in

Keywords:Value chain

es sg aut

aspects of value creation and appropriation, and the operational supply chain view that considersstrategies and tools for designing and operating efcient inter-rm networks. Commonly these views donot interact: value creation has the aim of capturing the maximum value-added in nancial terms, the

ain MManaghristop

morphology according the type of product being supplied

literatures wheretural aspects ofin rms, but may

ongitudinal casetecture, i.e. thee see the valuess architectures,

1995). In proposing a value chain architecture we seek to build a

Contents lists available at ScienceDirect

w.e

Int. J. Productio

Int. J. Production Economics 147 (2014) 230238work seeking to link product and process architectures withE-mail address: phelo@uva. (P. Helo).(Lamming et al., 2000), a general classication of value chain bridge between the research on value creation strategies in thesupply chain context (Normann and Ramirez, 1993; Pil andHolweg, 2006), and the work on supply chain design (Beamon,1998; Meixell and Gargeya, 2005). Our paper builds on earlier

0925-5273/$ - see front matter & 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.ijpe.2013.06.015

n Corresponding author. Tel.: +358 505 562668.While the overall language is well developed, and there is insur-mountable evidence that value chains differ substantially in their

where such purposeful design has been proposed before(Baldwin and Clark, 1997; Hayes and Wheelwright, 1979; Ulrich,Awealth of studies and reviews, too numerous to cite here, coupledwith the remarkable successes of companies such as Wal-Mart,Amazon, Dell, Toyota, Zara and the like, provide an unequivocalcase that a company's ability to design and manage its value chainhas become a cornerstone of contemporaneous management. Themanagement literature covers these cases well, as for example inthe cases of Dell's build-to-order value chain strategy (cf. Fugateand Mentzer, 2004; Holweg and Pil, 2001; Kapuscinki et al., 2004)or Zara's quick response manufacturing (Ferdows et al., 2004).

Commonly these views do not interact. We argfundamental gap in bridging the strategy and OMwe lack conceptual understanding of the strucwhy these supply chain structures work for certanot for others.

In this paper we are using an exploratory, lstudy to investigate the analogy of an archipurposeful design of a value chain. Conceptually wchain as an equivalent to product and proceno longer compete but entire value chains (Christopher, 1998). and tools for designing and operating efcient supply chains.ue that this is a1. Introduction

The rapid evolution of Supply Chsub-discipline within Operationsstrong evidence that, as Martin Csupply chain view aims for designing operationally efcient supply chains. In contrast to their treatisewithin the academic literature, from a practical point of view these two aspects are both necessary (andthus in their own right insufcient) components to a rm's supply chain strategy. In this paper we thusturn to an exploratory case study to identify what such a combined view of the value and supply chainwould entail. We refer to this purposeful creation as the value chain architecture and propose vefundamental decisions that dene the latter: (1) the nature of value provision (driven by the corecompetence of the rm), (2) the operational footprint decisions for manufacturing, sourcing anddistribution, (3) the approach to risk management, (4) the order fulllment strategy (and implicit inthat, the type of product customization), and (5) the buffering strategy. We conclude with an explorationof the application and utility of the value chain architecture concept in both academia and practice.

& 2013 Elsevier B.V. All rights reserved.

anagement (SCM) as aement (OM) providesher put it, companies

architecturesat the rm levelis sill amiss. In fact three decadesinto SCM research there remains an interesting dichotomy betweenthe value chain view that considers aspects of value creating,appropriation and nancial aspects of the supply chain, and theoperational supply chain view that largely considers strategiesSupply chain managementOperations strategyDening value chain architectures: Linkto operational supply chain design

