowledge diffusion, this role has to be taken by an internal manufacturing site with access to know-how (skills and qualified personnel) as primary strategic reason. As a knowledge hub, the actual production volume of the lead factory is of minor importance. • Advanced suppliers are external leverage or prime supplier types similarly performing a wide range of processes for the Elevator NW, especially including own design and development activities. Leverage supplier types are small volume suppliers who provide special parts or competencies that add to the internal product portfolio. Thereby, a certain degree of access to know-how is required. Prime supplier types are large volume suppliers for low cost mass production; they require access to low cost labour to control manufacturing costs and proximity to consolidation hubs for transportation cost reduction. • Simple suppliers perform a limited range of processes only concentrated on production, quality, service & maintenance, and engineering; other competencies are provided by the lead factory. They are critical supplier types due to their large amount in production volume; this makes their access to low cost sources a prerequisite. Simple suppliers are preferably internal sites, but also tightly controlled and coordinated contract manufacturers are feasible. In this set-up, the lead factory becomes a central player. Fig. 41 outlines its intended authority along the introduced process landscape. It varies with respect to the other plants’ roles. Towards the internal simple suppliers, the lead factory shall be: • reactively supportive in operational processes, e.g., production and engineering, • proactively prescriptive by defining tools and setting and controlling standards for critical processes which execution still requires a certain degree of local freedom, i.e., for quality management and high volume service & maintenance, • responsible for carrying out core activities, such as design for cost / manufacturing, process and production design, and the strategic management of the 2nd tier supplier base.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
149
Towards external simple suppliers, the lead factory is supposed to act as prescriptive instance that defines and controls standards for most of the processes; towards external advanced suppliers, only a supportive role is considered attainable in a first step. Supportive & reactive
Prescriptive & proactive
Responsible
Lead factory authority (external) none
none
Lead factory authority (internal)
Supportive & reactive
Prescriptive & proactive
Responsible
DFM / DFC
Prototyping
Process design (h. vol / l. var) Production design (h. vol / l. var) Production (h. vol / l. var) Quality (h. vol / l. var) Service & maint. (h. vol / l. var) Process design (l. vol / h. var) Production design (l. vol / h. var) Engineering (l. vol / h. var) Production (l. vol / h. var) Quality (l. vol / h. var) Service & maint. (l. vol / h. var) Strategic SCM
Towards internal simple suppliers
Towards external simple suppliers
Towards external advanced suppliers
Fig. 41: Elevator NW: Lead factory authority for the Sheet Metal technology line (TO-BE)
The aspired plant role portfolio for the Sheet Metal line induces some changes in the current network structure and specialisation. The previous evaluation of the supplier base exposed Sp. 3 as the most promising lead factory candidate. The decision is grounded in the site’s leading design, engineering, and production competencies, especially in the “high segment”; though, it raises the need to strengthen the site’s capabilities for high volume / low variety production. The nomination of the lead factory comes along with the necessity to streamline the scattered supplier base. First, the foundation of an internal simple supplier for large scale low cost production is currently brought to decision along with the consolidation of the internal players. This low cost site shall be guided by the lead factory according to its authority profile. Moreover, the number of external players has to be limited to selected advanced suppliers, either as second source for large scale production, or for the provision of special parts. With the recent elimination of SP. 1, this evolution is in progress, too. The directions for the development of the other three technology lines were derived accordingly. Their ideal network configurations are summarised by the plant role portfolios as depicted in Fig. 42, details are discussed subsequently.
150
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
Fig. 42: Elevator NW: Plant role portfolio per technology line (TO-BE)
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
151
• For Mechanics & Components, a fundamental reduction of the own Mechanics in-house activities is planned, except for safety relevant devices as promoted by SP. 16. Yet, due to the large and variant number of mechanical components and related processes, a lead factory approach turned out to be inappropriate. Instead, cross-functional competence teams shall be dedicated to provide support and guidance for European component suppliers in selected critical processes; among those particularly the production and process design, the design for cost / manufacturing, and 2nd tier sourcing. Two subsidiaries of this support team are currently being established. One is located close to the external advanced supplier SP. 20, which is assumed to become the consolidated source for large scale low cost supply. The other is tied to the internal SP. 16, striving to move this towards a low cost site by exploiting the potential of low cost 2nd tier sources. Additionally, distinct advanced suppliers for exotic components will be necessary as leverage sources. • For Electronics, the dominating position of the internal SP. 7 will be strengthened by its official nomination as lead factory. Thereby, the low volume / high variety processes for the “high segment”, so far promoted by SP. 3, will be transferred in order to concentrate competencies. In turn, a large amount of the roughly 95% overall purchasing order volume provided by Sp.7 will be shifted to an internal or external simple supplier as low cost source. For this supplier, a subtle degree of know-how access is required to master the complex Electronics manufacturing processes. Access to the consolidation hubs is of minor importance because of a high packing density for Electronics devices. Additionally, there is little need to add external advanced suppliers since critical process competencies are fully internally available. • Finally for Drives, it is again SP. 3 which will be preferred to SP. 16 as lead factory due to its competencies in the complex low volume / high variety processes. The choice is supported by the network’s historical evolvement with a transfer of the Drives production from initially SP. 3 to SP. 16; this makes most of the related Electronics competencies grounded and still available at SP. 3. In turn, the lead factory shall facilitate the development of SP. 16 towards a simple supplier for low cost provision by taking the responsibility for the 2nd tier supplier management for leveraging low cost material access. In the shortrun, the dependence on external suppliers will remain, hence constituting the position of SP. 21 as external advanced supplier. In the long-run, an insourcing of the Electronics business might be reasonable. This would induce a shift of production volume to the internal SP. 16 and cause a gradual downgrading of the advanced suppliers to special parts providers only.
152
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
Overall, the plant role portfolios display ideal settings which guide the direction for the development of the network configuration. Although first measures are in progress to support this direction, the transformation of the AS-IS towards the TO-BE state requires time and endurance and might be subject to unforeseen modifications. Aligning the Rules The targeted modifications of the network configuration raised the need for an adaption of the coordination dimensions, i.e., an alignment of the rules. Again, the management frameworks were applied as “playground” to discuss possible changes and for iteratively sketching the TO-BE scenario. Though, it has to be noted that the following description of the coordination layer had to account for the transition phase in the network configuration. Since the implementation of the lead factory approach and the subsequent adaptations of the plant role portfolios constitute an on-going transformation, the characteristics of the coordination dimensions are required to already underline the change phase and not only the final and ideal TO-BE state. This induces uncertainties in the concept; the not yet fully defined number and ownership of the supplier base, for instance, prevents the final design of the resource strategy. As to centralisation and standardisation, the framework depicted in Fig. 43 translates the lead factory’s authority as sketched in Fig. 41 to the network perspective. In line with the understanding of Tykal (2009), for this, the lead factory is considered as organisational middle way between centralisation and local autonomy. Internally, the aspired TO-BE state constitutes a paradigm shift for manufacturingrelated responsibility categories from autonomy to centralisation and standardisation; this is particularly necessary to control for internal complexity. Responsibility for systems in prime and support activities will be assigned to the lead factories to promote a harmonised network-wide implementation; only the creation and maintenance of product development data (2a) and the management system (4) remain under the Group’s control. The lead factories are designated to drive operational excellence in the technology lines by carrying responsibility for the coordination and implementation of improvement programs in and besides production (1a, 1b, 5). Moreover, the concentrated development of quality and maintenance tools & methods (3b, 3c) is supposed to enable a homogeneous roll-out of the respective systems, beyond the prescription of standards only. Similarly, most of the manufacturing-related decisions regarding process and technology allocation (14, 15) and also the make-or-buy strategy for existing products (11b) will be concentrated at the lead factories, complemented by the assignment of authority for manufacturing IT decisions (10). Decisions affecting the product
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
153
allocation (12) and capacity development (16, 17) will remain under regional or Group control. Additionally, a process is required to manage the sites’ strategies & roles (8). Concentrating manufacturing decisions supports the formalisation of operational processes, e.g., for custom design engineering (24c) and manufacturing (24f). In combination with the central authority for manufacturing IT decisions, such standardisation is prerequisite for a harmonised production-related data management (2b). This system, in turn, allows for transparency and comparability of the operational performance as source of continuous improvement. More specific processes will be brought directly under the lead factories’ control, meaning a harsh cut for the sites’ autonomy; they range from prototyping (24a), process and production design (24d, 24e), to design to cost / manufacturing (24b). Especially the latter carries on the latest movement towards a closer connection between manufacturing and design functions as imperative for cost cutting. Accordingly, the 2nd tier supply chain management (18) is designated as lead factory process in order to leverage material costs; with above 60% these account for the biggest share of the total manufacturing costs. Responsibility areas
Each site individually
Centralisation / Responsibility
S
System
D
Decision P Process
23
Several sites
Autonomous
!
24 c
Lead factory
11
Region Europe (SC-Europe)
10 b 17
9
Centralised
Standardised 24 c
6 7
3b 1a 1b 2b 3b 3c 5 7 24 24 24 14 15 18 a b 24 e d 12
Centralised & Standardised21
16 6 No / Local standardisation
Documented rules, guidelines & processes
Audited / Controlled processes & routines
Individual tools / heterogeneous implementation level at each site
Individual tools / homogeneous implementation level at each site
Standardised tools / heterogeneous implementation level in the network
3a 3a
2a
4
11 a
Standardised (IT-) tools or methods Standardised tools / homogeneous implementation level in the network
Internal view External view
Production system (layout) Production system (processes) Product data mgt. system (dev.) Product data mgt. system (prod.) Quality Mgt. (system & stand.) Quality Mgt. (tools & methods) Maintenance system Management system Improvement programs (besides production) HR system Know-how exchange system
Decision
19 22 19 21 22
13
20
1a 1b 2a 2b 3a 3b 3c 4 5
24 b
18 24 f
8
Group level
Degree of standardisation for the network
!
System
1-24 from external view excl. 3a, 3b, 18, 19, 21, 22, and 24b, 24c
P D S
8 Site strategy & roles 9 Organisational structure 10 Manufacturing IT decisions 11a Make-or-buy (new product) 11b Make-or-buy (existing product) 12 Product allocation decisions 13 Transfer pricing 14 Production process decisions 15 Manufacturing techn. decisions 16 Long-term capa. development 17 Short-term capacity adjust.
Process
18 Strategic sourcing (2nd tier) 19 Strategic logistics 20 Product cost calculation 21 Long-term S&OP 22 Intern. SC-planning / Order alloc. 23 Short-term manuf. planning 24a Prototyping 24b Design for cost / manuf. 24c Custom design engineering 24d Process design 24e Production design 24f Manufacturing
Fig. 43: Elevator NW: Centralisation & standardisation framework (TO-BE)
Externally, simple suppliers will be treated similar to internal sites, with the lead factories not being responsible but prescriptive for any production-related issues, i.e., the production, quality, and maintenance system, production and technology decisions, and the design of operational processes. Regarding external advanced suppliers, as
154
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
explicitly highlighted in the framework, the lead factories’ authority is intended to be limited to critical areas at first, i.e., to the most crucial cost drivers by acting as prescriptive instance for design to cost / manufacturing (24b) and to the 2nd tier supply chain management (18), as well as to quality management by defining methods and tools (3b) in order to drive the harmonisation of the quality system (3a). The resource allocation underlines the concentration efforts that go beyond the centralisation of authority. The internal view, as outlined by the “bubbles” in Fig. 44, targets two directions: (1) the promotion of cooperation and (2) a reduction of duplications and facilitation of resource sharing by departmentalisation and pooling. Scarcity in the network
!
Limited amount Rather limited amount Balanced amount
2b, 4b,5b
4.2 a 4.2 4.2 4.2 c d
Sufficient amount
1
Competition
4.1
Rather sufficient amount
Lead factory resource
2a
2,5
2c
!
4.1 a 4.1 4.1 d c
Cooperation 3
2d
5.1 a 5.1 d
5.1 c 6b
Dedicated Resources
6b
3
Resource Pool
Possessing sites = requiring sites Possessing sites < requiring sites Internal view
External view
2 3 4
a
a large extent
5 Custom design engineers (CDE) Strategic supply chain specialists (SCM) Basic manufacturing capacity 6 4.1 Manufacturing capacity 4.2 Standard manuf. equipment & machines
Sheet Metal
b
Mech. & Comp.
5.2 a 5.2 d 5.2 c
Intensity of sharing / exchange Frequently and to
No exchange / Seldom and to sharing a small extent Extent of provision as proportion of possessing Resource categories: sites vs. requiring sites: 1 R&D capacity Possessing sites > requiring sites
1
c
Special manufacturing capacity 5.1 Process design specialists (PE) 5.2 Production design specialists (PS) Support team incl. 3 and 5
Electronics d Drives
All tech. lines
Fig. 44: Elevator NW: Resource allocation & sharing framework (TO-BE)
Regarding the pooling attempts, the concentration of resources is assumed to leverage synergies between functional specialists and to reduce their scarcity. This primary affects the bundling of the process and production design specialists (5.1, 5.2) and supply chain managers (3) at the lead factories for the Sheet Metal, Electronics, and Drives technology. Activities for the Mechanics & Components line will basically be outsourced, with only limited internal resources dedicated to SP. 16 (2b, 4b, 5b) or pooled at the two support teams (6b incl. 3b and 5b). Regarding the cooperation attempts and for those resources requiring some local proximity, the lead factory approach is considered as prerequisite for exchangeability by promoting and assuring a higher degree of standardisation. This particularly accounts for a stronger cooperation between those specialists tied to the customisation process, i.e., the custom design
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
155
engineers (2), which will be coordinated by the respective lead factories. Manufacturing capacity (4) will only sporadically be subject to exchange since order allocation remains under control of the order centre. Moreover, the scarce and centralised R&D capacity (1) at the headquarters shall be discharged by limiting access to the lead factories solely; these get deeper involved into the design process. From an external view, simple suppliers will again be handled analogously to internal sites. They are strongly guided by the lead factories which use their resources to proactively prescribe and promote competencies beyond production, quality, service & maintenance, and engineering to contract manufacturers. On the contrary, advanced suppliers, as depicted in the framework, remain autonomous entities with dedicated resources. Support can be given on demand but not in terms of a joint and unlimited resource provision. Exceptions are only the lead factories’ prescriptive function for 2nd tier supply chain management (3) and the support teams (6) as designated resources for the Mechanics & Components. The discussion on the rules is ended by the modifications of the internal and external information & knowledge flows; they are shown in Fig. 45 and Fig. 46. 49 I
Exchange structure
External Info.
Centrally provided
I
Internal Info. 1
9
6
4
Transparency 15
3
12
2
Limitation Centrally coordinated
Lead factory competence
Knowledge
K
11
8
10
13b
5
14
7 13a
Centralised & decentralised
!
Isolation
Networking
Decentralised
No exchange
No access
Access to limited data / info.
Access to most data / info.
Access to full data / info.
No access
Access limited to selected sites
Access for all requiring sites
Access for all sites
I: Availability of information K: Intensity of sharing High
Medium Low
Internal view External view
Exchange mechanisms information Informal channels, e.g.: • ad hoc telephone calls & e-mails, meetings • social activities
I
Information & knowledge categories External information 1 2 3
Markets / Customers Competitors Suppliers
4 5 6 7 8 9
Site strategy & roles Financial site performance Market & sales performance Operational site performance Sales & operations planning Manufacturing planning data
Internal information
Knowledge
10 Product innovations 11 Product changes / improvements 12 Technology / process innovations 13a Manufacturing best practice 13b Engineering best practice 14 Management experience & practice 15 Business & supp. proc. improv.
I Degree of transparency K
Exchange mechanisms knowledge
Formal channels, e.g.: • databases & sharepoints, intranet • regular & formal meetings
Moving people / job rotation
Customised projects / project support
K Competence Training & groups qualification Manuals, systems, databases Mechanism not used
Fig. 45: Elevator NW: Information & knowledge sharing framework - internal (TO-BE)
Internally, the isolation position of the players in the “knowledge network” shall be 49
It has to be noted that modifications in the exchange mechanisms were not discussed.
156
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
resolved by fostering transparency and cooperation. Hence, as implied by the centralisation and standardisation framework, the lead factories are considered as internal knowledge hubs responsible for the collection, execution, and distribution of best practices in the manufacturing, engineering, and support processes of their technology line (13a, 13b, 15). But since they are manufacturing sites themselves, it will become the central Group’s objective to coordinate and promote an internal performance benchmarking to avoid any bias. Moreover, the concentration of manufacturing-related decision making designates the lead factories as central providers for manufacturing technology and process innovations (12). Similarly, by taking the responsibility for the design to cost / manufacturing process, they come closer to the R&D function, hence intensifying the spreading of product innovations (10) in the network. Additionally, the concentration of the 2nd tier supply chain management enables a central coordination of supplier data (3) by the lead factories, but in tight cooperation with the other internal sites, which are closer to the actual supplier performance. To empower the lead factory concept, management commitment and a transparent communication of the plant roles & strategies (4) by the Group function is crucial for supporting the players’ roles and authorities in the network. I
Exchange structure
External Info.
Centrally provided
I
Internal Info. 4
7
8
Limitation
K
13a
12
Lead factory competence
Knowledge 13b
3
11
Centralised & decentralised
2
1 Isolation
Networking
No access
Access to limited data / info.
Access to most data / info.
Access to full data / info.
No access
Access limited to selected sites
Access for all requiring sites
Access for all sites
I: Availability of information K: Intensity of sharing High
Medium Low
Internal view External view
Exchange mechanisms information Informal channels, e.g.: • ad hoc telephone calls & e-mails, meetings • social activities
I
1 2 3
Markets / Customers Competitors Suppliers
4 5 6 7 8 9
Site strategy & roles Financial site performance Market & sales performance Operational site performance Sales & operations planning Manufacturing planning data
Knowledge
Decentralised
No exchange
External information
Internal information
15
Transparency
Centrally coordinated
Information & knowledge Categories
10 Product innovations 11 Product changes / improvements 12 Technology / process innovations 13a Manufacturing best practice 13b Engineering best practice 14 Management experience & practice 15 Business & supp. proc. improv.
I Degree of transparency K
Exchange mechanisms knowledge
Formal channels, e.g.: • databases & sharepoints, intranet • regular & formal meetings
Moving people / job rotation
Customised projects / project support
K Competence Training & groups qualification Manuals, systems, databases Mechanism not used
Fig. 46: Elevator NW: Information & knowledge sharing framework - external (TO-BE)
Externally, the long-term strategy intends a development from literally no cooperation in the current state towards the networking position by establishing formal and
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
157
informal lateral ties, in particular, with the advanced suppliers. Thereby, the Elevator NW is ready to set a good example, proving these with access to best practice knowledge (13a, 13b, 15) and to selected technology or process innovations (12) as starting point. To underpin this approach, the lead factories will also communicate selected operational performance data of internal sites (7) to externals, revealing both internal strengths but also own improvement potential. Additionally, the prescriptive role for the 2nd tier supplier management requires a bilateral information exchange with externals regarding supplier information and performance (3). Overall, again an open communication of selected information about the own site strategies & roles (7) in the supplier base is required – at least about the designated lead factories. Summarising the TO-BE Concept The network players, their added value, and the rules have to match with each other and with the defined tactics. In other words, the decision dimensions of the configuration and coordination layer have to FIT among each other and with the targeted network capabilities. The obtained TO-BE scenario is supposed to meet these requirements. With regards to the shift in the network capabilities, as shown in Fig. 39, the main features of the intended concept can be resumed shortly as follows: • Economies of scale and scope and a reduction of duplications shall be exploited by consolidating the supplier base and defining three plant roles: Production volume is bundled at large scale simple or advanced suppliers, exceptional competencies are provided and specialists pooled at internal lead factories, “exotic” components are sourced by small scale advanced suppliers. Moreover, the internal semi-finished goods spectrum and the supplier portfolio are streamlined by a retraction from the Mechanics & Components technology. • The selection of large scale manufacturers is subject to their provision of low cost labour and proximity to consolidation hubs. The lead factories assure access to skilled labour and know-how; a concentration of the 2nd tier supply chain management leverages the potential of raw material costs. • Internal learning is facilitated by the lead factories as central hubs for best practice identification and promotion and as providers of process and technology innovations. Output and behavioural control is extended to operational performance, creating transparency in processes. • Finally, not the manufacturing mobility but the degree of standardisation is targeted by limiting site autonomy and increasing the degree of centralisation, harmonisation, and formalisation for manufacturing-related systems, strategic decisions, and processes, thus reducing internal complexity.
158
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
4.2.3 The “Chocolate NW” Case The presentation of the Elevator NW case aimed at a methodical application of the management architecture and the single frameworks to systematically map and conceptually (re-)design the manufacturing network. The main pillars for the targeted TO-BE scenario were predefined by the top management in advance. The following discussion of the Chocolate NW case is concentrated on the derivation of these very pillars. Coming from literally no “network thinking”, the creation of a common vision regarding the aspired TO-BE scenario turned out to be necessary for bridging the gap between the AS-IS analysis and the design of the TO-BE state. To avoid any repetition of what has been demonstrated so far, findings from the AS-IS analysis are resumed only shortly; deeper insights are given to the methodical approach of sketching and evaluating potential TO-BE scenarios as guiding visions for the network (re-)design. The Chocolate NW represents the collectivity of eight largely independent production companies of a traditional European prime chocolate manufacturer; each company is attached to an autonomous sales & distribution organisation covering the biggest market regions served. The holding structure of the overall Group is complemented by ten additional sales & distribution organisations without own production activities as well as an international department located at the headquarters and representing the Group interests; this department also hosts the global operations function. With an average turnover of about €2 billion and €250 million operating profit (EBIT), the Group has more than 7000 employees; about 50% of these are related to operations. 4.2.3.1 Network Analysis & Target Setting Players & Added Value The evolvement of the Chocolate NW was subject to inorganic growth within the past years. A wave of acquisitions and the integration of licensees in the course of the Group’s global expansion strategy led to a fragmented network configuration. Today, the eight production companies comprise nine manufacturing sites in two world regions: Six of them are concentrated in Western Europe, two in North America. The multiplant strategy reflects a mixture of market area and product plants: Coming from a pure market orientation, the plant’s original focus on local needs has been increasingly diluted by the growing importance of the intercompany business in recent times. Today, intercompany deliveries from a manufacturing site to a sales & distribution company different than the one it is attached to account for about one third of the overall production volume. The products themselves are grouped in eight product lines with a substantial degree of locally dedicated variants for each, but the growing intercompany business is more and more questioning this market dedication.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
159
Nonetheless, still only selected variants are originally designated as global products. The plants serve direct customers, group-owned retail shops and flagship stores, as well as the 3rd party retail business. Manufacturing can be described by six process levels: Cocoa beans, cocoa butter, sugar, milk powder, and other ingredients are processed to cocoa liquor in level 1 and transformed into chocolate mass by a conching process in level 2. Level 3 comprises the provision of filling mass for distinct product types. The final product shape is generated in level 4 by an on-line molding or enrobing process. The finished products are wrapped and packed in level 5 and 6, again on-line or off-line. Manufacturing is complemented by R&D competencies which cover product development, i.e., the design of new receipts, shapes, and packaging solutions, technology development for product launches, and process engineering. Machine design and tooling is outsourced, concentrating the core competencies on the products and manufacturing processes. Owed to the network’s historical development, most sites cover the full process spectrum, except for cocoa liquor manufacturing in Europe, which is concentrated at two distinct sites; one of them is a process plant purely dedicated to level 1 production, the other performs the full bandwidth of production levels. Packaging capacity is widely scattered and provided both internally and by external co-packers. Except for manufacturing at the headquarters, which – as the biggest site – guarantees for the company’s historical heritage, hosts the international department, and is the only source for overseas business, each player’s added value is limited to contributions to the own sales & distribution companies mainly. The characteristics of the Chocolate NW are summarised in Tab. 22. Network scope
Company Division Business Other level level unit level Chocolate NW
HQ Industry n.a. Food products
Core products Chocolate & pralines
Core processes Food processing & packaging
Network characteristics
# Sites # Employ. Global (operat.) (operat.) dispersion 9 > 3000 Multinat. (2 regions)
Network structure* Local for local / World products
Network category** IV
* According to Meyer and Jacob (2008) ** According to the survey classification of the network capability level and conformance
Tab. 22: Chocolate NW: Network characteristics
Rules The coordination layer underpins the impression from the network configuration. The manufacturing sites / production companies are organised as cost centres attached to the respective sales & distribution companies; these are run as independent profit centres. Although the international department is supposed to assure a certain degree of central control on operations, the intended matrix organisation with an even distribution of power between local and global operations is not yet balanced; instead, weight is put on the authority of the local sales & distribution companies. Findings from the AS-IS analysis on the coordination dimensions can be resumed as follows:
160
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
• Generally, the network shows a limited degree of centralisation and standardisation for most of the operations-related systems, decisions, and processes. It ranges from no common production system, over the plants’ authority on process and IT decisions, to a decentralised receipt management. The allocation of responsibility in manufacturing, technology, and planning accounts for the autonomy of the sales & distribution companies, striving for a local optimisation but lacking any network-wide formalisation. Only some long-term-related strategic issues are centralised and partially standardised, among those, the quality system, the financial management and KPI system, as well as the long-term sales & operations planning. • Similarly, resources allocation is primary dedicated to the independent sites with little sharing. Except for the pooling of the cocoa liquor production, the network reveals redundancies in the allocation of the production capacity, the R&D and engineering specialists, and the support functions. Hence, distribution companies compete for the investment budget in assets and specialists. • Consequently, the incentive system supports an individualism strategy. It is tailored to the autonomy of the sales & distribution companies, setting targets on the single site’s level and linking them to outcome-related performance categories. More precisely, the profit-oriented sales & distribution companies are incentivised by financial and market performance, which is broken down into operational targets for the manufacturing sites run as cost centres. • The design of the “information & knowledge network” underpins the insular set-up. While information exchange about external markets and competitors is not part of the manufacturing network, most of the other internal information is centralised but with limited access. Knowledge exchange is stuck in the isolation position, with little and only decentralised exchange of product innovations and management experience and only basic best practice sharing as part of annual regular operations meetings. Overall, the analysis of the configuration and coordination layer reveals a loose aggregation of isolated sites, which autonomously provide mostly local products, rather than an integral and cooperative network with a common vision and mission; accordingly, any common “network thinking” is weak. Tactics & Scope With respect to the network’s history and past challenges, the AS-IS state indeed seems consequent. Manufacturing sites are tied to single sales & distribution companies serving local markets as entrepreneurs with tailored products mainly, thus
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
161
making any “network thinking” superficial. The current manufacturing strategy underlines this approach by emphasising specification quality and product range / design flexibility as order winners, complemented by quality conformance and delivery reliability as most critical market qualifiers. This strong emphasis on customer orientation constitutes the Group’s position as leading prime chocolate manufacturer and allows for a distinct price premium in the luxury goods segment. But with respect to the emerging challenges, the necessity for a mind change comes up. On the one hand, the products’ nature as luxury goods and the Group’s brand recognition inhibit substantial changes regarding the current order winner strategy. On the other hand, the pressure on some of the market qualifiers grows steadily; among those is rising cost sensitivity due to increasing and volatile raw material prices but also due to currency fluctuations in the last years, both eating up the price premium. Additionally, the considerable business with the large 3rd party retailers gives more priority to delivery speed and flexibility in order to meet the smaller time fences for product supply; a situation that is even accelerated by a limited planning accuracy and stability as a consequence of seasonal and volatile demands. Finally, the growing intercompany business questions the network’s loose coordination; this already highlights some weaknesses. To give examples: Since the incentive system targets the sales & distribution companies’ market performance only, in good times, these prefer meeting local demands by offering products with a high price mark-up instead of covering the intercompany demand with lower transfer prices; a decision that might be suboptimal from a network perspective. In bad times, the plants fight for intercompany volume to cover their fixed costs. Moreover, the combination of central transfer price decisions but autonomous and less standardised product cost calculations evokes nontransparency and internal conflicts for the intercompany volumes. Likewise, the dedication and lacking coordination of resources for the receipt and packaging development wastes potential; not only by letting isolated entities working on similar solutions but also by missing the opportunity to make innovations quickly to diffuse in the network. Similarly, the potential of best practice exchange remains unexploited even though most plants perform an identical set of manufacturing processes. The outlined changes in the network “scope” require an alignment of the network capabilities with the altered manufacturing strategy, i.e., a (re-)shaping of the network tactics. Fig. 47 outlines the current and targeted network capability profiles as perceived by the operations management team of the international department. The current profile emphasises the results from the AS-IS analysis: The network offers only limited exploitation of thriftiness, mobility, and learning abilities due to the insular organisations. On the contrary, the focus is set on customer proximity owed to the local sales & distribution companies. The Group additionally capitalises its image
162
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
by promoting its local heritage as selling argument. Overall, the profile reveals only little utilisation of the network’s potential, confirming its weak position in the network capability level and conformance evaluation in Fig. 31. low
Network capability level
high
Markets / Customers
Assure access to strategic markets and competitive factors, like …
Competitors Socio political factors Image
Accessibility
Supplier / Raw material
Assure access to Best cost resources of labour strategic importance, Skilled labour like …
Thriftiness ability
Increase efficiency by …
Manufact. mobility
Provide mobility of …
Learning ability
Explore and exploit know-how and innovation about …
External know-how Economies of scale Economies of scope Reduction of duplication Products, processes, personnel Production volume & orders External factors Internal factors
AS-IS state TO-BE state
Fig. 47: Chocolate NW: Network capability profile (AS-IS vs. TO-BE)
The capabilities’ TO-BE profile primary targets an efficiency increase as reaction on shrinking margins. It addresses a consolidation of the small manufacturing volumes to benefit from scale effects, a bundling of similar product types to achieve economies of scope, and a subsequent reduction of administrative duplications. Such steps, in turn, might negatively affect the access to markets. However, it is not solely the production function but more the local packaging that requires customer proximity to fulfil the expectations regarding tailored package displays and languages. Technically, the products’ shelf life and cost structure allows for a certain degree of concentration as long as the increase in transportation costs is overcompensated by thriftiness effects. Moreover, fostering cooperation and internal learning is intended to facilitate best practise exchange and drive operational excellence. It also accounts for the growing intercompany business by constituting a common “network thinking”. Access to low cost labour and raw material, on the contrary, are considered as minor cost levers due to a high degree of automation in processes, global sourcing channels, strict quality requirements for ingredients, and due to only low transportation cost for packaging material. Finally, facilitating internal learning is supposed to drive standardisation,
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
163
allowing for some flexibility in the short-term order allocation but mainly, as risk mitigation, in the long-term mobility of products, processes, and personnel. 4.2.3.2 Scenario Development The lacking “network thinking” so far motivated the need for an initial formulation of a joint vision as common denominator for the future network design. For this, possible TO-BE scenarios were sketched and evaluated, and the bricks of the most promising scenarios consolidated to an aspired TO-BE vision for the subsequent network design. Themes for the scenario creation were derived with respect to the dynamics in the manufacturing strategy, leading to the selection of a (1) “customisation”, (2) “best cost”, and (3) “speed” scenario. These were contrasted by the current AS-IS state. To provide a fundamental understanding of the themes, a more detailed description of the characteristics of their underlying configuration and coordination decision dimensions was required. Yet, using the management frameworks for this step would have meant a very high effort necessary at the early stage of the network development. Instead of doing so, the “essence” of each framework was synthesised by a set of core decisions which were translated into so-called “tension lines”. Tension lines represent a wellknown concept characterising a decision by “… an extreme position at the left and right ends of the line” (Friedli et al., 2010, p. 205). The process of developing the scenarios along the tensions lines allowed for a target-oriented discussion ending in drawing a unique profile for each. The isolation of the core decisions and the creation of the tension lines themselves were conducted with respect to usability and applicability, but having the theoretical foundation of the frameworks in mind. Basically, just the bipolar ends of each tension line were defined, enabling to point out tendencies; only for decisions requiring a definite answer, the middle positions were added. It was further differentiated between first order and second order decisions, reflecting a growing depth of detail. The applied tension lines and the individual profiles for the four scenarios (including a summary of the AS-IS state) are outlined in Tab. 23, the characteristics of their respective network architectures will be resumed as follows. The “customisation” scenario aims at providing the highest degree of customer orientation and market responsiveness, targeting a fundamental sales growth due to the maximisation of local product customisation that over-compensates the network’s structural inefficiencies. Basically, the scenario emphasises the AS-IS strategy by ignoring the emerging intercompany business. Its configuration, therefore, empowers the insular market area plants, leaving them under the control of the profit-oriented sales & distribution companies; in the long-run, local proximity might even call for the establishment of new plants, which would increase the network’s dispersion. To assure
164
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
local responsiveness, the isolated sites continue performing almost the full bandwidth of processes with little central control and standardisation. Consequently, development competencies remain decentralised while a site’s autonomy is actually extended by permitting customer-specific technology and equipment. Thus, product design and allocation is focused on local demands, restricting any load balancing due to tailored processes and little standardisation. Likewise, resources are dedicated without sharing, and slack resources and capacity cushions are accepted to increase a site’s reactivity. The degree of both cooperation and competition between the sites is naturally low due to their non-overlapping scope, except for the coverage of new markets. Although information & knowledge exchange remain centralised, transparency and sharing is limited owing to structural differences between the sites and markets served. Overall, the scenario benefits from an optimal fulfilment of market requirements and improved customer relations, but it entails the risks of inefficiencies and a flat learning curve of the plants as a consequence of their isolated R&D activities. It also reveals vulnerability regarding order fluctuations due to the lacking possibility of load balancing and might dilute the global brand recognition and reputation as a consequence of an ample share of local specialities in the product portfolio. The “best cost” scenario aims at globally optimising the direct and indirect manufacturing cost structure, targeting a maximisation of the product margin by increasing efficiency and standardisation. The underlying network configuration strives for exploiting economies of scale and scope and reducing administrative duplications. Compared to the AS-IS state, the level 1 and 2 processes will be even more distracted and bundled, feeding distinct world plants for the subsequent process steps; however, the configuration is not subject to best cost labour access because of the high degree of automation. Further, engineering, supply chain, and support functions, as well as, in particular, the product development will be pooled, constituting a single source for product innovation and a controlling instance for product variants. A functional organisation runs the sites as cost centres and carries the responsibility for the network’s coordination and global optimisation. Therefore, central authority is strengthened to foster a network-wide harmonisation and global standardisation of processes and decisions; only the product orientation of the plants calls for some local freedom in manufacturing-related areas. Process standardisation, in turn, enables short-term load balancing; although, as a consequence of the product concentration, it is restricted to exceptions only (e.g., for risk mitigation). The sites are supposed to act as embedded “network players” that basically cooperate; only a subtle degree of competition regarding the operational performance is intended to leverage cost improvements. Finally, information & knowledge exchange are kept centralised but will be facilitated. Regarding information sharing, full transparency is not
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
165
considered necessary since the product plants are lacking full comparability. To amplify cost-savings, for knowledge sharing, a networking position is aspired based on both centrally coordinated and decentralised informal exchange. Overall, the scenario substitutes customer orientation in terms of product variety and responsiveness by efficiency. It benefits from reducing the costs of goods sold by leveraging the network’s thriftiness and from controlling the capital employed. The central development activities are supposed to drive break-through innovations but entail the risk of lagging behind for incremental improvements. Finally, the “speed” scenario aims at providing a fast, flexible, and reactive order fulfilment, targeting sales growth specifically in the 3rd party retailer business by minimising the customer lead times. Therefore, the network configuration requires local responsiveness, thus adopting the footprint of the AS-IS state with general purpose plants close to the markets. But different from the current state, a high degree of standardisation in manufacturing processes is intended, allowing for production volume flexibility, coordinated by a central supply chain management. Similar to the “best cost” scenario, sites are again attached to a functional organisation in order to substitute local optimisation with global interests. They have to be run as cost centres to sacrifice their individual market performance for a central and flexible order allocation. Accordingly, a high degree of parental control with full centralisation and standardisation for systems, decisions, and processes is needed, especially for manufacturing-related areas and also for the strategic and operational internal supply chain management. Manufacturing resources are dedicated but with much sharing; product development, engineering, supply chain, and support functions will be pooled. To meet the joint network goals, cooperation is necessary, particularly regarding market coverage. Concerning operational performance, some degree of cooperation is promising due to the comparability of processes while competition is beneficial to motivate operational excellence. Further, network coordination has to facilitate knowledge sharing but with a central exchange structure to prevent any dilution of global standardisation. Analogous, information exchange calls for central control, especially for planning and administrative data, and also for financial and operational performance; a high information transparency underlines the cooperative culture. Overall, the scenario is supposed to allow for an improved customer management regarding delivery speed and flexibility, coping with demand fluctuations. It enables a reduction of internal stock levels and a higher inventory turnover as well as a prevention of stock-outs at the customer shelves. The aspired standardisation, however, will negatively affect the customer orientation concerning product range and design flexibility.