Matthias Holweg a, Petri Helo b,n

a Judge Business School, University of Cambridge, UKb Department of Production, Faculty of Technology, University of Vaasa, Finland

a r t i c l e i n f o

Article history:Received 31 May 2012Accepted 13 June 2013Available online 28 June 2013

a b s t r a c t

Over the past three decadstrategic aspect of designinhowever one cannot help b

journal homepage: wwg strategic value creation

cholars have developed comprehensive insights into the operational andnd managing the supply chain. Reviewing this ample body of knowledgenotice a persistent disunion between the value chain view that considers

lsevier.com/locate/ijpe

n Economics

2. Literature review

2.2. Supply chain or value chain?

to talk about binary classications such as modular versus

M. Holweg, P. Helo / Int. J. Production Economics 147 (2014) 230238 231The original term supply chain management had a veryoperational connotation when Booz & Allen consultants Oliverand Webber rst proposed it in 1982 by telling rms to managetotal supply chain inventories (cited in Christopher, 1992). To thisdate this operation's focus has remained, aiming for a design andoperating stable and efcient supply chains. On the other handthere is the value chain concept originally proposed by Porter,which takes a nancial view of the sequential value creationprocess in a network of rms (Porter, 1985). Hence, even thoughoften used synonymously, there is a specic difference in perspec-tive on the same phenomenon: value is created in sequential stepsby a set of distinct rms.

The aspect of value creation and appropriation has beenintroduced as value constellation and value grid (Normannand Ramirez, 1993; Pil and Holweg, 2006), suggesting that rmsare well advised by going beyond the linear view of managingdyadic supplier relations. In fact, as the cases cited by these twoaforementioned studies show, considerable additional value canbe captured by thinking across value streams, as well as bythinking how managing or inuencing other tiers than theimmediately adjacent ones can yield considerable benet to2.1. A need for more denitions?

As Gibson et al. (2005) argue, SCM has evolved over time in thesame way as all academic elds evolve as the eld evolves, so dothe related key concepts under investigation. In this sense onemight rightfully argue that yet another paper on the (natural)evolution and of the supply chain concepts adds little value to theadvancement of the eldgiven that a set of comprehensivereviews on all aspects of supply chain management is available(Barney, 2012; Beamon, 1998; Bertrand, 2003; Burgess et al., 2006;Christopher and Ryals, 1999; Cooper et al., 1997; Holweg and Pil,2008; Lambert et al., 1998). We sympathize with this argument,however would argue that a key gap remains in the disunion ofbridging the operations-strategy gap in SCM. This review will thusrst and foremost focus on this aspect.supply chain design (Fixson, 2005a; Rungtusanatham and Forza,2005). We also draw upon the initial quantitative descriptions ofsupply networks (Harland et al., 2004; Harland, 1996; Lamminget al., 2000), and build on this prior work along three dimensions:rstly, we extend the often restricted focus of supply (chain)management beyond the immediate supplier tiers, to encompassthe entire value chain, in other words, the system that includes allvalue-adding steps, from raw materials to the distribution systemthat delivers the product or service to the end customer. Secondly,in addition to the mostly qualitative description of value chainarchitectures, a combination of qualitative and quantitative mea-surements and descriptors of a value chain conguration ispresented. Thirdly, we argue that a key tenet of dening valuechain architectures is to consider the alignment between process,product and supply network conguration.

The paper is organized as follows: in Section 2 we willintroduce the distinction between a value chain and a supplychain view in more detail, before introducing the method used inthis case study in Section 3. The case ndings are presented inSection 4, followed by the discussion of key ndings and implica-tions in Section 5.the rm.integral value chain structures. Another factor is the fact thatthe structure of a value chain determines its dynamic behavior, asfor example in the bullwhip effect shows, where removing anechelon has strong benets for the dynamics behavior (Forrester,1958; Lee et al., 1997; Metters, 1997).

The purpose dening a value chain architecture is to provide acommon language and structure for this mapping process byproposing ways in which to classify them. Such a denition isneeded for strategic planning and alignment of the elements, andthus should not only be of academic interest, but also haveimportant managerial implications.