166
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
Network specialisation
Network structure
Configuration & coordination decision dimension(s) 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Manufacturing footprint design driven … Multiplant strategy (products) Multiplant strategy (processes) Geographic dispersion of the network
by efficiency "product plants" "process plants" simple (national)
Dominating strategic site reason Legitimation for manufacturing sites Degree of site embeddedness into the network
access to low costs access to internal skills short term persp./demand network players
Standard manufacturing competence (level 6) Standard manufacturing competence (level 3-5) Special manufacturing competence (level 2) Special manufacturing competence (level 1) Strategic SC competence (purchasing, logistics & SCM) Operative SC competence (purchasing, logistics & SCM) Basic development competence (engineering)
outstanding at single site(s) outstanding at single site(s) outstanding at single site(s) outstanding at single site(s) outstanding at single site(s) outstanding at single site(s) outstanding at single site(s) outstanding at single site(s)
Centralisation & standardisation
Network organisation
15 Core development competence (products, technology)
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
Organisation of the manufacturing function Responsibility for financial site performance (P&L) Responsibility for operational site performance Responsibility for financial network performance Responsibility for operational network performance Design of administrative functions and structures
complex (matrix) corporate corporate corporate corporate copied structures
Degree of parental control in the network Degree of site authority Degree of standardisation
tight low high
Authority for systems for primary activities Degree of harmonisation of systems for primary activities Authority for systems for support activities Degree of harmonisation of systems for support activities Authority for organisational decision making Degree of standardisation of organisational decision making Authority for product decision making Degree of standardisation of product decision making Authority for manufacturing decision making Degree of standardisation of manufacturing decision making Authority for the execution of strategic processes Degree of standardisation of strategic processes Authority for the execution of operational processes
central unit high central unit high central unit high central unit high central unit high central unit high central unit high
Resource allocation & sharing
38 Degree of standardisation of operational processes
39 Allocation of resources 40 Intensity of sharing / exchange 41 Resource availability 42 43 44 45 46
Allocation & sharing of basic manuf. capacity (level 3-5) Allocation & sharing of special manuf. capacity (level 1-2) Allocation & sharing of product & techn. development specialists Allocation & sharing of engineering specialists Allocation & sharing of supply chain specialists
47 Allocation & sharing of support functions
Incentive system
Cooperation in financial areas Competition in financial areas Cooperation in market & sales areas Competition in market & sales areas Cooperation in operational areas
55 Competition in operational areas
Information sharing
56 Information transparency in the network 57 Exchange structure of information in the network 58 59 60 61 62 63 64
Transparency of external inform. (custom., comp., supplier) Exchange structure of external inform. (custom., comp. supplier) Transparency of financial / market & sales performance Exchange structure of financial / market & sales performance Transparency of operational performance Exchange structure of operational performance Transparency of planning and administrative data
65 Exchange structure of planning and administrative data
Knowledge sharing
66 Intensity of knowledge sharing in the network 67 Exchange structure of knowledge in the network 68 69 70 71 72 73
"AS-IS"
Intensity of product knowledge sharing (innov. & improv.) Exchange structure of product knowledge (innov. & improv.) Intensity of techn. & process knowledge sharing (innov. & improv.) Exchange structure of techn. & process knowledge (innov. & improv.) Intensity of non-production knowledge sharing (support & mgt.) Exchange structure of non-production knowledge (support & mgt.)
"Customisation"
"Best cost"
regional
multinational
by customer proximity "market area plants" "general purpose plants" complex (worldwide)
access to exter. knowledge proximity to markets / cust. long term persp. / demand isolated similar at every site similar at every site similar at every site similar at every site similar at every site similar at every site similar at every site similar at every site
divisional mix division division division division
region / company region / company region / company region / company region / company
concentrated high sufficient
48 Degree of cooperation in the network 49 Degree of competition in the network 50 51 52 53 54
Tension line(s)
concentrated & much sharing concentrated & much sharing concentrated & much sharing concentrated & much sharing concentrated & much sharing concentrated & much sharing
direct (functional) site site site site context dependent struct. loose high low every site / company low every site / company low every site / company low every site / company low every site / company low every site / company low every site / company low
dedicated low scarce dedicated & little sharing dedicated & little sharing dedicated & little sharing dedicated & little sharing dedicated & little sharing dedicated & little sharing
high high
low low
high high high high high high
low low low low low low
high transp. / open access centralised high transp. / open access centralised high transp. / open access centralised high transp. / open access centralised high transp. / open access centralised
high centralised high centralised high centralised high centralised
low transp. / limited access decentralised low transp. / limited access decentralised low transp. / limited access decentralised low transp. / limited access decentralised low transp. / limited access decentralised
low decentralised low decentralised low decentralised low decentralised
"Speed"
Tab. 23: Chocolate NW: Scenario development via tension lines
The common understanding and rough description of each scenario defined the basis for their evaluation. This evaluation aimed at sustaining FIT with the aspired network
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
167
capabilities (tactics) as well as with selected contextual factors (scope). The FIT with manufacturing strategy was not checked explicitly since manufacturing strategy was supposed to be aligned with and reflected by the targeted network capabilities already. Tab. 24 summarises the evaluation approach and the results of the assessment. 50 Scale Assessment criteria
weight
-2
1
much worse
similar to AS-IS
much better
4
much worse
similar to AS-IS
much better
2
much worse
similar to AS-IS
much better
4 FIT with corporate culture
2
much worse
similar to AS-IS
much better
Availability of internal resources to manage the 5 change (competence and time)
3
much worse
similar to AS-IS
much better
6 FIT with food safety and regulatory requirements
5
much worse
similar to AS-IS
much better
7 FIT with traceability requirements
5
much worse
similar to AS-IS
much better
8 FIT with current plant roles and competencies
1
much worse
similar to AS-IS
much better
3
much worse
similar to AS-IS
much better
1
much worse
similar to AS-IS
much better
11 FIT with product quality expectations
4
much worse
similar to AS-IS
much better
12 Proximity to markets / customers
3
much worse
similar to AS-IS
much better
13 Proximity to competitors
1
much worse
similar to AS-IS
much better
14 Exploitation of socio political factors
3
much worse
similar to AS-IS
much better
15 Exploitation of image effects
3
much worse
similar to AS-IS
much better
16 Access to supplier / raw material
2
much worse
similar to AS-IS
much better
17 Access to best cost labour
2
much worse
similar to AS-IS
much better
18 Access to skilled labour
4
much worse
similar to AS-IS
much better
19 Access to external know-how
1
much worse
similar to AS-IS
much better
20 Realisation of economies of scale
4
much worse
similar to AS-IS
much better
21 Realisation of economies of scope
4
much worse
similar to AS-IS
much better
22 Reduction of duplication besides manufacturing
4
much worse
similar to AS-IS
much better
23 Mobility of products, processes, and personnel
4
much worse
similar to AS-IS
much better
2
much worse
similar to AS-IS
much better
25 Improvement of learning about external factors
1
much worse
similar to AS-IS
much better
26 Improvement of learning about internal factors
5
much worse
similar to AS-IS
much better
1 FIT with current organisational structure FIT with expectations of the financial community / shareholders FIT (synergies) with other internal programs and 3 developments
FIT with network scope
2
FIT with network tactics
FIT with current asset equipment and allocation 9 between plants FIT with current incentive system and plant 10 managers expectations
24
Mobility and flexibility of production volumes & orders
"Customisation"
"Best cost"
-1
AS-IS
1
2
"Speed"
Tab. 24: Chocolate NW: Scenario assessment
50
The evaluation was conducted in a cross-functional workshop with the operations management team of the international department. The selection of the contextual factors and their weights reflect the team’s common opinion. The selection and the weights of the network capabilities display the previously discussed capability profile of the Chocolate NW. The evaluation of the scenarios was done in relation to each other and with respect to the AS-IS state. Two scores were calculated per scenario for both dimensions of FIT.
168
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
Given a promising definition of the manufacturing strategy and its proper translation into network capabilities, the scores of the “best cost” and the “speed” scenario clearly dominated the AS-IS state and its amplification by the “customisation” scenario. Since the tension line approach facilitates the creation of extreme positions, the following vision was derived as consolidation of these two best scoring scenarios. The TO-BE vision strives for margin increase by both leveraging sales and reducing costs. Hence, its configuration partially sacrifices responsiveness by efficiency. With regards to the process competencies, this affects the centralisation of strategic supply chain activities, a pooling of the R&D and support functions, and of engineers; operational supply chain management remains decentralised. The cocoa liquor manufacturing (level 1) is kept at the two selected sites, a bundling of the chocolate mass production (level 2) requires further investigation, level 3-5 processes remain dedicated. The necessity for customisation is met by local packaging and tailored displays (level 6). Further, the centralisation of the development activities shall better control for internal product variety. The scenario also calls for a realigned product allocation with respect to economies of scale and scope; the production of local specialities will be reduced but kept dedicated, for high running products a dual source strategy is aspired which moves selected sites towards global product plants, hence enabling load balancing for some product lines. A functional organisation is established which runs the sites as cost centres and is responsible for the network’s coordination and global optimisation. Consequently, a high degree of parental control with strong central authority and standardisation for systems, decisions, and strategic processes is targeted. The network coordination has to foster the embeddedness of the sites and a cooperative culture to achieve common network goals, especially for flexible market coverage. A subtle degree of competition is aspired only for operational areas, utilising an internal performance benchmarking. Thus, operational performance needs to be made transparent, and the incentive system has to be aligned accordingly. More general, information transparency on planning and administrative data, knowledge sharing, and best practice exchange is facilitated, but as for the “speed” scenario, the exchange remains centralised to maintain process discipline. 4.2.3.3 Conceptual Network (Re-)Design The definition of a common vision for the targeted TO-BE scenario clarified the direction for the subsequent conceptual network (re)design. As shown in the Elevator NW case, this phase can be structured similar to the analysis, i.e., along the PARTS elements of the network architecture and supported by the management frameworks. Due to the initial intention of this case study and the methodical analogy to the Elevator NW, a closer description of this phase is omitted.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
169
4.3 Summary & Discussion 4.3.1 Findings from the Application of the Management Architecture The case studies confirm the practical utility of the network management architecture. Complementary to the isolated validation of the management frameworks, it was now shown how these can be linked as bricks for a holistic approach. Two detailed examples were given how the architecture was applied as descriptive model to (1) analyse and map a manufacturing network, to (2) define the pillars for its future state, and as a tool to (3) support its (re-)design. For this, the architecture serves as discursive anchor raising the main questions and decisions for the conceptual network development. It facilitates the network management to derive a common understanding of the AS-IS situation and to lead a joint discussion towards the TO-BE state; the “tangible” illustration by the single frameworks allows for an integration of the relevant stakeholders. In this context, the architecture focuses on the “Whats” to address and not yet on the “Hows” in terms of a procedural design approach. As to these “Whats”, the main features of the architecture will be summarised subsequently. The summary is led along a set of distinct evaluation characteristics. Since literature from many disciplines is rich in providing various criteria for the assessment of models, frameworks, and artefacts, this step follows the selection of Friedli (2000). As depicted in Fig. 48, he suggests a collection of six main criteria for the evaluation of his own architecture for designing intercompany cooperation. The selection reflects the condensation of inputs from different sources in the field of enterprise modelling, among those Fox et al. (1993), Vernadat (1996), Weston (1999), and the ISO/DIS 15704 (ISO, 1999). Architecture
Holistic model
• Applicability • Consistency
• Completeness
Competency
• Identification of object and mission
Efficacy
• Clarity
• Transformability • Separation between process structure & -content
(Re-)usability
• Genericity • Accuracy
• Extendibility
Conformity
• Evolutionary approach
• Separation of behaviour and functionality • Modularity • Scalability
Fig. 48: Evaluation criteria (adapted from Friedli (2000))
Complexity handling
• Separation of areas
• Functional decomposition • Visualisation
170
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
The utilisation of his collection is justified by the similarity in the research focus and research stream, and newer contributions do not significantly add to or alter this sample. The design science community, for instance, proposes the evaluation of artefacts according to three main dimensions: utility, quality, and efficacy (e.g., Hevner et al., 2004; Vom Brocke and Buddendick, 2006). Hevner et al. (2004) break these dimensions into “… functionality, completeness, consistency, accuracy, performance, reliability, usability, fit with the organization …” but also narrows their exclusiveness by pointing out that “… also other relevant quality attribute” might be applicable (Hevner et al., 2004, p. 85). In fact, the characteristics cited can be directly linked with Friedli’s (2000) selection. The following evaluation itself is observational (Hevner et al., 2004), based on previous in-depth studies of the architecture in detailed case settings and the application of the single frameworks in multiple projects: • Holism, with respect to the generalisability of the architecture and its applicability for different intra-company manufacturing networks, is constituted by the range of network types influencing the design and serving as objects for validation. They differ in their global dispersion, industry type, size, multiplant strategy, products, as well as in the core processes and technologies. The Elevator NW case even demonstrated applicability for inter-company networks by integrating the external supplier base. Consistency of the derived network designs is assured by the discursive nature of the approach; any inconsistencies are supposed to be clarified within a joint discussion. Completeness can never be fully guaranteed, but it is striven for by including the most important decision dimensions for the network configuration and coordination derived from literature and by mirroring and complementing them in field-discussions. • Competency, in terms of the creation of an integral approach that supports the design and management of intra-company manufacturing networks from the network level perspective, is demonstrated on several dimensions: starting with the systematic identification of the object including the isolation of the network players and the analysis of their added value, continuing with the definition of the network tactics and the formulation of a common vision / mission for the TO-BE scenario, and ending in underpinning the conceptual network (re-)design. A separation between process structure and content is not explicitly fulfilled. Although the application of the architecture implicitly provides guidance for a (de-)composition of the manufacturing network, the focus has been on the content, and a predefined procedural approach has not yet been provided: Neither are single process steps explicitly defined, nor are their content or the underlying roles, inputs, and outputs fully operationalised.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
171
• Efficacy is achieved by providing clarity for a common understanding of, and defined language for, a manufacturing network’s architecture, its layers, and underlying decision dimensions. Thereby, particularly the integration and elaboration of the coordination layer is novel. Transformability on different design problems and preconditions can be achieved on several dimensions. First, by tailoring the elements on the single framework’s level, e.g., by adding distinct responsibility, resource, performance, and / or information & knowledge categories, second, by emphasising elements of the architecture, as shown in the Chocolate NW case with the formulation of the network vision, and third, by skipping elements, as done with the incentive system framework in the Elevator NW case. Owed to the discursive nature of the approach, efficacy is proved with respect to guiding the conceptual design of potential scenarios – but not with respect to the prescription of an optimal architecture subject to contextual preconditions, not with respect to concrete measures, or to a detailed performance evaluation. • (Re-)usability is guaranteed by the architecture’s modularity, splitting it into distinct elements, i.e., layers, decision dimensions, and variables. Each decision dimension is supported by a predefined but customisable management framework. The genericity and repetitive applicability of the frameworks was shown. Analogous, the scalability of the frameworks and of the architecture itself was demonstrated on various network sizes. Extendibility, in terms of a possible integration of new or other concepts, is given; in particular, by pointing out interfaces to those elements not yet fully operationalised (like the interplay with the site level). However, the boundaries between functionality and organisational behaviour are not clearly separated. While the architecture and its elements are independent of organisational idiosyncrasies, its application, e.g., for the formulation of the network capabilities and the conceptual rearrangement of the network design, is biased by historical and contextual dependencies. Accuracy, in turn, is assured again by the discursive nature of the approach: Results always represent a common understanding of the future network design; hence, it is not the task of the architecture but the responsibility of the participants to clarify potential redundancies and ambiguity. • Conformity, as a model’s or artefact’s conformance with the intended object to be displayed, can be argued similar to completeness. The iterative design process for the single frameworks combined literature research with findings from practical discussions. The dynamics and evolutionary character of any design approach is accounted for by the illustrative and “playful” approach; this makes any changes in the decision dimensions easily to go through.
172
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
• Complexity handling and reduction are achieved again by the modular design of the architecture and its functional decomposition: structuring its elements along the PARTS method, distinguishing between the strategy, configuration, and coordination layer, and operationalising their main decision dimensions. Thereby, the demand for a separation of areas is met by distinguishing between the configuration and coordination decisions on network level and the structural and infrastructural decisions on site level. The transformation into illustrative management frameworks meets the requirements regarding visualisation and provides an integrative character. The evaluation of the architecture basically confirms the adequacy of the “Whats” but also shifts the scope to the “Hows”. Although the construction of a strict and sequential network design process with delimited steps is considered illusory due to the complexity and ambiguity of the corresponding tasks, a basic structure for a guided discussion turned out to be helpful in practical usage. It has already been outlined by the architecture’s black edging – the “striving for FIT” – but is not yet elaborated in terms of strategic design and management approaches as introduced in Section 2.4.2. 4.3.2 Implications for a Strategic Design & Management Approach Implicitly, a procedural character became obvious when applying the architecture in the two case studies. The Elevator NW case gave insights on how to structure the network analysis and pass through the conceptual redesign phase. This redesign followed a predetermined network vision as a reaction on predefined modifications in the network architecture, i.e., a consolidation of the players and the implementation of a lead factory strategy. With respect to the contextual changes and their impact on the targeted network capabilities, these directives became necessary to sustain the position in the network capability level and conformance evaluation as shown in Fig. 49. “Striving for FIT” in this context meant to “sustain FIT”. The Chocolate NW had not yet actually tried to leverage its network potential. Its position in Fig. 49 revealed a network capability and conformance level far below the survey average. This position, as initially derived from the survey results, was confirmed by the operations management team. It is owed to the limited degree of network capabilities and the gaps between the AS-IS and TO-BE profile in Fig. 47. Hence, before starting with the conceptual network design, a joint understanding of the tactics and the formulation of a mutual vision for a TO-BE scenario supporting these tactics became necessary. “Striving for FIT” in this context meant first of all to understand and “achieve FIT”.
FROM THE COORDINATION LAYER TO A MANAGEMENT ARCHITECTURE
Network capability conformance as deviation from mean
Overall network capability level & conformance
173
Overall network capability level as deviation from mean
Low capability level / High conformance
High capability level / High conformance
II
I
“Achieving FIT"
IV
?
Low capability level / Low conformance
!
! ?
"Sustaining FIT"
III
? High capability level / ! Low conformance
= Cause = Problem = Target = Solution
Fig. 49: Aspects of striving for FIT
A procedural approach has to integrate both aspects of “striving for FIT”. It has to provide a structured methodology to achieve a fitting TO-BE state, i.e., by designing the network from scratch according to an agreed direction. It, on the other hand, must be able to sustain that steady state subject to internal or external contextual changes. The network architecture and the management frameworks support the descriptive analysis and conceptual (re-)design phases. The tension line approach for sketching and evaluating possible TO-BE scenarios, in turn, proved value in bridging the gap between both phases by deriving a common understanding of the aspired future direction. Both concepts will be seized in the following chapter, condensing the various practical experiences to a “suggested practice approach” for strategic network design and management.
174
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
5 From a Management Architecture to a Strategic Design & Management Approach This chapter is finally dedicated to the practitioners facing the challenge to (re-)design their manufacturing network. Findings, experiences, and impressions from literature and field-work are condensed to transform the management architecture into a “suggested practice approach” (Schmid, 2011) for strategic network design and management. The promotion of such approach is motivated by the need to guide operations managers through the stages of analysing, developing, and improving their network. Its nature thereby accounts for the ambiguity of the task: Instead of promoting a “conclusively strict corset” with rigid process steps, the approach is understood as discursive frame open for creative work and iterative loops, sensitising for and stimulating the core decisions of strategic network (re-)design. Section 5.1 introduces the “suggested practice approach” and its distinct stages. Section 5.2 positions the approach against the methods as derived from the literature analysis in Section 2.4.2. Pointing at some open questions left to be addressed, it also bridges the gap to this study’s end.
5.1 Presenting a Strategic Network Design & Management Approach A detailed theoretical discussion and repetition of the existing literature on strategic network design and management processes and methods is omitted here; it has already been given in Section 2.4.2. Instead, the following presentation of the approach occupies the practitioners’ perspective mainly. The “suggested practice approach” is sketched in Fig. 50. It is built around the introduced network management architecture. The overall goal of the approach is formulated as “Striving for FIT”, thereby reflecting both the achievement and the sustainment of the network FIT. Five main stages with distinct steps each are comprised. Stages in white have explicitly been addressed in this study; stages shaded in grey require further elaboration. For the latter, only basic recommendations are given as summary of the impressions from field-work.
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
Fig. 50: “Suggested practice approach” for strategic network design & management
175
176
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
Stage 1: Network Analysis & Target Setting The initial stage targets a detailed analysis constituting a common understanding of the network’s AS-IS state. It is conducted along the elements of the network’s PARTS architecture and starts with the identification of the players and their added value. Both elements are tackled when elaborating on the configuration layer. First, grasping the structure is necessary to isolate the network’s players; a step particularly crucial for larger companies with different product groups to segment possible networks. Getting an overview of the sites’ geographical dispersion, their product portfolio, and volumes manufactured, their market allocation, processes performed, and technology applied, as well as about the physical linkages and material flows is a precondition to separate between different value chains in a company and to draw the distinct networks’ system boundaries. Second, enhancing the structure by the network specialisation means to concretise the added value of each player. A stepwise approach was demonstrated in the Elevator NW and the Profile NW cases how to proceed for the evaluation of the site competencies and strategic site reasons in order to design the plant role portfolio. Complementary to the configuration layer is the analysis of the network rules; it integrates the organisational structure and the network coordination dimensions. From a network perspective, it can be conducted along the coordination frameworks which shed light on the centralisation & standardisation, the resource allocation & sharing, the incentive system, and the information & knowledge sharing. The internal view on the network architecture, as derived so far, is condensed by a sketch of the current network capability profile as shown in the Chocolate NW case. Revealing the AS-IS profile as summary of the network tactics proved to be valuable for several reasons: First, the management’s probably biased perception of the capabilities can be contrasted with the findings from the AS-IS analysis indicating strengths, weaknesses, and gaps in the current network configuration and coordination. Moreover, a mutual understanding of the actual network capabilities serves as common denominator for the future network design. Yet, defining the pillars for the TO-BE state requires more than just a look at the internal capabilities. Following the established opinion in operations research, strategy has to match a company`s internal capabilities with the external (market) requirements (e.g., Hax and Majluf, 1995; Slack and Lewis, 2002). Thus, as outlined in the management architecture, manufacturing strategy is expected to align the network (and site) capabilities with the company’s contextual requirements. Consequently, shaping the network capabilities is directly influenced by the manufacturing strategy itself, but it is also framed by the contextual environment that makes a deeper understanding of the market situation, competitors, mega trends, and current developments necessary.
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
177
Similarly, internal contextual boundaries and limitations stemming from the products, processes, and technology should also be taken into consideration. In both of the indepth case studies, analysing the network scope was presented only with regards to the link between the network capabilities and manufacturing strategy. Nonetheless, existing literature is rich in providing tools for a structured assessment of the company’s opportunities and threads; among those the well-known “Porter’s five forces” (Porter, 1979) and the original PARTS analysis (Brandenburger and Nalebuff, 1995) when it comes to the evaluation of the strategic environment, or roadmapping techniques, as proposed by Christodoulou et al. (2007) in the “Cambridge approach”, when it comes to the analysis of products and technology. Discussing external and internal forces also adds dynamics to a system looked at. For the analysis of the network’s AS-IS situation, this means not only to statically map the current state but also to anticipate and project the latest changes and forecasts. Finally, based on the results of the AS-IS mapping, the joint perception of the network capabilities, the deeper understanding of the manufacturing strategy, and the contextual framing, the capability targets for the TO-BE state can be defined as shown in the Elevator NW and the Chocolate NW case. Stage 2: Scenario Development Shaping the network capabilities is realised by adjusting the configuration and coordination decision dimensions. When doing so, initially formulating a concrete vision of the targeted network scenario, or at least defining its main pillars, is valuable. In the Chocolate NW case, it was demonstrated how such vision for a TO-BE scenario can be developed systematically. For this, the tension line approach proved to be a helpful tool as “mediator” between the analysis stage and the subsequent conceptual (re-)design. It comprises the (1) formulation of themes for possible future scenarios, their (2) definition along the tension lines, the (3) evaluation of the scenarios, and the (4) development of the TO-BE scenario as network vision. • Themes can usually be derived with a closer look at the manufacturing strategy. Formulating antithetic and concise themes that address the mostly conflicting goals of the aspired manufacturing strategy is useful to capture the full range of possible directions. Further, also the AS-IS state should be considered. • Elaborating the themes is done by defining the scenarios along the tension lines; these reflect the “essence” of the configuration and coordination decision dimensions. The procedure allows for a target-oriented discussion ending in drawing a unique profile for each potential scenario. Thereby, starting with the AS-IS scenario and positioning the others against it turned out as helpful.
178
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
Basically, the collection of tension lines as depicted in Tab. 23 can be applied for most companies with only a slight modification and renaming necessary. • For the scenario evaluation, different methods are conceivable. In the Chocolate NW case, a simple assessment along a set of weighted criteria was conducted. 51 Each scenario was evaluated in relation to the others and with respect to the AS-IS state. The procedure could be carried out easily in a workshop setting and allows for a qualitative estimation rather than an “indisputable” quantification. Generally, criteria for assessment are manifold; in any case, they should be distinctive and clearly formulated, reflecting the common understanding of the participants. With respect to contingency theory, they should enable to judge the FIT with the network tactics, by integrating the capability targets, and the FIT with the network scope, by integrating criteria from the contextual environment. Explicitly considering the manufacturing strategy might bias the results since it is supposed to be already aligned with and reflected by the targeted network capabilities. • The second stage is completed by the development of the TO-BE scenario. Accounting for the network’s “historical heritage” and the usual trade-offs in the manufacturing strategy, this is normally a blending of the best scoring scenarios from the evaluation step and the AS-IS state, rather than a pure form. In the end, explicitly formulating the TO-BE scenario as vision provides guidance for the subsequent network design, it demonstrates the commitment of the decision makers, and it also serves as mission statement for the communication of the project. Stage 3: Rough Cut Network Design The realisation of the aspired TO-BE scenario progresses from a “rough cut” to a detailed network design; it starts with the conceptual modification of the network architecture as content of the third stage. As shown in the Elevator NW case, this very stage follows a similar approach than for the analysis: passing through the distinct frameworks and sketching the intended changes. Working with the frameworks turned out as discursive and integrative way to transform the TO-BE vision into a consistent TO-BE concept; their illustrative and “playful” character makes any changes easily to rethink. However, since the conversion of the TO-BE scenario into the management frameworks is subject to multiple alignments between the underlying decision dimensions and variables, the procedure should be highly iterative, avoiding decisions in one dimension to prejudice decisions in others. Owed to such risk of a strictly 51
Various other approaches can be applied for weighting the criteria and evaluating the scenarios. In any case, their choice and the effort for their execution should match with the depth and utility of the results obtained.
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
179
successive approach, any recommendations regarding the proceeding shall not be understood as prescription, leaving room for iterative loops instead. Given the network capability targets and visions, the rough cut design begins best with the configuration layer, more precisely, with the adjustment of the network players and their added value in the plant role portfolio, and it should continue with the subsequent alignment of the rules along the coordination frameworks. The plant role portfolio allows concretising the future network structure and specialisation in terms of the number of sites needed, their roles, competencies, and their contribution to the network. As shown by Fig. 33 for the Elevator NW, it might be useful to lay down the requirements and expectations for a generic plant role, or even a concrete site, by a strategic site profile. Designing the rules, then, switches between the alignment of the coordination dimensions and the adoption of the formal organisational structure. It can be initiated by defining the intended centralisation and standardisation and the resource allocation and sharing strategy, followed by the design of the information & knowledge flows and the alignment of the incentive system. The organisational structure manifests the changes in configuration and coordination. Finally, to avoid any prejudication, with the conceptual changes in all dimensions at hand, a consistency check should be conducted where the frameworks are reviewed altogether. The stage ends with a concept for the TO-BE scenario and general implications for its detailed design. Stage 4: Detailed Network & Site Design The second stage of the design phase has not been addressed explicitly in this study. It is basically about the refinement of the TO-BE concept, its implementation, and about the alignment with the site level. Therefore, the stage could be split into two parts. On the network level, it now comes to a detailed quantitative elaboration of the TO-BE architecture and its validation by “hard facts”. In this step, the boundaries between the rough cut concept and the detailed TO-BE architecture become blurred. Validating the architecture should integrate monetary aspects to estimate the potential of the new design and its return on invest, but it should also elucidate impacts on the other strategic manufacturing priorities, e.g., on quality, delivery performance, etc. Quantitative (optimisation) methods, like mathematical programming or simulation, can be beneficial for this detailed evaluation, especially to substantiate time and cost impacts. Some of them also enable a testing for robustness of the network design by integrating uncertainties when modelling the internal and external environment. However, such approaches are concentrated to changes in the network configuration mainly. Changes in the coordination layer, in turn, tackle modifications of the
180
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
organisational structure; they probably allow for a reduction of bureaucracy, a streamlining of administrative interfaces, or an acceleration of business process. Hence, they require other methods for quantification. Further, the impact of synergies by resource bundling or the benefit of information and knowledge exchange should not be left aside. In the end, the quantitative validation of the TO-BE architecture supports the formulation of concrete actions, and project management has to assure their implementation. Moreover, redesigning the network level induces changes on the site level, e.g., when altering a plant role, when modifying the sites’ degree of autonomy, etc. And also the manufacturing strategy is not solely supported by network capabilities but by the site capabilities, too (Colotla et al., 2003). Thus, the challenge of translating the manufacturing strategy into site capabilities, aligning the site capabilities with the network level, and shaping the sites’ structural and infrastructural decision dimensions triggers a similar approach than for the strategic network design. In terms of a delimited (re-)design / restructuring project, the approach ends here; it closes the loop by a realigned TO-BE architecture. Stage 5: Institutionalisation of Network Management As argued, passing through the stages of the illustrated approach supports a systematic network (re-)design from scratch, i.e., “achieving FIT”. Yet, considering a manufacturing network as “moving target” subject to contextual dynamics also requires to “sustain FIT” over time. Operations managers have to be empowered both to act proactively and react quickly on any external and internal changes. Correspondingly, Christodoulou et al. (2007) call for “… a repeatable long-term process (…) that needs to be fully integrated in the business planning cycle and needs to be the definitive basis for all manufacturing network decisions” (Christodoulou et al., 2007, p. 4 and p. 40). Capturing their claim, analogous to the evolution of the manufacturing management on site level, the network management needs to be institutionalised and professionalised. Rather than as a delimited restructuring project, it has to be understood as a replicable business process embedded into the company’s organisation (Christodoulou et al., 2007). Such process synthesises inputs from other global functions, challenges and refines the current network vision and mission, and works on its implementation. Moreover, appropriate tools and KPIs need to be defined, allowing not just to evaluate the one-time success of a (re-)design project but to track the network performance on a regular basis and to initiate actions in case of deviations. In the end, facilitating a continuous improvement philosophy to evolve from the site to the network level assures on-going incremental optimisation.
How
Where
What
Why
A
Deriving footprint options
7
Colotla (2002) and (2003)
2 Formulating global mfg. strategy and scenarios
1b Analysing network and factory capabilities
1a Analysing business strategy & scanning the environment
B
Strategic approaches
Tab. 25: Comparison of the “suggested practice approach” (part I)
14 Embedding the new process
12 Transition Implementation 13 Measuring the success
4 Detailed programming and budgeting
4 Transforming network config.
3 Identifying mfg. mission and designing network config.
2b Assessing current network configuration 2c Analysing network capabilities
0 Introducting international mfg. network and strategy
1 Identifing requir. of globalis. 2a Analysing prod. and markets
Ude (2010), Lanza & Ude (2010) "Karlsruhe approach"
Quantitive approaches
1e
1b
1c
1a
1d
E
Defining capability targets for TO-BE network Scenario development
Analysing the rules
Analysing the tactics
Analysing the players & added values
"Suggested practice approach" for strategic network design and management Analysing the scope
"Suggested practice approach"
4 Testing robustness
5
Institutionalisation of network management
4.2a- Rough cut and detailed site 4.2c design
4.1c Testing robustness of the TO-BE network architecture 4.1d Actions for implementation
2a2d 2 Designing global mfg. 3a- Rough cut network design network options 3c 4.1a Detailing the TO-BE network architecture 3 Evaluating options by simula- 4.1b Validating the TO-BE tion and multi-criteria analy. network architecture
1 Determining multidimensional target system
Shi et al. (1997), Shi & Gregory C (1998), Shi et al. (2001), Shi D (2003)
b) Network design, management & optimisation approaches
Assessing footprint options based on 3a Evaluating scenarios based on cost & network capabilities financial and non fin. Impact Aggregating product or business unit strategies into company vision Testing robustness of the options 3b Evaluating scenarios according based on contextual changes to sustainability and risk Mobilising for change
Analysing & determining coordination principles
6
5
Creating a common framework for analysis (variables & assumptions) Analysing & designing plant roles
Mapping the strategic and environmental context for the network Embracing the change (burning platform, right people, etc.) Defining make-or-buy strategy
Christodoulou et al. (2007) "Cambridge approach"
4
Assess 8 options Aggregate 9 solution Analyse 10 robustness 11
Identify network options
Preconsideration
Make-or- 3 buy analysis
1 Understand underlying 2 motivation
a) Generic steps
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
181
5.2 Summary & Discussion
As a last point, the “suggested practice approach” is positioned against that network design, management, and optimisation approaches discussed in the literature review. Falling back on the structure of the previously used Tab. 7, the first part of the comparison, as given in Tab. 25 (area b), contrasts the distinct stages (and steps) of the different approaches and processes.
Analysing the tactics
Analysing the rules
Defining capability targets for TO-BE network Scenario development
1c
1b
1e
5
Tab. 26: Comparison of the “suggested practice approach” (part II)
Cambridge approach Colotla Shi bzw. Shi & Gregory Karlsruhe approach Suggested practice approach
A B C D E
e) Aggregation of comparison
Institutionalisation of network management
4.2a- Rough cut and detailed site 4.2c design
4.1c Testing robustness of the TO-BE network architecture 4.1d Actions for implementation
2a2d 3a- Rough cut network design 3c 4.1a Detailing the TO-BE network architecture 4.1b Validating the TO-BE network architecture
Analysing the players & added values
1a
o o
1d
Network capab.
Business & mfg. strategy
Markets, techn., environ., compet.
+
++ o + o
o + o + o o
"Suggested practice approach" for strategic network design and management Analysing the scope
++
o
o o + o
+ + +
+ + ++
o o
Site capab.
Network Structure
+
Network Specialisation
+ o +
E
+ +
o
+ o
++ o
+ + + o
++ ++
Flows
Institutional Persp.
Coordination
++
o o
Configuration
c) Elements of manufacturing network (management)
o o
o
++ ++ + + o +
++ + ++ ++ o
++ ++ ++ o
Network Perform.
Perform.
o
o o
+ + o o
PjM, QL,QT
d) Method
QL PM QL QL QL
+ o
Strategy
QL
QL QL QL
+ o
QL QL PM QL/QT
QL QL QL/QT QL QL QL
QL QL QL QL QL QL QL/QT QL
Context
+ o + o o
++ o + o
o
+
+ + ++
o
+ +
+
+ o ++ ++
++
+ o
++
o
++
+ o o
o
o + + +
QL QL QL QT QL
PM = Project management not adressed
"Suggested practice approach"
QT = Quantitative approach covered and explicitly elaborated
+ ++
covered and partially elaborated
QT QT
A B C D E A B C D E A B C D E A B C D E A B C D E A B C D E A B C D E A B C D E A B C D E A B C D E
QL = Qualitative approach
o adressed but not elaborated
182 FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
The second part, as outlined in Tab. 26, evaluates to what extent a distinct step tackles the elements and decision dimensions of manufacturing network management (area c). Moreover, the dominant method underlying the very step (area d) is classified. The aggregated comparison (area e) gives an estimation of the quality and individual contribution in each dimension based on the own perception. For the following comparison, mainly the qualitative approaches are in the focus.
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
183
Glancing at the process steps first, the structure of the “suggested practice approach” reveals some special features. Starting with an in-depth analysis of the network along the PARTS elements allows for creating a sound understanding of the initial network state. Continuing with the development of a common vision for the TO-BE scenario, the conceptual rough cut design of the TO-BE architecture, and its subsequent refinement enables a stepwise concretisation by successively increasing the level of detail in the design phase. Thus, compared with the “Cambridge approach”, more emphasise has been put on the tasks of mapping and conceptually designing the network. On the contrary, less effort has been dedicated to the general analysis of its strategic and contextual scope, which other approaches are considered worthwhile for, to shaping the make-or-buy strategy, and to creating the organisational conditions for a transformation project (e.g., to any project management tasks). When getting deeper into the content of each step, the network architecture and the “suggested practice approach” claim superiority particularly regarding the elaboration of the network specialisation and the coordination layer. With the operationalisation of the network specialisation by a novel concept for the plant role portfolio construction, the “mountain model” of the “Cambridge approach” is enhanced in terms of proved practical applicability. But admittedly, this plant role portfolio limits the sites to the network management’s aggregated perspective solely. Thus, aligning the interplay between the site and network level, breaking up the plants’ microcosm, and linking the network and the site capabilities to support manufacturing strategy have to be further concretised. By operationalising the decision dimensions of the coordination layer using replicable management frameworks, the “suggested practice approach” also fills a second main gap left open so far. Albeit, for modelling the network structure, only a simple approach is presented; other methods known are more sophisticated, providing support for a detailed evaluation of value / volume streams. However, the detailed network design stage, including the quantitative evaluation of the TO-BE architecture, remains basically unaddressed. For this, the “suggested practice approach” provides an interface to quantitative optimisation approaches. It puts the strategic considerations in front and limits the number of options to be quantitatively evaluated to the most promising ones; thus, it reduces the effort of data gathering and handling. Moreover, the need of concepts for a suitable performance evaluation in manufacturing networks becomes emergent. These cannot be realised as aggregation of the KPIs on site level; they also call for a more specific assessment of the network’s capabilities. 52 Being able to measure and monitor the network FIT on an 52
Taking flexibility as an example: Measuring it on site level is based on efficiency, overcapacity, and changeover times; assessing it on the network level also integrates the manufacturing mobility in terms of a flexible order allocation and alternative manufacturing locations.