It is implicitly known that the alignment of architectures is anecessary, but not sufcient condition for success. We only makethis explicit in product-process (Hayes and Wheelwright, 1979),The key assumption in this value view of the supply chain isthat rms can enhance their competitive position by consideringthe value streams they are operating in, as well as other parallelones that use the same supply and distribution and retail chains asa grid in which they operate. This assumption is fundamentallydifferent from the supply view, which argues that the most valuecan be added by ensuring the linear ows of information andmaterial occur with as little disruption as possible. Avoiding excessinventory, the bullwhip effect and long lead-times are key objec-tives here. As such these assumptions, as we will argue here, arenot contradictory but in fact complimentary. They have, however,so far not been merged into one conceptual framework. This papermarks a rst attempt in doing so.

2.3. Strategy or architecture?

A central question one might ask is: do we need another termor denition? Is not all of this part of supply chain strategy?In our view it is only partly so. An architecture is dened as aunifying or coherent form or structure in the English language,and we apply exactly this notion of a purposeful structure to thedesign of supply chains at rm level. We hereby consider twoobjectives: to capture maximum value, with the least operationalinefciencies in providing the former.

The value chain architecture thus is essentially a function of theoperations strategy. The decisions of the location of manufacturingoperations, the sourcing patterns, and the conguration andcustomization of the products are all questions that reside withinthe realm of the wider business or operations strategy. Thedifferential benet of determining the value chain architecturethat can be seen here is the ability to systematically describethe design of large network structures of value chain partners, andset overall objectives for the whole network or specic processes.In general, architecture describes the concept of creating an actualor apparent plan of a complex system, where it basically descri-bes a subjective mapping of the elements or components ofthe system, which considers the relationships among thesecomponents.

We hence propose the term architecture as an analogy to theway in which the term product architecture has been used todistinguish between modular and integral, as well as open andclosed product architectures (Ulrich, 1995), or process architec-tures to determine the appropriate manufacturing process inrelation to volume and variety (Hayes and Wheelwright, 1979).While both product and process architectures have been widelydiscussed, the complexity inherent in value chains (multiplelevels, network structure, dynamic and static complexity) haveso far meant that value chain architectures have not beenclassied in the same way. In many ways it is simply not possibleand product-value chain (for example modularity). The 3D model

versus point-to-point connections can lead to different perfor-mances because of structural changes. Also the systems applyingmerge-in-transit type of logistics designs can be perceived asstructural changes in network. Some sourcing models have astructural aspect, too. For example the purchasing decision touse policy of single supplier versus using multiple suppliers has areliability aspect in addition to costs. However, in our review of therespective literatures we found that these terms are isolateddescriptors only and do not match the need for a full description.For this reason we will propose a ve-factor model based on thesourcing strategy, the product and manufacturing strategy, thedistribution and customer interface, the customization and orderfulllment strategy, and the buffering mechanism used.

The material ow part of the proposed classication includessourcing, product type and customer aspects by varying local andglobal aspects. The key attributes are local and global which areperceived from the factory point of view. For example, localsourcing stands for suppliers located nearby manufacturing. Anexample of this could be the Zara clothes manufacturing in Spain,

M. Holweg, P. Helo / Int. J. Production Economics 147 (2014) 230238232makes this more explicit; however, there is a lack of a concisedenition of value-chain architecture.

This is important when considering rms and their success insupply chain management: Dell, Zara, Toyota all have engineeredsupply chains based on strategic choice. The key justication fordening value chain architectures is the need to acknowledgestrategic decisions that connect the structure and performance ofthe value chains. Strategic choices start from requirements, such asa competitive lead-time or cost-efcient delivery. These require-ments are then implemented by modifying the value-chain struc-ture. Implementing a build-to-order strategy, quantity exibilitycontracts, early-supplier involvement or hub-and-spoke distribu-tion logistics are examples of such actions. The actions change theperformance and these results can be compared to originalrequirements.