184
FROM A MANAGEMENT ARCHITECTURE TO A STRATEGIC DESIGN & MANAGEMENT APPROACH
on-going basis is a precondition for an institutionalisation of the network management. For this, the organisation, implementation, execution, and the continuous improvement of the network management as a strategic business process need to be concretised, too. Arguing with the findings from Platts (1994), such process raises questions regarding the procedure and stages, the participation, project management, and point of entry. With regards to these questions, only the procedural character has been tackled so far; but this was done less with respect to a repeatable business process and more in terms of providing guidance for a delimited network restructuring project. Regarding the latter, the analysis phase, the scenario development, and the rough cut design were carried out qualitatively in a series of cross-functional workshops embracing all key decision makers on the network level. Integrating representatives from the site level can be beneficial in terms of a broader foundation of the results, but it finally depends on the balance of power in the network. Discussions in the analysis stage and for the rough cut design were led along the management frameworks; these provide a “playground” to stimulate and discuss strategic options. They were refined iteratively, serving also as backbone for the project documentation. The detailed network analysis stage, then, requires quantitative input from functional specialists (e.g., financial, business planning, controlling).
SUMMARY & OUTLOOK
185
6 Summary & Outlook The last chapter ends the study by resuming its findings and impact. Section 6.1 starts with a critically reflection of its main results and contrasts them with the raised research question. Section 6.2 outlines the study’s contributions to theory and practice while Section 6.3 points at its limitations, providing stimulus for further research.
6.1 Critical Reflection Recalling the research gaps and superordinate motivation in the introductory section, it was argued that today’s operations leaders require a holistic system perspective and management paradigm for understanding, analysing, designing, and optimising their manufacturing network. Accounting for the magnitude of this postulation, the scope of this study was limited to the management’s tasks of analysing and conceptually redesigning intra-company manufacturing networks with respect to the integration of the coordination layer; or as condensed by the guiding research question: Q. 1: How can the strategic coordination of intra-company manufacturing networks be supported systematically and methodically from a network level perspective? Four second-order questions were formulated to predefine and structure the research. In the following, the results of this study will be judged according to their ability to answer these questions. Q. 1.1: What are the main decision dimensions and variables characterising the strategic coordination of intra-company manufacturing networks? Operations management literature on manufacturing networks served as starting point to isolate the central coordination decision dimensions in the first part of Chapter 3. These were: (1) centralisation & standardisation, (2) resource allocation & sharing, (3) information & knowledge sharing, and the (4) incentive system. Combining insights from literature with close field-discussions in a set-up of five case networks, the distinct decision dimensions were concretised by breaking them down into their underlying variables. Hence, a “multi-dimensional decision space” for network coordination was created as outlined by Tab. 18 in Section 3.4.1.
186
SUMMARY & OUTLOOK
Q. 1.2: How can decision making along these dimensions be supported systematically and methodically, elaborating the network coordination layer? Each coordination decision dimension was operationalised by a distinct management framework in the second part of Chapter 3. Empirical data from the five case networks, a cross-industry study, and three on-site interviews with operations managers underpinned their development and guaranteed for a validation in real industry settings. With regards to the requirements of the network management for strategic decision making, the frameworks provide an illustrative and tangible set of tools fostering conceptual thinking and the discursive derivation of strategic options for the design of coordination mechanisms. It was demonstrated how their application supports the mapping of the coordination layer, but also the shaping of the coordination decision dimensions systematically and methodically. Moreover, the completeness of the dimensions and the accuracy of the logic behind the frameworks were indicated by contrasting findings with the organisation theory’s perspective on coordination. Hence, the “multi-dimensional decision space” for network coordination was transformed into a “multi-framework based toolbox”. Q. 1.3: What are contextual factors to be considered for decision making along these dimensions, especially with regards to the network configuration? When targeting a holistic network management approach, not only the operationalisation of the coordination layer but also a profound understanding of the forces driving its design became necessary. Amongst others, especially the interplay between configuration and coordination lacks deeper insights. Within the course of this study, considering contextual linkages was concentrated on these two layers’ relationship mainly. Therefore, influences of the network organisation, the network structure, and the network specialisation on the coordination decision dimensions were highlighted. Similar to the design of the coordination frameworks and with respect to existing concepts in operations management research, these dimensions were operationalised in the first part of Chapter 4. This accounted particularly for the network specialisation by providing a novel approach for the construction of a plant role portfolio. Hence, the “multi-framework based toolbox” was linked with the decision dimensions of the configuration layer.
SUMMARY & OUTLOOK
187
Q. 1.4: How can decision making support be integrated into a generic network management architecture and design approach? The “multi-framework based toolbox” defined the bricks for a generic network management architecture. Enhancing the heuristic research framework from the initial literature analysis, a model for such architecture was promoted by Fig. 30 in Section 4.1.2. For practical usage mainly, it was presented along the five elements of the PARTS analysis developed by Brandenburger and Nalebuff (1995), but tailored to the specialities of manufacturing networks. In terms of the PARTS analysis, the network is understood as a system comprising individual sites (players) with each having a certain role and providing distinct competencies (added value). Coordination mechanisms (rules) need to be set in place to organise the interaction between the players and with the superordinate network management. Modifying the players, their added value, and the rules, i.e., the configuration and coordination decision dimensions, is necessary to shape the network capabilities (tactics) subject to the manufacturing strategy and the internal and external contextual environment and preconditions (scope). Two in-depth case studies in the second part of Chapter 4 demonstrated the generic applicability and utility of the architecture for the analysis and conceptual network (re-)design, and its assessment along a set of accepted evaluation characteristics strengthened the belief in its adequacy. Finally, findings, experiences, and impressions from these two in-depth case studies in particular, and from the other field-work in general, were synthesised in Chapter 5, transforming the management architecture into a methodical approach for strategic network design and management. Rather than as rigid step-wise process, this approach has to be understood as discursive frame for network analysis and conceptual development by raising the operations managers’ awareness for the core decisions. Overall, the “multi-framework based toolbox” was anchored in a holistic network architecture as core of a “suggested practice approach” that supports the strategic network design and management.
6.2 Contribution to Theory & Practice With regards to this study’s foundation in Ulrich’s (1984) postulation of business administration as applied science to development normative models for designing and changing the social reality, i.e., for solving real industry problems, the contribution to theory and practice cannot be clearly separated. Consequently, answering the research question had to integrate both perspectives. Hence, the assessment of the main results regarding their contribution to theory and practice given in Tab. 27 does not reflect an explicit segmentation between both stakeholders but is overlapping instead.
188
SUMMARY & OUTLOOK
Contribution to theory
Results 1
“Multi-dimensional decision space” for network coordination
2
“Multi-framework based toolbox” for network coordination
3
Linkage to the configuration layer
4
Holistic network management architecture
5
“Suggested practice approach” for strategic network design and management
High
Medium
Contribution to practice
Low
Tab. 27: Results & their contribution to theory & practice
• “Multi-dimensional decision space” for network coordination: By breaking coordination into its single decision dimensions and variables, the multidimensional decision space constitutes a research arena to understand, model, and study coordination aspects. Since prior approaches neglected or failed in defining the nature and content of coordination from a network management perspective, this contribution is of high theoretical relevance; it provides a sound foundation of coordination in manufacturing networks. • “Multi-framework based toolbox” for network coordination: By transforming the decision dimensions into management frameworks, practitioners are provided with a toolbox to analyse, understand, visualise, discuss, and develop the network’s coordination layer. The frameworks give deeper insights into the logic and causalities of coordination mechanisms in the context of manufacturing networks. For researchers, their explorative development provides spacious “playground” for in-depth hypothesis formulation and quantitative explanation on the single dimensions and their interrelations. • Linkage to the configuration layer: To the best of the own knowledge, little systematic research is known having addressed the linkages between the configuration and coordination layer yet. This study contributes to this gap by linking the network organisation, the network structure, and especially the network specialisation with coordination. Regarding the latter, the novel elaboration of the plant role portfolio to systematically create and keep track of plant roles and the demonstration of its alignment with the coordination frameworks adds value to practitioners but also to theory. • Holistic network management architecture: A holistic network management architecture is presented that anchors the single frameworks and provides an
SUMMARY & OUTLOOK
189
integral view on manufacturing networks to both practice and research. It is constructed from the operations manager’s perspective, thereby explicitly separating between the network and the site level. Existing models are enhanced by its integral nature, the operationalisation of its elements, its generic validity, as well as by its proved applicability. Further, the condensed depiction of the architecture by tension lines reflecting the “essence” of the management frameworks might allow for a quantitative construction of network typologies and the derivation of taxonomies. • “Suggested practice approach” for strategic network design and management: The practical contribution of the architecture is enhanced by its transformation into a strategic network design and management approach. Rather than a restrictive process with prescriptive steps, operations managers are equipped with a discursive approach focusing on the main elements and decisions to make when conceptually (re-)designing their network. Existing processes are extended by the approach’s main features: (1) by the successively increasing depth of detail in network design, (2) by the elaboration of the coordination layer, and (3) by the foundation of the network specialisation, suggesting a novel operationalisation of the plant role portfolio.
6.3 Limitations & Further Research The study is completed by having a closer look at its limitations. These, in turn, outline indications for further research. The following discussion is two-fold: It starts with the closer limitations stemming from the way the research question was tackled and ends with the wider limitations coming from the study’s superordinate motivation. First, regarding the closer limitations: • A qualitative architecture (and design approach) is provided embracing and integrating the central elements of manufacturing network management. Its exploratory character so far calls for additional quantitative manifestation, thus defining a sound starting point for hypothesis testing. This accounts for the assumptions reflected by the logic of the single coordination frameworks, for a more substantiated examination of the linkages between the distinct decision dimensions, and for the influence of contextual factors. • Likewise, an explicit description of a “fitting network” is left open. With the central decision dimensions isolated, a quantitative derivation of typologies and a foundation of promising network taxonomies would be beneficial. Finding statistical evidence to conclude from the manufacturing strategy and contextual forces to the “ideal” design of the network configuration, and from the
190
SUMMARY & OUTLOOK
configuration to an “ideal” design of the coordination dimensions respectively could empower the design approach. With respect to contingency theory, such evaluation requires a closer examination of contingency factors and the consideration of the network performance as measure to judge the FIT of potential taxonomies. • Moreover, the research methodology applied is based on a static observation of manufacturing networks that relies on multiple case studies. The restricted time horizon for this study and the companies’ gradualness in implementing changes did not allow for a longitudinal study, thus permitting repeated observations. Hence, the actual impact of implementing a derived concept in general, or modifying a certain coordination dimension in specific, was not observable in a real industry setting. Accordingly, although dynamics were recommended to be anticipated for the analysis and target setting stage of the “suggested practice approach”, it was not further elucidated how mangers (should) actually respond to them in practice. • Finally, the impact of cultural aspects was not integrated explicitly. Shaping the corporate culture can be a coordination target itself. Cultural differences between the scattered manufacturing locations, on the other hand, can be understood as contextual drivers influencing the design of the coordination mechanisms. In particular, further research should focus on the cultural imbalance in a network and how coordination measures can be used for evening or exploiting it. Second, regarding the wider limitations: • More research should urgently be dedicated to the question of performance evaluation in manufacturing networks; 53 not only to support the derivation of taxonomies. Instead, knowing the current network performance, being able to monitor deviations, and quickly initiate counter-actions becomes a prerequisite for the on-going network management. As argued before, such measurement cannot be realised as aggregation of the site KPIs only, but it calls for a more specific set of indicators tailored to the network’s idiosyncrasies instead. • Additional research should put effort on a deeper understanding of the interplay between the network and the site level. It can be argued that competitive advantage from manufacturing superiority stems from an exploitation of the 53
Certain measures were defined for the cross-industry survey to calculate the overall performance on strategic manufacturing priorities and the overall network capability level as outlined in Appendix B.2 and B.3. These provide a starting point, but require closer examination and enhancement to be suitable for a KPI system.
SUMMARY & OUTLOOK
191
network’s and the sites’ potential. Considering both as interlinked parts of a system requires an alignment of their capabilities in order to support manufacturing strategy best. • Overall, the need for a professionalisation of the network management in general becomes necessary. Little is known about the implementation of the network management; neither in industry nor in practice. Independent research, in turn, allows for identifying “success stories” that might serve as guidelines and benchmarks for the institutionalisation of the network management and its foundation as business process.
Galbraith, 1977; Fine and Hax, 1985; Mefford, 1986; Hayes et al., 1988; van de Ven, 1989; Sugiura, 1990; Baden-Fuller and Stopford, 1991; Samson, 1991; Vos, 1991; DuBois et al., 1993; Kogut and Kulatilaka, 1994; Sweeney, 1994; Miltenburg, 1995; Bolisani and Scarso, 1996; Skinner, 1996; Dasu and de la Torre, 1997; Ferdows, 1997b; Lee and Lau, 1999; Bartlett and Ghoshal, 2000; Hill, 2000; Prasad and Babbar, 2000; Mauri and Sambharya, 2001; Prasad et al., 2001; Ferdows, 2003; MacCarthy and Atthirawong, 2003; Ferdows et al., 2004; Ketokivi and Jokinen, 2006; Noori and Lee, 2006; Bonaglia et al., 2007; Karlsson and Sköld, 2007; Riis et al., 2007; Abele et al., 2008; Coe et al., 2008; Feldmann et al., 2009; Gray et al., 2009; Mauri, 2009; Vecchi and Brennan, 2009; Wright et al., 2009
192
REFERENCES
References Abele, E., T. Meyer, U. Näher, G. Strube and R. Sykes (Eds.) (2008): "Global Production: A Handbook for Strategy and Implementation", Ed. 1, Springer, Berlin and Heidelberg. Aikens, C. H. (1985): "Facility location models for distribution planning", European Journal of Operational Research, Vol. 22 (3), pp. 263–279. Baden-Fuller, C. W. F. and J. M. Stopford (1991): "Globalization frustrated: the case of white goods", Strategic Management Journal, Vol. 12 (7), pp. 493-507. Bartlett, C. A. and S. Ghoshal (1989): "Managing Across Borders: The Transnational Solution", Ed. 1, Harvard Business School Press, Boston (MA). Bartlett, C. A. and S. Ghoshal (2000): "Going global: lessons from late movers", Harvard Business Review, Vol. 78 (2), pp. 132-142. Bartmess, A. and K. Cerny (1993): "Building competitive advantage through a global network of capabilities", California Management Review, Vol. 35 (2), pp. 78-103. Bartol, K. M. and A. Srivastava (2002): "Encouraging knowledge sharing: the role of organizational reward systems", Journal of Leadership and Organizational Studies, Vol. 9 (1), pp. 64-76. Bhatnagar, R. and P. Chandra (1993): "Models for multi-plant coordination", European Journal of Operational Research, Vol. 67 (2), pp. 141-160. Birkinshaw, J. (2001): "Strategies for managing internal competition", California Management Review, Vol. 44 (1), pp. 21-38. Bolisani, E. and E. Scarso (1996): "International manufacturing strategies: experiences from the clothing industry", International Journal of Operations & Production Management, Vol. 16 (11), pp. 71-84. Bonaglia, F., A. Goldstein and J. A. Mathews (2007): "Accelerated internationalization by emerging markets' multinationals: the case of the white goods sector", Journal of World Business, Vol. 42 (4), pp. 369-383.
REFERENCES
193
Brandenburger, A. M. and B. J. Nalebuff (1995): "The right game: use game theory to shape strategy", Harvard Business Review, Vol. 73 (4), pp. 57-71. Brandenburger, A. M. and B. J. Nalebuff (1996): "Co-opetition", Ed. 1, Currency Doubleday, New York (NY) et al. Canel, C. and B. M. Khumawala (1996): "A mixed-integer programming approach for the international facilities location problem", International Journal of Operations & Production Management, Vol. 16 (4), pp. 49–68. Canel, C. and B. M. Khumawala (2001): "International facilities location: a heuristic procedure for the dynamic uncapacitated problem", International Journal of Production Research, Vol. 39 (17), pp. 3975-4000. Cerrato, D. (2006): "The multinational enterprise as an internal market system", International Business Review, Vol. 15 (3), pp. 253-277. Cheng, Y., S. Farooq and J. Johansen (2011): "Manufacturing network evolution: a manufacturing plant perspective", International Journal of Operations & Production Management, Vol. 31 (12), pp. 1311-1331. Cheng, Y., J. Johansen and H. Boer (2008): "Coordinating knowledge transfer within manufacturing networks", 9th International CINet Conference, September 5-9, Valencia, Spain, pp. 173-183. Chew, W. B., T. F. Bresnahan and K. B. Clark (1990): "Measurement, coordination, and learning in a multiplant network", in: R. S. Kaplan (Ed.): Measures for Manufacturing Excellence, Ed. 1, Harvard Business School Press, Boston (MA), pp. 129-162. Christodoulou, P., D. Fleet, P. Hanson, R. Phaal, D. Probert and Y. Shi (2007): "Making the Right Things in the Right Places: A Structured Approach to Developing and Exploiting 'Manufacturing Footprint' Strategy", Ed. 1, University of Cambridge Institute for Manufacturing, Cambridge. Coe, N. M., P. Dicken and M. Hess (2008): "Global production networks: realizing the potential", Journal of Economic Geography, Vol. 8 (3), pp. 271-295. Colotla, I. (2002): "Global manufacturing strategy formation: a proposal for a process approach integrating factory and network perspectives", 7th Annual Cambridge International Manufacturing Symposium, September 12-13, Cambridge, England, pp. 1-16.
194
REFERENCES
Colotla, I. (2003): "Operation and Performance of International Manufacturing Networks", Doctoral Thesis University of Cambridge, Cambridge, England. Colotla, I., Y. Shi and M. J. Gregory (2003): "Operation and performance of international manufacturing networks", International Journal of Operations & Production Management, Vol. 23 (10), pp. 1184-1206. Dasu, S. and J. de la Torre (1997): "Optimizing an international network of partially owned plants under conditions of trade liberalization", Management Science, Vol. 43 (3), pp. 313-333. De Toni, A. F. and M. Parussini (2010): "International manufacturing networks: a literature review", 17th International Annual EurOMA Conference, June 6-9, Porto, Portugal, pp. 1-10. Deflorin, P. (2007): "Fähigkeiten zur Umsetzung hybrider Produktionsstrategien in der Schweiz", Doctoral Thesis University of St.Gallen, St. Gallen, Switzerland. Denzin, N. K. (1978): "The Research Act", Ed. 2, McGraw-Hill, New York (NY). Diederichs, R., T. Meyer, M. Leopoldseder and F. Jacob (2008): "Management: applying best-practice structures and processes", in: E. Abele, T. Meyer, U. Näher, G. Strube and R. Sykes (Ed.): Global Production: A Handbook for Strategy and Implementation, Ed. 1, Springer, Berlin and Heidelberg, pp. 270-322. Doty, D. H., W. H. Glick and G. P. Huber (1993): "Fit, equifinality, and organizational effectiveness: a test of two configurational theories", The Academy of Management Journal, Vol. 36 (6), pp. 1196-1250. Doz, Y. L., J. Santos and P. Williamson (2001): "From Global to Metanational: How Companies Win in the Knowledge Economy", Ed. 1, Harvard Business School Press, Boston (MA). Drazin, R. and A. H. Van de Ven (1985): "Alternative forms of fit in contingency theory", Administrative Science Quarterly, Vol. 30 (4), pp. 514-539. DuBois, F., B. Toyne and M. Oliff (1993): "International manufacturing strategies of U.S. multinationals: a conceptual framework based on a four-industry study", Journal of International Business Studies, Vol. 24 (2), pp. 307 – 333.
REFERENCES
195
Dyer, J. H. and K. Nobeoka (2000): "Creating and managing a high-performance knowledge-sharing network: the Toyota case", Strategic Management Journal, Vol. 21 (3), pp. 345-367. Easterby-Smith, M., R. Thorpe and A. Lowe (2002): "Management Research: An Introduction", Ed. 2, Sage Publications, Thousand Oaks (CA) et al. Eisenhardt, K. M. (1989): "Building theories from case study research", The Academy of Management Review, Vol. 14 (4), pp. 532-550. Ernst, D. and L. Kim (2001): "Global production networks, knowledge diffusion, and local capability formation: a conceptual framework", Nelson & Winter Conference, June 12-15, Aalborg, Denmark, pp. 1-25. Ernst, D. and L. Kim (2002): "Global production networks, knowledge diffusion, and local capability formation", Research Policy, Vol. 31 (8-9), pp. 1417-1429. Feldmann, A. (2011): "A Strategic Perspective on Plants in Manufacturing Networks", Doctoral Thesis Linköping University, Linköping, Sweden. Feldmann, A. and J. Olhager (2009a): "Bundles of site competences in defining plant roles", Working Paper, Linköping University, Department of Management and Engineering, Linköping, Sweden. Feldmann, A. and J. Olhager (2009b): "Plant roles and decision-making in manufacturing networks", 16th International Annual EurOMA Conference, June 1417, Göteborg, Sweden, pp. 1-10. Feldmann, A. and J. Olhager (2011): "Distribution of manufacturing strategy decisionmaking in manufacturing networks", 18th International Annual EurOMA Conference, July 3-6, Cambridge, England, pp. 1-11. Feldmann, A., J. Olhager, D. Fleet and Y. Shi (2010): "Linking networks and plant roles: the impact of changing a plant role", 17th International Annual EurOMA Conference, June 6-9, Porto, Portugal, pp. 1-10. Feldmann, A., J. Olhager and F. Persson (2009): "Designing and managing manufacturing networks - a survey of Swedish plants", Production Planning & Control, Vol. 20 (2), pp. 101-112. Ferdows, K. (1997a): "Making the most of foreign factories", Harvard Business Review, Vol. 75 (1), pp. 73-88.
196
REFERENCES
Ferdows, K. (1997b): "Made in the world: the global spread of production", Production and Operations Management, Vol. 6 (2), pp. 102-109. Ferdows, K. (2003): "New world manufacturing order", Industrial Engineer, Vol. 35 (2), pp. 28-33. Ferdows, K. (2006): "Transfer of changing production know-how", Production and Operations Management, Vol. 15 (1), pp. 1-9. Ferdows, K. (2009): "Shaping global operations", GCG Georgetown University Universia, Vol. 3 (1), pp. 136 - 148. Ferdows, K., M. Lewis and J. Machuca (2004): "Rapid-fire fulfillment.", Harvard Business Review, Vol. 82 (11), pp. 104-110. Fine, C. H. and A. C. Hax (1985): "Manufacturing strategy: a methodology and an illustration", Interfaces, Vol. 15 (6), pp. 28-46. Flick, U. (1999): "Qualitative Forschung: Theorie, Methoden, Anwendung in Psychologie und Sozialwissenschaften", Ed. 1, Rowohlt, Reinbek. Flick, U. (2002): "An Introduction to Qualitative Research", Ed. 2, Sage Publications, London et al. Fox, M. S., J. F. Chionglo and F. G. Fadel (1993): "A common-sense model of the enterprise", Proceedings of the 2nd Industrial Engineering Research Conference, Norcross (GA), pp. 425-429. Friedli, T. (2000): "Die Architektur von Kooperationen", Doctoral Thesis University of St.Gallen, St. Gallen, Switzerland. Friedli, T. (2006): "Technologiemanagement: Modelle zur Wettbewerbsfähigkeit", Ed. 1, Springer, Berlin and Heidelberg.
Sicherung
der
Friedli, T., T. Gronauer and J. Werani (2010): "A model for the sustainable implementation of operational excellence", in: T. Friedli, P. K. Basu, T. Gronauer and J. Werani (Ed.): The Pathway to Operational Excellence in the Pharmaceutical Industry: Overcoming the internal inertia, Ed. 1, Editio Cantor, Aulendorf, pp. 202219.
REFERENCES
197
Friedli, T., S. Heinzen, A. Mundt and S. Thomas (2011): "Strategisches Management globaler Produktionsnetzwerke: Erschliessung des Netzwerkvorteils", Zeitschrift für wirtschaftlichen Fabrikbetrieb (ZWF), Vol. 106 (9), pp. 610-614. Fusco, J. P. and M. Spring (2003): "Flexibility versus robust networks: the case of the Brazilian automotive sector", Integrated Manufacturing Systems, Vol. 14 (1), pp. 2635. Galbraith, J. R. (1977): "Organization Design", Ed. 1, Addison-Wesley, Reading (MA). Gassmann, O. (1999): "Präxisnähe Wissenschaftsmanagement, Vol. 5 (3), pp. 11-16.
mit
Fallstudienforschung",
Ghoshal, S. and C. A. Bartlett (1988): "Creation, adoption, and diffusion of innovations by subsidiaries of multinational corporations", Journal of International Business Studies, Vol. 19 (3), pp. 365-388. Giddens, A. (1984): "The Constitution of Society: Outline of the Theory of Structuration", Ed. 1, University of California Press, Berkeley (CA). Grallert, S., D. Fleet, G. Lanza, R. Moser, Y. Shi and J. Ude (2010): "Configuring the right global manufacturing network: comparison of two practical processes", 15th Annual Cambridge International Manufacturing Symposium, September 23-24, Cambridge, England, pp. 1-22. Gray, J. V., A. V. Roth and B. Tomlin (2009): "The influence of cost and quality priorities on the propensity to outsource production", Decision Sciences, Vol. 40 (4), pp. 697-726. Gulati, R., N. Nohria and A. Zaheer (2000): "Strategic networks", Strategic Management Journal, Vol. 21 (3), pp. 203-215. Gupta, A. K. and V. Govindarajan (1991): "Knowledge flows and the structure of control within multinational corporations", The Academy of Management Review, Vol. 16 (4), pp. 768-792. Hallgren, M., J. Olhager and R. G. Schroeder (2011): "A hybrid model of competitive capabilities", International Journal of Operations & Production Management, Vol. 31 (5), pp. 511-526.
198
REFERENCES
Hax, A. C. and N. S. Majluf (1995): "The Strategy Concept and Process: A Pragmatic Approach", Ed. 2, Prentice Hall, Upper Saddle River (NJ). Hayes, R. H., G. P. Pisano, D. M. Upton and S. C. Wheelwright (2005): "Operations, Strategy, and Technology: Pursuing the Competitive Edge", Ed. 1, Wiley, Hoboken (NJ) et al. Hayes, R. H. and S. C. Wheelwright (1984): "Restoring Our Competitive Edge: Competing Through Manufacturing", Ed. 1, Wiley, Hoboken (NJ) et al. Hayes, R. H., S. C. Wheelwright and K. Clark (1988): "Dynamic manufacturing: Creating the Learning Organization", Ed. 1, Free Press, New York (NY). Henderson, R. I. (2005): "Compensation Management in a Knowledge-Based World", Ed. 10, Prentice Hall, Upper Saddle River (NJ). Hevner, A. R., S. T. March, J. Park and S. Ram (2004): "Design science in information systems research", MIS Quarterly, Vol. 28 (1), pp. 75-105. Hill, T. (1993): "Manufacturing Strategy: The Strategic Management of the Manufacturing Function", Ed. 2, Macmillan Press Ltd., London. Hill, T. (2000): "Manufacturing Strategy: Text and Cases", Ed. 2, Palgrave Macmillan, Basingstoke and New York (NY). ISO (1999): "Industrial Automation Systems: Requirements for Enterprise-Reference Architectures and Methodologies", ISO/DIS 15704, International Organisation for Standardisation. Jacob, F. (2006): "Quantitative Optimierung dynamischer Produktionsnetzwerke", Doctoral Thesis RWTH Aachen, Aachen, Germany. Jacob, F. and G. Strube (2008): "Why go global? The multinational imperative", in: E. Abele, T. Meyer, U. Näher, G. Strube and R. Sykes (Ed.): Global Production: A Handbook for Strategy and Implementation, Ed. 1, Springer, Berlin and Heidelberg, pp. 2-33. Jaehne, D. M., M. Li, R. Riedel and E. Mueller (2009): "Configuring and operating global production networks", International Journal of Production Research, Vol. 47 (8), pp. 2013-2030.
REFERENCES
199
Justus, A. (2009): "Management globaler Produktionsnetzwerke", Doctoral Thesis University of St.Gallen, St. Gallen, Switzerland. Karlsson, C. and M. Sköld (2007): "The manufacturing extraprise: an emerging production network paradigm", Journal of Manufacturing Technology Management, Vol. 18 (8), pp. 912-932. Ketokivi, M. and M. Jokinen (2006): "Strategy, uncertainty and the focused factory in international process manufacturing", Journal of Operations Management, Vol. 24 (3), pp. 250-270. Khurana, A. and B. Talbot (1999): "Plant missions in global manufacturing networks: a resource-based perspective with evidence from the global color picture tube industry", Working Paper, University of Michigan Business School, Ann Arbor (MI), USA. Kim, J. S. and P. Arnold (1996): "Operationalizing manufacturing strategy: an exploratory study of constructs and linkage", International Journal of Operations & Production Management, Vol. 16 (12), pp. 45–73. Kinkel, S. (2004): "Erfolgsfaktor Standortplanung: In- und ausländische Standorte richtig bewerten", Ed. 1, Springer, Berlin and Heidelberg. Kinkel, S. and S. Maloca (2009): "Produktionsverlagerung und Rückverlagerung in Zeiten der Krise", Modernisierung der Produktion: Mitteilungen aus der ISI-Erhebung, Fraunhofer-Institut für System- und Innovationsforschung ISI. Klassen, R. D. and D. C. Whybark (1994): "Barriers to the management of international operations", Journal of Operations Management, Vol. 11 (4), pp. 385396. Kleijnen, J. P. C. (2005): "Supply chain simulation tools and techniques: a survey", International Journal of Simulation & Process Modelling, Vol. 1 (1), pp. 82-89. Kogut, B. and N. Kulatilaka (1994): "Operating flexibility, global manufacturing, and the option value of a multinational network", Management Science, Vol. 40 (1), pp. 123-139. Kubicek, H. (1977): "Heuristische Bezugsrahmen und heuristisch angelegte Forschungsdesigns als Elemente einer Konstruktionsstrategie empirischer Forschung", in: R. Köhler (Ed.): Empirische und handlungstheoretische Forschungskonzeptionen in der Betriebswirtschaftslehre, Ed. 1, C. E. Poeschel, Stuttgart, pp. 3-36.
200
REFERENCES
Lanza, G. and J. Ude (2010): "Multidimensional evaluation of value added networks", CIRP Annals - Manufacturing Technology, Vol. 59 (1), pp. 489-492. Lee, W. B. and H. C. W. Lau (1999): "Factory on demand: the shaping of an agile production network", International Journal of Agile Management Systems, Vol. 1 (2), pp. 83-87. Leong, G. K., D. L. Snyder and P. T. Ward (1990): "Research in the process and content of manufacturing strategy", Omega, Vol. 18 (2), pp. 109-122. Lewin, K. (1946): "Action research and minority problems", Journal of Social Issues, Vol. 2 (4), pp. 34-46. Love, R. F., J. G. Morris and G. O. Wesolowsky (1988): "Facilities Location: Models and Methods", Ed. 1, North-Holland, Amsterdam. Luo, Y. (2005): "Toward coopetition within a multinational enterprise: a perspective from foreign subsidiaries", Journal of World Business, Vol. 40 (1), pp. 71-90. MacCarthy, B. L. and W. Atthirawong (2003): "Factors affecting location decisions in international operations - a Delphi study", International Journal of Operations & Production Management, Vol. 23 (7), pp. 794-818. Malone, T. W. and K. Crowston (1994): "The interdisciplinary study of coordination", ACM Computing Surveys, Vol. 26 (1), pp. 87-119. Malone, T. W., K. Crowston, J. Lee, B. Pentland, C. Dellarocas, G. Wyner, J. Quimby, C. S. Osborn, A. Bernstein, G. Herman, M. Klein and E. O'Donnell (1999): "Tools for inventing organizations: toward a handbook of organizational processes", Management Science, Vol. 45 (3), pp. 425-443. Mapes, J., C. New and M. Szwejczewski (1997): "Performance trade-offs in manufacturing plants", International Journal of Operations & Production Management, Vol. 17 (10), pp. 1020-1033. March, J. G. and H. A. Simon (1958): "Organizations", Ed. 1, Wiley, New York (NY). Maritan, C. A., T. H. Brush and A. G. Karnani (2004): "Plant roles and decision autonomy in multinational plant networks", Journal of Operations Management, Vol. 22 (5), pp. 489-503.
REFERENCES
201
Martinez, J. I. and J. C. Jarillo (1989): "The evolution of research on coordination mechanisms in multinational corporations", Journal of International Business Studies, Vol. 20 (3), pp. 489-514. Mascarenhas, B. (1984): "The Coordination of Manufacturing Interdependence in Multinational Companies", Journal of International Business Studies, Vol. 15 (3), pp. 91-106. Mauri, A. J. (2009): "Influence of MNC network configuration patterns on the volatility of firm performance: an empirical investigation", Management International Review, Vol. 49 (6), pp. 691-707. Mauri, A. J. and R. B. Sambharya (2001): "The impact of global integration on MNC performance: evidence from global industries", International Business Review, Vol. 10 (4), pp. 441-454. Mefford, R. N. (1986): "Determinants of productivity differences in international manufacturing", Journal of International Business Studies, Vol. 17 (1), pp. 63-82. Meijboom, B. and H. Voordijk (2003): "International operations and location decisions: a firm level approach", Tijdschrift voor economische en sociale geografie, Vol. 94 (4), pp. 463-476. Meijboom, B. and B. Vos (1997): "International manufacturing and location decisions: balancing configuration and co-ordination aspects", International Journal of Operations & Production Management, Vol. 17 (8), pp. 790-805. Meijboom, B. and B. Vos (2004): "Site competence dynamics in international manufacturing networks: instrument development and a test in Eastern European factories", Journal of Purchasing & Supply Management, Vol. 10 (3), pp. 127-136. Meredith, J. (1998): "Building operations management theory through case and field research", Journal of Operations Management Vol. 16 (4), pp. 441-454. Meyer, T. and F. Jacob (2008): "Network design: optimizing the global production footprint", in: E. Abele, T. Meyer, U. Näher, G. Strube and R. Sykes (Ed.): Global Production: A Handbook for Strategy and Implementation, Ed. 1, Springer, Berlin and Heidelberg, pp. 140-190.
202
REFERENCES
Mills, J. and K. Platts (2011): "Performance improvement and resource co-ordination", 18th International Annual EurOMA Conference, July 3-6, Cambridge, England, pp. 110. Miltenburg, J. (1995): "Manufacturing Strategy: How to Formulate and Implement a Winning Plan", Ed. 1, Productivity Press, Portland (OR). Miltenburg, J. (2005): "Manufacturing Strategy: How to Formulate and Implement a Winning Plan", Ed. 2, Productivity Press, New York (NY). Miltenburg, J. (2008): "Setting manufacturing strategy for a factory-within-a-factory", International Journal of Production Economics, Vol. 113 (1), pp. 307–323. Miltenburg, J. (2009): "Setting manufacturing strategy for a company's international manufacturing network", International Journal of Production Research, Vol. 47 (22), pp. 6179-6203. Mintzberg, H. (1989): "Mintzberg on Manangement", Ed. 1, Free Press, New York (NY). Nassimbeni, G. (1998): "Network structures and co-ordination mechanisms: a taxonomy", International Journal of Operations & Production Management, Vol. 18 (6), pp. 538-554. Nonaka, I. (1991): "The knowledge-creating company", Harvard Business Review, Vol. 69 (6), pp. 96-104. Noori, H. and W. B. Lee (2006): "Dispersed network manufacturing: adapting SMEs to compete on the global scale", Journal of Manufacturing Technology Management, Vol. 17 (8), pp. 1022-1041. Petersen, C. G., G. R. Aase and D. R. Heiser (2011): "Journal ranking analyses of operations management research", International Journal of Operations & Production Management, Vol. 31 (4), pp. 405-422. Platts, K. W. (1994): "Characteristics of methodologies for manufacturing strategy formulation", Computer Integrated Manufacturing Systems, Vol. 7 (2), pp. 93-99. Platts, K. W. and M. J. Gregory (1990): "Manufacturing audit in the process of strategy formulation", International Journal of Operations & Production Management, Vol. 10 (9), pp. 5 - 26.