In this paper we thus dene the value chain architecture as aconscious design of the network structure consisting of suppliers,manufacturers, distributors and customers in order to maximumthe value creation for the focal rm. This denition takes intoaccount the material ow aspect such as location issues andtransport decisions, information ow and control part as well asthe dynamic nature of the chain. In this respect it is different fromprevious concepts such as strategic network design, which onlycovers long-term strategic issues such as sourcing and outsour-cing, and supply chain management in general, which often hasthe connotation of being limited to the operational execution only.

The management literature implicitly already classies value chainarchitectures by discussing specic aspects in isolation, such as orderfulllment and product customization strategies (Gilmore and Pine,1997; Pil and Holweg, 2004), sourcing congurations and supplierrelations (Dyer, 1998; Hines, 1997), global sourcing and outsourcing(Handeld, 1994; Monczka and Trent, 1991), and the like. Furthermore,as outlined above, the need to align product, process and value chainlayout has been proposed. Thus the question arises as to whether amore comprehensive classication would contribute to either theacademic debate or industrial practice.

The contributions to both areas can be made by providing amore rigid denition than available at present, as it would allowfor a more focused interpretation of the root causes for the successor failure of certain value chain strategies, and with respect toindustrial practice, provide a much stronger guideline as how toachieve this rather elusive three-dimensional alignment. Verybasic yet strategically important questions such as why no one hasyet managed to replicate either Dell's or Zara's value chain model,are currently not being addressed to any satisfactory detail. Also,the advise that practitioner's can derive from the value chainresearch to date often is very detailed (e.g. with regards toestablishing long-term collaborative relationships), and lacks thehigh-level overview that should guide any company's value chainstrategy.

2.4. The 3D alignment model

In the context of architectures in management it is interestingto note that value chains architectures are far less well dened aseither product or process architectures. Still, a recurring model isthe so-called 3D model, which argues that all three architecturesneed to be aligned for maximum competitiveness of the rm.Lamming et al. (2000) provide an initial impetus towards linkingvalue chain architectures and product architecture by developing aclassication of supply networks based on Fisher's (1997) distinc-tion of functional versus innovative products. A more explicit andrecent link is the discussion of the so-called 3D model, whichargues that the value chain, product and process dimensions needto be aligned for the organization to work at an optimal level (Fine,

2000a; Fine et al., 2005), see Fig. 1.The model argues that the nexus of competitiveness lies atthe point of alignment value chain, product and process dimen-sions, and that there are multiple workable strategies within the3D space in terms of process architecture, technology architectureand business systems / value chain architecture. The operationaland nancial performance of the value chain is depending on all ofthese internal parameters. We will return to this point later.

2.5. Summary

To date, the concept of value chain architecture has been usedin a limited context of structural decision making in the logisticsand distribution subsystems (Gow et al., 2002; Walker, 2005), andin conjunction with product and process architectures (Fine,2000b; Fixson, 2005b). And while value chain architectures maynot have been formally described, there nonetheless is an existingtypology that is often used to describe it. Some key decisionsrelated with the value chain architecture include the locationdecision, i.e. whether factories are classied into regional factoriessupplying local markets and those feeder factories supplying keycomponents for a large geographical area. This decision is con-nected to product structurehow product is decomposed intoparts, as for example the well-documented HewlettPackard InkJetprinter design case illustrates, where a separation of the electricaltransformer enabled use of high-volume feeder factory for theprinter part (Feitzinger and Lee, 1997). Furthermore, the transpor-tation related decisions, such as applying a hub-and-spoke model

Fig. 1. 3D model of aligning architectures within the rm.or Dell computers partnership with SCI Solectron. On the other

is located closer to customer may present problems in case of high

as we have worked with the rm from 2004 to 2009.

M. Holweg, P. Helo / Int. J. Production Economics 147 (2014) 230238 233variation or drastic demand changes. According to Holweg and Pil(2001), this issue might be the key reason behind the problems inthe automotive industry. The other part of control is the predo-minant buffering mechanism. In addition to stock, a company mayhedge against supply unreliability and market unpredictability byapplying time buffers or slack capacity. The exibility to adjustoperations to cope in uncertainty of volume, product mix changesand new product ramp-ups help in maintaining protability alsoin changing environments.