REFERENCES
203
Platts, K. W., J. F. Mills, M. C. Bourne, A. D. Neely, A. H. Richards and M. J. Gregory (1998): "Testing manufacturing strategy formulation processes", International Journal of Production Economics, Vol. 56-57, pp. 517–523. Polanyi, M. (1958): "Personal Knowledge: Towards a Post-Critical Philosophy", Ed. 1, University of Chicago Press, Chicago (IL). Pontrandolfo, P. and O. G. Okogbaa (1999): "Global manufacturing: a review and a framework for planning in global corporation", International Journal of Production Research, Vol. 37 (1), pp. 1-19. Porter, M. E. (1979): "How competitive forces shape strategy", Harvard Business Review, Vol. 57. (March-April), pp. 137-156. Porter, M. E. (1986): "Changing patterns of international competition", California Management Review, Vol. 28 (2), pp. 9-40. Porter, M. E. and V. E. Millar (1985): "How information gives you competitive advantage", Harvard Business Review, Vol. 63 (4), pp. 149–160. Prasad, S. and S. Babbar (2000): "International operations management research", Journal of Operations Management, Vol. 18 (2), pp. 209-247. Prasad, S., S. Babbar and J. Motwani (2001): "International operations strategy: current efforts and future directions", International Journal of Operations & Production Management, Vol. 21 (5/6), pp. 645-665. Riis, J. O., J. Johansen, B. V. Waehrens and L. Englyst (2007): "Strategic roles of manufacturing", Journal of Manufacturing Technology Management, Vol. 18 (8), pp. 933-948. Rogers, D. L. and D. A. Whetten (1982): "Interorganizational Coordination: Theory, Research, and Implementation", Ed. 1, Iowa State University Press, Ames (IA). Rudberg, M. and J. Olhager (2003): "Manufacturing networks and supply chains: an operations strategy perspective", Omega, Vol. 31 (1), pp. 29-39. Rudberg, M. and B. M. West (2008): "Global operations strategy: coordinating manufacturing networks", Omega, Vol. 36 (1), pp. 91-106.
204
REFERENCES
Rüegg-Stürm, J. (2005): "The New St. Gallen Management Model: Basic Categories of an Approach to Integrated Management", Ed. 1, Palgrave Macmillan, Basingstoke and New York (NY). Salter, M. S. (1973): "Tailor incentive compensation to strategy", Harvard Business Review, Vol. 51 (2), pp. 94-102. Samson, D. (1991): "Manufacturing and Operations Strategy", Ed. 1, Prentice Hall, Upper Saddle River (NJ). Schill, C. O. (1990): "Industrielle Standortplanung: Eine theoretische Konzeption und deren praktische Anwendung", Ed. 1, P. Lang, Frankfurt et al. Schmenner, R. W. (1982): "Multiplant manufacturing strategies among the fortune 500", Journal of Operations Management, Vol. 2 (2), pp. 77-86. Schmid, L. (2011): "Erfolgsdynamik: Ein Konzept für die Erstellung systemdynamischer Modelle zur Entscheidungsunterstützung im strategischen Management", Doctoral Thesis University of St.Gallen, St. Gallen, Switzerland. Shi, Y. (2003): "Internationalisation and evolution of manufacturing systems: classic process models, new industrial issues, and academic challenges", Integrated Manufacturing Systems, Vol. 14 (4), pp. 357-368. Shi, Y., H. Briggs and M. J. Gregory (2001): "Global manufacturing strategy: contents, processes and challenges", 8th International Annual EurOMA Conference, June 3-5, Bath, England, pp. 46-59. Shi, Y. and M. J. Gregory (1998): "International manufacturing networks - to develop global competitive capabilities", Journal of Operations Management, Vol. 16 (2-3), pp. 195-214. Shi, Y. and M. J. Gregory (2005): "Emergence of global manufacturing virtual networks and establishment of new manufacturing infrastructure for faster innovation and firm growth", Production Planning & Control, Vol. 16 (6), pp. 621-631. Shi, Y., M. J. Gregory and M. Naylor (1997): "International manufacturing configuration map: a self-assessment tool of international manufacturing capabilities", Integrated Manufacturing Systems, Vol. 8 (5), pp. 273-282. Skinner, W. (1996): "Three yards and a cloud of dust: industrial management at century end", Production and Operations Management, Vol. 5 (1), pp. 15-24.
REFERENCES
205
Slack, N. and M. Lewis (2002): "Operations Strategy", Ed. 1, Pearson Education, Harlow. Smith, K. G., S. J. Carroll and S. J. Ashford (1995): "Intra- and interorganizational cooperation: toward a research agenda", The Academy of Management Journal, Vol. 38 (1), pp. 7-23. Sousa, R. and C. A. Voss (2008): "Contingency research in operations management practices", Journal of Operations Management, Vol. 26 (6), pp. 697-713. Stremme, U. (2000): "Internationales Strategisches Produktionsmanagement", Ed. 1, P. Lang, Frankfurt et al. Stuart, I., D. McCutcheon, R. Handfield, R. McLachlin and D. Samson (2002): "Effective case research in operations management: a process perspective", Journal of Operations Management, Vol. 20 (5), pp. 419-433. Sugiura, H. (1990): " How Honda localizes its global strategy", Sloan Management Review, Vol. 31 (1), pp. 77-82. Sweeney, M. T. (1994): "A methodology for the strategic management of international manufacturing and sourcing", The International Journal of Logistics Management, Vol. 5 (1), pp. 55-65. Thomas, S. (2013): "Produktionsnetzwerksysteme: Ein Weg zu effizienten Produktionsnetzwerken", Doctoral Thesis University of St.Gallen (in print), St. Gallen, Switzerland. Thompson, J. D. (1967): "Organizations in Action: Social Science Bases of Administrative Theory", Ed. 1, McGraw-Hill, New York (NY). Tomczak, T. (1992): "Forschungsmethoden in der Marketingwissenschaft: Ein Plädoyer für den qualitativen Forschungsansatz", Marketing-ZFP, Vol. 14 (2), pp. 7787. Tsai, W. (2001): "Knowledge transfer in intraorganizational networks: effects of network position and absorptive capacity on business unit innovation and performance", The Academy of Management Journal, Vol. 44 (5), pp. 996-1004.
206
REFERENCES
Tsai, W. (2002): "Social structure of "coopetition" within a multiunit organization: coordination, competition, and intraorganizational knowledge sharing", Organization Science, Vol. 13 (2), pp. 179-190. Tsai, W. and S. Ghoshal (1998): "Social capital and value creation: the role of intrafirm networks", The Academy of Management Journal, Vol. 41 (4), pp. 464-476. Tykal, D. (2009): "Die strategische Rolle von Produktionsstandorten in Hochlohnländern: Relevanz und Gestaltungsmerkmale von Lead Factories in Produktionsnetzwerken", Doctoral Thesis University of St.Gallen, St. Gallen, Switzerland. U.N. (2011): "World Investment Report 2011: Non-Equity Modes of International Production and Development", United Nations, New York (NY) and Geneva. Ude, J. (2010): "Entscheidungsunterstützung für die Konfiguration globaler Wertschöpfungsnetzwerke: Ein Bewertungsansatz unter Berücksichtigung multikriterieller Zielsysteme, Dynamik und Unsicherheit", Doctoral Thesis Karlsruhe Institut of Technology, Karlsruhe, Germany. Ulrich, H. (1984): "Die Betriebswirtschaftslehre als anwendungsorientierte Sozialwissenschaft", in: T. Dyllick and G. Probst (Ed.): Management, Ed. 1, Haupt, Bern, pp. 168-199. van de Ven, A. D. M. (1989): "International manufacturing: an analysis on factory level on the centre-periphery relations", Engineering Costs and Production Economics, Vol. 16 (1), pp. 11-33. Vecchi, A. and L. Brennan (2009): "A cultural perspective on innovation in international manufacturing", Research in International Business and Finance, Vol. 23 (2), pp. 181-192. Vereecke, A. and A. De Meyer (2009): "The dynamic management of manufacturing networks", Working Paper, Vlerick Leuven Gent Management School, Gent, Belgium. Vereecke, A. and R. Van Dierdonck (2002): "The strategic role of the plant: testing Ferdows’s model", International Journal of Operations & Production Management, Vol. 22 (5), pp. 492-514. Vereecke, A., R. Van Dierdonck and A. De Meyer (2006): "A typology of plants in global manufacturing networks", Management Science, Vol. 52 (11), pp. 1737-1750.
REFERENCES
207
Vernadat, F. (1996): "Enterprise Modeling and Integration: Principles and Applications", Ed. 1, Chapman & Hall, London. Vickery, S. K. (1991): "A theory of production competence revisited", Decision Sciences, Vol. 22 (3), pp. 635–643. Vokurka, R. J. and R. A. Davis (2004): "Manufacturing strategic facility types", Industrial Management & Data Systems, Vol. 104 (6), pp. 490-504. Vom Brocke, J. and C. Buddendick (2006): "Reusable conceptual models: requirements based on the design science research paradigm", First International Conference on Design Science Research in Information Systems and Technology (DESRIST 2006), February 24-25, Claremont (CA), pp. 576–604. Vos, G. C. J. M. (1991): "A production-allocation approach for international manufacturing strategy", International Journal of Operations & Production Management, Vol. 11 (3), pp. 125-134. Voss, C., N. Tsikriktsis and M. Frohlich (2002): "Case research in operations management", International Journal of Operations & Production Management Vol. 22 (2), pp. 195-219. Weston, R. H. (1999): "Reconfigurable, component-based systems and the role of enterprise engineering concepts", Computers in Industry, Vol. 40 (2-3), pp. 321–343. Wheelwright, S. C. (1984): "Manufacturing strategy: defining the missing link", Strategic Management Journal, Vol. 5 (1), pp. 77-91. Wheelwright, S. C. and R. H. Hayes (1985): "Competing through manufacturing", Harvard Business Review, Vol. 63 (1), pp. 99-109. Wright, C., C.-S. Suh and C. Leggett (2009): "If at first you don't succeed: globalized production and organizational learning at the Hyundai Motor Company", Asia Pacific Business Review, Vol. 15 (2), pp. 163–180. Yin, R. K. (2003): "Case Study Research: Design and Methods", Ed. 3, Sage Publications, Thousand Oaks (CA) et al.
GER Electrical Vacuum equipment cleaners & devices GER Medical & Dental technical prosthetics & products restoration material HUN Food Pet food products
Floor Care NW
* ** *** ****
Chocolate NW
Elevator n.a. Mechanical Elevators NW*** & electrical products 9
4 int. / 17 ext.
7
8
6
8
12
18
5
> 3000
Local for local
Local for local / Web structure
Transition from world products to local for local World products / Local for local
Regional (Europe)
Multinat. Local for local / (2 regions) World products
Sequential or convergent
Web structure / Local for local
Multinat. World products (3 regions)
Multinat. Sequential or (2 regions) convergent
Worldw.
Worldw.
Worldw.
Worldw.
1000 int. / Regional 1000 ext. (Europe)
ca. 1000
ca. 400
> 1000
ca. 1500
ca. 1000
ca. 1700
ca. 400
Network characteristics
# Sites # Employ. Global Network (operat.) (operat.) dispersion structure* 12 ca. 250 Worldw. Hub & spoke
According to Meyer and Jacob (2008) According to the survey classification of the network capability level and conformance Supply Chain Europe organisation Participation only sporadic
n.a. Food products
Food processing & packaging
Component manufacturing & assembly Chemical process engineering
Extrusion
Extrusion
Chipping technology
Core processes Engineering & assembly Engineering & assembly
1) Sheet Metal 2) Mech. & Comp. 3) Electronics 4) Drives Chocolate & Food processing pralines & packaging
Frames & profiles
n/a Polymer products
Profile NW
Edgeband
n/a Polymer products
Edgeband NW
GER Mechanical Mechanical products seals
56 Respondents from quantitative survey
Pet Food NW
Dental NW
Seals NW
CH
CH
HQ Industry
Core products Electrical Excitation equipment systems Electrical Voltage equipment drives
Network scope
Company Division Business Other level level unit level Excitation NW Drives NW
IV
I
III
II
I
n.a.
n.a.
n.a.
n.a.
Network category** n.a.
12 (avg. 3h)
10 (avg. 4h)
Head of Internat. Operations Manager SCM Manager SCM Manager Controlling
1 (ca. 7 h)
Head of Global Production Head of SCM
Site Head**** Assistant to Site Head Head of Engineering Head of SCM
1 (ca. 7 h)
1 (ca. 7 h)
7 (avg. 7 h)
9 (avg. 7 h)
8 (avg. 7 h)
Head of Global Production
Head of Production**** Manager Global Production Head of Operations Manager Production Head of SCM* Manager Business Devel. Head of Division**** Head of Business Unit**** Head of Operations Manager Production Manager Technics Head of Product Line Assistant to COO Site Head
# Interviews / Workshops Head of Global Prod. Line 13 (avg. 3 h) Head of Global Prod. Line 8 (avg. 3 h)
Empirical data base Functions involved
Research
Benchmarking Consulting
Benchmarking
Benchmarking
Benchmarking
Consulting
Research & consulting
Research & consulting
Research & consulting Research & consulting
Project
208 APPENDIX A: CASE STUDIES & INTERVIEWS
Appendix A: Case Studies & Interviews
APPENDIX B: QUESTIONNAIRE
209
Appendix B: Questionnaire B.1
Survey Questionnaire (Excerpt of Questions) General Information 1.1 According to ISIC Rev.4, in which industrial sector are you operating? 10 - Manufacture of food products 11 - Manufacture of beverages 12 - Manufacture of tobacco products 13 - Manufacture of textiles 14 - Manufacture of wearing apparel 15 - Manufacture of leather and related products 16 - Manufacture of wood and of products of wood and cork, except furniture; manufacture of articles of straw and plaiting materials 17 - Manufacture of paper and paper products 18 - Printing and reproduction of recorded media 19 - Manufacture of coke and refined petroleum products 20 - Manufacture of chemicals and chemical products 21 - Manufacture of basic pharmaceutical products and pharmaceutical preparations 22 - Manufacture of rubber and plastics products 23 - Manufacture of other non-metallic mineral products 24 - Manufacture of basic metals 25 - Manufacture of fabricated metal products, except machinery and equipment 26 - Manufacture of computer, electronic and optical products 27 - Manufacture of electrical equipment 28 - Manufacture of machinery and equipment n.e.c. 29 - Manufacture of motor vehicles, trailers and semi-trailers 30 - Manufacture of other transport equipment 31 - Manufacture of furniture 32 - Other manufacturing: 33 - Repair and installation of machinery and equipment
1.2 Please indicate the size of your manufacturing network´s business, the sales development of the last five years and the current number of employees. < 100
100 - 250
251 - 400
401 - 600
601 - 800
801 - 1.000
> 1.000
Total sales 2009 (Mio. €)
Far below market growth
Slightly below market growth
Equals market growth
Slightly above market growth
Far above market growth
Sales develop-ment (2005-2009)
< 50
50 - 200
201 - 500
501 - 1.000
1.001 5.000
5.001 10.000
Number of em-ployees (2009)
Share of employees located in HIGH-COST COUNTRIES*:
%
Share of employees located in LOW-COST COUNTRIES*: (should add up to 100%)
% ∑
0%
> 10.000
210
APPENDIX B: QUESTIONNAIRE Product and Process Information 1.3 Following questions give us a deeper understanding of your manufacturing network´s product and process characteristics. Please keep in mind always to answer the questions from your manufacturing network perspective. Product type manufactured (multiple
Industrial goods
Consumer goods
Market segments served (multiple answ ers
Low-end
Mid-range
answ ers possible)
High-end
possible)
1 line
Number of product lines sold (from
2-5 lines
6 - 10 lines
11 - 20 lines
21 - 30 lines
31 - 40 lines
> 40 lines
n/a
Very high
n/a
marketing perpective)
Average product line variety (i.e. number of
Very small
variants per product line)
Medium value products
High value products Average product value
Average length of the product life cycle Average number of annual design changes (per product line)
Low value products
(e.g. middle class cars)
(e.g. ships, aircrafts)
(e.g. plugs, pens)
n/a
<1 year
1-3 years
4-6 years
7 - 10 years
11 - 20 years
21 - 40 years
> 40 years
n/a
1-5
6 - 10
11 - 20
21 - 30
31 - 50
51 - 100
> 100
n/a
Organisational Structure 2.1 Below various types of schematic organisational structures are listed. Which organisational chart reflects the structure of your manufacturing network best? Functional Function
Divisional Division
Divisional mix Product Division
Matrix
Region Division
Product A
Plant 1 Plant 2
Plant 1
Plant 1
Plant 3
…
Plant 2
Plant 2
Plant 4
…
…
…
If you have a functional organisation, please specify the function your plants are assigned to
If you have a divisional organisation, please specify the division your plants are assigned to
If you have a divisional mix organisation, please specify the dimensions of this mix
Region 1 Region 2
If you have a matrix organisation, please specify the dimensions of the matrix
Direct CEO
Product division
Product & region division
Product-region matrix
Manufacturing
Regional division
Function & region division
Function-region matrix
International department
Functional division
Product & function division
Product-function matrix
Other, please specify:
APPENDIX B: QUESTIONNAIRE
211
2.2 According to which centre type are your sites managed within your manufacturing network? Cost centre
Site is responsible for costs. Central targets for the site have to be met by least resource consumption.
Profit centre
Site is responsible for profit and loss. Since targets are still set by corporate strategy, decisions mainly cover price and output.
Both
Other
Please specify:
Fragmentation of the Network 2.3 Below, various types of schematic network structures are listed. Please indicate which of the following structures describe your manufacturing network best by ticking the box on the right side. Global Structure
Market structure
Hub & Spoke
Value chain
Flexible order allocation
Regional Structure
Global structure: A world product* is produced at a single site. Markets are served by export.
Regional structure: Similar to Global Structure, but each site is serving a certain region only.
Global structure: The main share of manufacturing work is conducted at a single hub. In order to meet market-specific requirements products are adapted at local spokes.
Regional structure: As Global Structure, but on a regional base with regional hubs and spokes.
Global structure: The production process is organised as a chain, with process steps conducted at succeeding sites. Hence, sites do not necessarily serve a market.
Regional structure: As Global Structure, but on a regional base.
The network exists at a Global structure only. All sites have almost the full set of production competences. Orders are assigned to a specific site within the manufacturing network to achieve a high degree of capacity utilisation.
t no
is ex
ting
Other Annotations: Full range manufacturing site (performing all process steps)
Internal flow of non finished goods
Limited range manufacturing site (performing selected process steps)
Delivery of finished goods
Markets served
212
APPENDIX B: QUESTIONNAIRE Network Dispersion
Rest of Asia*
Australia & New Zealand
Rest of Asia
Australia & New Zealand
China
India
Africa
Russia etc.*
Middle East*
Central & South America*
North America*
Site
Eastern Europe*
Operating region
Western Europe*
2.6 Please indicate the geographic dispersion of your manufacturing network by (1) entering the number of sites located in each operating region listed below and (2) ticking the boxes naming the markets that are regularly served from sites in the corresponding regions.