Another obvious approach to describe a value chain architec-ture would be to employ well-established value stream mappingtools (Gardner and Cooper, 2003; Hines and Rich, 1997; Rother andShook, 1999). This approach clearly is capable of identifying veryimportant aspects of a value chain architecture (where value iscreated, where it is destroyed), however, these approaches aregenerally concerned with operational improvement, rather thanthe strategic design of a value chain, and hence will not beconsidered in this paper.

3. Method

3.1. Research design

An exploratory case study approach was used to capture thecomplex contextual issues of value chain architecture and to demon-strate a proposed framework as a longitude descriptive case study.Case-based research is widely used in the eld of operationsmanagement studies related to theory building (Eisenhardt, 1989;Flynn et al., 1990; Meredith, 1998; Voss et al., 2002). For the purposeof this paper we combine qualitative data based on interviews withkey stakeholders of the SCM Department in the focal case rm, withsupporting quantitative secondary data based on internal rm dataon supply network structures and performance. We report on alongitudinal case study which allows us to comment on the evolu-tion of the value chain architecture at ABB from 2004 to 2009.

3.2. Case selection

In order to capture value chain architectural decisions in a largescale company and its network, a global manufacturer of energyappliances was selected, the ABB Company. ABB presents an interest-hand global suppliers are used in electronics business, wherecomponents are generic or there are only few manufacturers ofkey components available. The product design/manufacturing partdescribes how products are tailored to regional markets from thefactory perspective. For example many electrical appliances needto be designed for various voltages to comply with local safetyregulations. Global products are in many cases highly standardizedor very exible generic products that can be used in multipleregions. The downstream or distribution aspect of the chaindescribes the location of customers from the factory. Local man-ufacturing may lead to improved time-to-customer whereas globalexport type of supply from a limited number of factories mayincrease the volume and improve the cost effectiveness.

Yet the material ow aspect does not give a complete picture ofthe network architecture. The information ow part is equallyimportant since it is dening the focal point of carrying inven-tories. The order fulllment strategy and point of customizationare related with the operational policy, whether the chain isoperated by make-to-stock, assembly-to-order, make-to-order orfully customized by engineering-to-order. The point of customiza-tion and product range is dened by this important productdecision. Make-to-stock type of operations where the inventorying context of a global manufacturing company that is builtABB is a global company supplying a wide range of products inthe power and automation technologies sector. Power products areused to distribute electricity with transformers, switchgear, circuitbreakers, cables and associated equipment. These components areused in building large systems consisting of power grids, substations,and generation. The automation side of the company is producingcomponents such as electric motors, frequency converters, powerelectronics and other low voltage products. In addition to these, ABBsupplies specic industries, such as oil and gas, marine, pulp andpaper, in the eld of automation solutions. ABB can be considered asa large multi-national company with revenues of US$ 31billion(2011) and more than 130,000 employees. The company operatesin more than 100 countries and has more than 150 factories in vecontinents. The global presence of the company has resulted fromseveral mergers and frequent structural changes. Initially ABB wasformed at merger of Swedish ASEA and Swiss BBC Brown Bovericompanies in 1988. Both the companies hold 50% of the shares.At that time ABB employed 170,000 employees.

3.3. Data collection

The data collection included several types of interaction: semi-structured interviews, data collection on-site at the global fac-tories, and participant observation in development projects withinthe ABB Corporate Research Center between a six-year periodbetween 2004 and 2009. This work includes interviews with12 factory managers, 3 development managers, and 4 internalconsultants. The interviewees have been selected because of theirdirect involvement in structural changes of ABB's network in termsof designing and updating the feasible network structure relatedto distribution, supply, inventory location and buffering methods.