(1) Number of sites located in region Pure R&D sites Manufacturing / engineering sites Mixed sites
(2) Markets regularly served from corresponding region: Western Europe* Eastern Europe* North America* Central & South America* Middle East* Russia etc.* Africa India China Rest of Asia* Australia & New Zealand
Share of total sales generated in region* [%]
0%
Import quota of final products* [%] Share of total manufacturing value* [%]
0%
Share of component value transfered to other regions* [%]
China
India
Africa
Russia etc.
Middle East
Central & South America
(USA & CAN)
North America
Eastern Europe
Western Europe
2.7 For each of the operating regions listed below, please specify the manufacturing and sales structure. Please base your answers upon your financial volume.
APPENDIX B: QUESTIONNAIRE
213
Strategic Network Focus From today's perspective, please rate the strategic importance of the following statements in order to achieve competitiveness through your network. 3.2 With respect to competitiveness, our manufacturing network should provide us with favourable access to: Not important
Very important
n/a
… markets & customers (e.g. proximity to customers, product adaptation to local taste)
… competitors (e.g. proximity to main competitors, maintaining market share) … socio-political factors (e.g. tax
breaks,overcome trade barriers, hedge exchange rate fluctuations)
… image
(e.g. marketing, tradition, "made in …")
… suppliers (e.g. rapid delivery, low cost suppliers, raw material access)
… low-cost labour (e.g. best cost w orkforce) … skilled labour (e.g. qualified w orkers, enigneers, specialists)
… external know-how (e.g. universities,
competence clusters, engineering services)
3.3 With respect to competitiveness, our manufacturing network should provide us with cost advantages and contribute to efficiency improvements by realizing: Not important
Very important
n/a
… economies of scale* (i.e. realising cost benefits by concentrating identical products) … economies of scope* (i.e. realising cost benefits by concentrating products w ith similar processes)
… reduction of duplications (i.e. realising cost
benefits by concentrating support functions, e.g. IT)
3.4 With respect to competitiveness, our manufacturing network should enable us to quickly respond to changes in the market through economically transferring: Not important
Very important
n/a
… product, process or personnel (i.e.
exchanging products, processes or personnel fast & efficiently)
… production volumes (i.e. exchanging production volumes and orders fast & efficiently)
3.5 With respect to competitiveness, our manufacturing network should provide us with excellent learning ability through unlocking and transferring knowledge about: Not important … external factors (e.g. local know ledge about markets, customers & culture)
… internal factors (e.g. internal know ledge about products, processes & technology)
Very important
n/a
214
APPENDIX B: QUESTIONNAIRE Centralisation of the Network The next questions are focusing on the degree of CENTRALISATION and STANDARDISATION of selected systems, decisions and processes in your manufacturing network. 5.1 At your manufacturing network, who holds the responsibility for DEFINING AND DEVELOPING the following SYSTEMS*? Each site individually
Selected sites related to site competences
Each region individually
Central unit
n/a
Production system (e.g. production methods, guidelines, principles) Product Data Management system (i.e. handling & storage of product data)
Systems*
Quality management system
(e.g. rules, guidelines, methods, standards)
Management system (e.g. management methods, scorecards, KPI's) Improvement programs besides production (e.g. support processes) HR system (e.g. job descriptions, employee development) Know-how exchange system
5.2 At your manufacturing network, who holds the responsibility for MAKING the following DECISIONS? Each site individually
Selected sites related to site competences
Each region individually
Central unit
n/a
Make or buy decisions Production process decisions (i.e. selection & allocation of process types)
Decisions
Manufacturing technology decisions (i.e. technology selection)
Long-term production capacity development decisions Short-term production capacity allocation (e.g. machine transfers) Organisational structure decisions (e.g. organisational size & structure)
Product allocation decisions (i.e. assigning products to sites)
5.3 At your manufacturing network, who holds the responsibility for EXECUTING the following PROCESSES? Each site individually Strategic sourcing (e.g. supplier selection & qualification)
Processes
Strategic logistics (e.g. selection of logistics partners)
Short-term manufacturing planning & control Long-term sales & operations planning (i.e. production planning) Internal supply chain planning (i.e. allocation of orders)
Selected sites related to site competences
Each region individually
Central unit
n/a
APPENDIX B: QUESTIONNAIRE
215
Standardisation of the network 5.4 Please indicate the degree of STANDARDISATION for the following SYSTEMS* in your network. Individual tools / heterogeneous implementation level at each site
Individual tools / homogeneous implementation level at each site
Standardised tools / heterogeneous implementation level in the network
Standardised tools / homogeneous implementation level in the network
n/a
Production system (e.g. production methods, guidelines, principles) Product Data Management system (i.e. handling & storage of product data)
Systems*
Quality management system
(e.g. rules, guidelines, methods, standards)
Management system (e.g. management methods, scorecards, KPI's) Improvement programs besides production (e.g. support processes) HR system (e.g. job descriptions, employee development) Know-how exchange system
5.5 Please indicate the degree of STANDARDISATION in your network for MAKING the following DECISIONS. No / local standardisation in the network
Documented rules & guidelines in the network
Audited / controlled decision routines in the network
Standardised (IT-) tools or methods in the network
n/a
Make or buy decisions
Decisions
Production process decisions (i.e. selection & allocation of process types) Manufacturing technology decisions (i.e. technology selection) Long-term production capacity development decisions Short-term production capacity allocation (e.g. machine transfers) Organisational structure decisions (e.g. organisational size & structure)
Product allocation decisions (i.e. assigning products to sites)
5.6 Please indicate the degree of STANDARDISATION in your network for EXECUTING the following PROCESSES. No / local standardisation in the network Strategic sourcing (e.g. supplier selection & qualification)
Processes
Strategic logistics (e.g. selection of logistics partners)
Short-term manufacturing planning & control Long-term sales & operations planning (i.e. production planning) Internal supply chain planning (i.e. allocation of orders)
Documented rules & processes in the network
Audited / controlled processes in the network
Standardised (IT-) tools or methods in the network
n/a
216
APPENDIX B: QUESTIONNAIRE Incentive System 5.7 Below, performance dimensions are listed which companies typically use for their incentive setting. For each of the performance dimensions please indicate how and on which organisational level TARGETS FOR INCENTIVES are set in your manufacturing network. Incentives based on … Performance dimension
No targets / incentives set
Individual targets set for each site
Common targets set for each site
Targets set for the network
Targets set above network level (e.g. on
n/a
corporate level)
Overall financial performance (e.g. EBIT. ROI)
Market / sales performance
(e.g. market share, sales figures)
Operational performance
(e.g. manuf. cost, lead time, OEE)
Contribution to learning
(sharing know -how & information)
Conformance with strategic goals Others, please specify:
5.8 For each of the performance dimensions listed below, please specify your incentive system in terms of: (1) the reward mechanism applied (multiple answers possible) (2) how the rewards are allocated between sites (multiple answers possible) (1) Reward mechanism Financial payments
Reputation & awards
(2) Reward allocation
Autonomy & degree of responsibility
Equally between sites
Based on individual contribution / achievement of sites
Overall financial performance (e.g. EBIT. ROI)
Market / sales performance
(e.g. market share, sales figures)
Operational performance
(e.g. manuf. cost, lead time, OEE)
Contribution to learning
(sharing know -how & information)
Conformance with strategic goals
Exchange of Information 5.10 For those types of information that are exchanged in your network, please indicate (1) the quality and extent of information provided and (2) the extent to which sites have access to information as well as the degree of details they can access.
… markets / customers
(e.g. market trend, grow th & share)
… competitors (e.g. competitor actions, new competitors) … suppliers (e.g. new suppliers, supplier quality performance)
Most sites have access to full data
Most sites have access to limited data
Selected sites have access to full data
Selected sites have access to limited data
(2) Access to information*
No access
Comprehensive data provided by most sites
Basic data provided by most sites
Comprehensive data provided by few sites
Basic data provided by few sites
Information about …
No provision
(1) Provision of information*
APPENDIX B: QUESTIONNAIRE
217
Sourcing Strategy 6.1 For each of the PROCUREMENT TYPES listed below, please declare (1) its share of your total sourcing volume and (2) the percentage your manufacturing network sources in low- or high-cost countries. (1) Volume Monetary annual sourcing vol./ total sourcing vol.
MRO supplies* (i.e. maintenance, repair, operating mat.)
%
Raw material
%
Simple parts & high volume standard components
%
Complex parts
%
Core components
%
(2) Low-cost vs. high-cost sourcing 100% low-cost
75% low-cost 25% high-cost
50% low-cost 50% high-cost
25% low-cost 75% high-cost
100% high-cost
n/a
6.2 For each of the procurement types listed below, please indicate which strategy reflects your SOURCING PRACTICE best in terms of (1) geographic supply structure and (2) the number of suppliers involved. (1) Geographic supply structure Most sites supplied by local sources within own region
Certain sites supplied by local sources within own region
Most sites supplied by sources outside own region
(2) Number of suppliers
n/a
Single source
Limited sources
Multiple sources
n/a
MRO supplies* (i.e. maintenance, repair, operating mat.) Raw material Simple parts & high volume standard components Complex parts Core components
Performance Factors 7.1 Finally, please provide us with some information about the performance of your manufacturing network. Please keep in mind always to answer the questions from your manufacturing network perspective. From today´s perspective, please rate your SALES PERFORMANCE according to the following items. Average return on sales (EBIT-Margin*)
< 5%
5 - 10%
11 - 15%
16 - 20%
21 - 25%
26 - 30%
> 30%
n/a
Average return on capital (ROCE*)
< 5%
5 - 10%
11 - 15%
16 - 20%
21 - 25%
26 - 30%
> 30%
n/a
Service sales / total sales
< 5%
5 - 15%
16 - 25%
26 - 35%
> 35%
n/a
0 - 20%
21 - 40%
41 - 60%
61 - 80%
81 - 100%
n/a
Sales won through price competition / total sales
218
APPENDIX B: QUESTIONNAIRE 7.2 From today´s perspective, please rate your COST PERFORMANCE according to the following items. Manufacturing costs / total sales
< 30%
30 - 45%
46 - 60%
61 - 75%
> 75%
n/a
Warranty costs / total sales
< 5%
5 - 15%
16 - 25%
26 - 35%
> 35%
n/a
Total R&D expenditures / total sales
< 1%
1 - 5%
6 - 10%
11 - 15%
> 15%
n/a
External R&D expenditures / total R&D expenditures
< 5%
5 - 20%
21 - 35%
36 - 50%
> 50%
n/a
Total overhead costs / total costs*
< 10%
10 - 20%
21 - 30%
31 - 40%
> 40%
n/a
7.3 From today´s perspective, please estimate your MANUFACTURING COST STRUCTURE (should add up to 100%). Material costs / total manufacturing costs*
%
Direct labour costs / total manufacturing costs
%
Manufacturing overhead costs* / total manufacturing costs
%
∑
0%
7.4 From today´s perspective, please estimate the share of DISTRIBUTION COSTS. Total distribution costs* / total manufacturing costs
< 1%
1 - 5%
6 - 10%
11 - 15%
> 15%
n/a
7.5 From today´s perspective, please rate the degree of BUNDELING OF ACTIVITIES according to the following statements Bundeling of production volume per product Bundeling of similar products in the network Bundeling of support functions* in the network
Production volume scattered on many sites
Production volume concentrated on single site
n/a
Similar products scattered on many sites
Similar products concentrated on single site
n/a
Every site performs support functions
Central support functions concentrated on single site
n/a
7.6 With respect to your current network set-up, please rate the impact of following factors on your business Socio-political factors
(e.g. taxes, subsidies, customs, unions)
Image factors (e.g. "made in …", local for local)
Constrain our business
Have no influence on our business
Are exploited & facilitate our business
n/a
Constrain our business
Have no influence on our business
Are exploited & facilitate our business
n/a
7.7 From today´s perspective, please rate your DELIVERY & ORDER PERFORMANCE according to the following indicators. Average order lead time* / desired customer order lead time
Shorter
Slightly shorter
Balanced
Slightly longer
Longer
n/a
Minimum order lead time compared to average order lead time
Equals average lead time
Up to 10% faster
Up to 30% faster
Up to 50% faster
More than 50% faster
n/a
Average delivery reliablity* (on time in full)
< 75%
75 - 85%
86 - 95%
96 - 98%
> 98%
n/a
< 1%
1 - 5%
6 - 10%
11 - 15%
> 15%
n/a
No. of total annual customer complaints* / Number of annual orders
APPENDIX B: QUESTIONNAIRE
219
7.8 From today´s perspective, please rate your NETWORK PERFORMANCE according to the following indicators. Productivity differences between sites
No productivity difference
Deviation of gross margin* (as a percentage of sales) between sites
No difference
Average capacity utilisation (referred to typical
< 50%
50 - 70%
Very high productivity differences
71 - 85%
86 - 95%
n/a
Very high difference
n/a
> 95%
n/a
Very high difference
n/a
level of capacity availability)
Deviation of capacity utilisation between sites
No difference
Annual training hours per employee / total working time
< 1%
1 - 5%
6 - 10%
11 - 15%
> 15%
n/a
Annual average cost savings through continous improvement initiatives
< 1%
1 - 5%
6 - 10%
11 - 15%
> 15%
n/a
0 - 20%
21 - 40%
41 - 60%
61 - 80%
81 - 100%
n/a
< 1 week
1 week - 1 month
1 - 3 months
3 - 6 months
> 6 months
n/a
No alternative
Few sites
Certain sites
Many sites
All sites
n/a
Product innovations motivated by sites outside home region Ramp-up time of product transfer (transfering stable pro-duction from one site to another)
Economically alternative manufacturing sites per product
* Terms are defined in the glossary © 2010 Transfer Center for Technology Managem ent (TECTEM), University of St. Gallen
Excerpt of selected questions of X-GO survey questionnaire
20 / 20
220
B.2
APPENDIX B: QUESTIONNAIRE
Measures for the Calculation of the Network Capability Level & Conformance
Network capability items KPIs / Items evaluated in the survey
1. Accessibility Markets / Customers Assure access to strategic markets and competitive factors, like …
Competitors
Socio political factors Exploitation of socio political factors Image Supplier / Raw material
Exploitation of image factors Share of local sourcing volume Share of low-cost sourcing volume
Best cost labour
Assure access to resources of strategic importance, like … Skilled labour
External know-how
Initial type of measure*
∑ (share of total sales per region [%] x Average import ratio of final products per region import quota of final products per region [%]) Quality and extent of market or customer scale [defined steps] information provided by sites Total distribution costs / scale [% in defined steps] total manufacturing costs Quality and extent of competitor scale [defined steps] information provided by sites scale [defined steps] scale [defined steps] ∑ (sourcing volume per non-finished goods type x share of local sourcing per non-finished goods type) ∑ (sourcing volume per non-finished goods type x share of low-cost sourcing per non-finished goods
Share of employees located in low-cost countries % Deviation of gross margin between sites
scale ["no" to "very high"]
Productivity differences between sites
scale ["no" to "very high"]
Annual training hours per employee / total working time External R&D expenditures / total R&D expenditures
scale [% in defined steps] scale [% in defined steps]
2. Thriftiness ability Economies of scale
Increase efficiency by …
Economies of scope
Reduction of duplication
Manufacturing overhead costs / total manufacturing costs Bundling of production volume per product in the network Manufacturing overhead costs / total manufacturing costs
% scale ["scattered" to "concentrated"] %
Bundling of similar products in the network
scale ["scattered" to "concentrated"]
Total overhead costs / total costs
scale [% in defined steps]
Bundling of support functions in the network
scale ["scattered" to "concentrated"]
Ramp-up time for product transfer
scale [defined steps]
Economically alternative manufacturing sites per product
scale [defined steps]
Average capacity utilisation
scale [% in defined steps]
Deviation of capacity utilisation between sites
scale [% in defined steps]
3. Manufacturing mobility
Provide mobility of …
Products, processes, personnel Production volume & orders
4. Learning ability Explore and External factors exploit know-how and innovation Internal factors about …
Product innovations motivated by sites outside domestic region Annual cost savings through continuous improvement initiatives
* All measures were either directly asked on or transformed to a 1 to 5 scale
scale [% in defined steps] scale [% in defined steps]
APPENDIX B: QUESTIONNAIRE
B.3
221
Measures for the Calculation of the Performance Level on Strategic Manufacturing Priorities
Strategic manuf. priorities Price / Costs Quality
Delivery
Flexibility
Innovation Service
KPIs / Items evaluated in the survey
Initial type of measure*
Markets / Custom.
Sales won through price competition / total sales
scale [% in defined steps]
Specification quality
No. of total annual customer complaints / number of annual orders
scale [% in defined steps]
Conformance quality
Warranty costs / total sales
scale [% in defined steps]
Speed
Average order lead time / desired customer order lead time
scale [defined steps]
Reliability
Average delivery reliability (on time in full)
scale [% in defined steps]
Product range / Design Flex.
Number of product lines sold
scale [defined steps]
Average product line variety
scale ["very small" to "very high"]
Average number of annual design changes
scale [defined steps]
Minimum order lead time compared to average order lead time
scale [defined steps]
Total R&D expenditure / total sales
scale [% in defined steps]
Service sales / total sales
scale [% in defined steps]
Order size / Delivery Flex. Product & process innovation
* All measures were either directly asked on or transformed to a 1 to 5 scale
222
CURRICULUM VITAE
Curriculum Vitae Name:
Andreas Gerhard Mundt
Date and place of birth:
3rd of June 1980 in Alzenau (GER)
Nationality:
German
Practical Experience 2009 – 2012:
University of St.Gallen, St. Gallen (SUI) Institute of Technology Management, Chair of Production Management Group Coordinator & Research Associate
2008 – 2009:
Schaeffler Technologies AG & Co. KG, Herzogenaurach (GER) Specialist Continuous Process Improvement Manufacturing & SCM
2007:
Porsche Consulting GmbH, Bietigheim-Bissingen (GER) / Parma (IT) Internship: Lean Development & Production
2004:
Dr. Ing. h.c. F. Porsche AG, Hemmingen (GER) Internship: Product Cost Controlling Porsche Cayenne
Education 2009 – 2012:
University of St.Gallen (HSG), St. Gallen (SUI) Doctoral Studies in Business Innovation
2001 – 2008:
Technische Universität Darmstadt, Darmstadt (GER) Mechanical Engineering & Business Administration (Dipl. Wirtsch. Ing.)
2005 – 2007:
Linköping University, Linköping (SWE) International Master’s Program in Manufacturing Management (M.Sc.)
1991 – 2000:
Franziskanergymnasium Kreuzburg, Großkrotzenburg (GER) Abitur (German A-Level Equivalent)