3.4. Validity and limitations

Validity of exploratory case study results depends on constructvalidity, face validity (external business executive review), internalvalidity (number of informants within the company), externalvalidity (replication of company interviews), and reliability (pro-tocols and triangulation) (Eisenhardt 1989). The internal validityhas been taken into account by interviewing both factory man-agers as well external consultants working within ABB at the time,providing a non-internal point of view to guard against biasedreporting. However, it must be acknowledged that the perspectivemay still be biased towards an operations development consultantpoint of view not taking into account business unit specic issuessuch as competition in products and technology. Supportingdescriptive quantitative data has been collected to demonstratesome behavior in development. Overall we argue that the limita-tions our research design incurs are those any case study wouldhave. The ndings below should be considered in this context.

4. Case ndings

4.1. Value chain design phases

During the past decades ABB has gone through several phasesin terms of value chain structure over the six year period of thisnon-organically from mergers and systematic acquisitions. As such,the value chain architecture did not grow out of a coherent set ofdecisions over time, but grew with the acquisitions. This contextprovides for a rich background for our research question. The case ofABB is also well suited to show the connection between strategicdecision making and formulation process of value chain architecture,case study. This does allow us to study and compare multiple

decision making instances over time: which criteria have beenused, how these criteria have been weighted in the decision, andwhat data is being consulted. We have observed considerablechange in the way that ABB has approached value chain architec-tural decisions, and will outline these chronologically.

Historically ABB has made about 15 major acquisitions mainlyin Europe and then continuing aggressively with 40 more compa-nies in 1989, including the purchase of Westinghouse ElectricCorporation power transmission and distribution operations inthe US. In the early 1990's the company faced problems inNorthern-Europe because of economic recession. The focus of

the local customers with their own products. This architectureworked well in a situation where a new market-player aimed toreach new customers and distribution partners for products.However, many operations have required the same work to bedone for each local product and the cost-efciency suffered.Products were stored in many cases as nished goods inventoryand the distribution side cannot use the risk pooling advantage.

The main reasons for changing network structures were marketconditions. In the early stages, the market presence was importantbecause the electrical utility market was very much national andin hands of monopolistic companies. When the legislation chan-

pancal m

re.

nce

andets.

M. Holweg, P. Helo / Int. J. Production Economics 147 (2014) 230238234expansion moved to building new companies in Eastern-Europeand restructuring the overall corporate structure. The rate ofacquisition slowed down signicantly.

In the mid 1990s, ABB started to build new networks inemerging markets of Asia. The operations in western worldEurope and USconcentrated on service and retrot businesses.The corporate objective was to increase prots and consolidate fornew volume. These operations included new joint-ventures.

Towards the late 1990's the Asian markets started to grow aswell as business areas in the oil and gas industries. At the sametime ABB started a major cost cut in the Western operations. Morethan 12,000 jobs were cut and investments were focused on Asia.At the beginning of new millennium ABB continued with acquisi-tions. A 5050 joint venture with French Alstom Group wasformed to create worlds largest power generation company. Atthis stage ABB discontinued its nuclear power business and soldstandard copper cable manufacturing. In the 2000s ABB wasfocusing on growing markets of Asia: new factories are built inChina and India, but also in South-America. Local presence ofoperations is required to ensure global growth.

The products, market areas, and supply chains have changedvery much during the formation of ABB. The design process forsupply chain architecture has varied focus from 1990's cost cuttingand inventory reduction to 2000s time-based objectives. The valuechain architecture has always been close to the strategy processand its objectives. From a historical perspective the developmentof the architecture can be described in four stages as described inTable 1 below.

4.2. The case of the electrical machine division

In order to capture what is happening on more detailed levelwe will focus on one division of ABB that we will refer to asElectrical Machines. This business area consisting of severalfactories is producing systems for electricity transmission. Theannual revenue of the business area is approximately 1 billionUSD. The products are classied in three categories: low voltage,medium voltage and high voltage products. Over the years, thestructure of this business area has changed from localized pro-ducts and local manufacturing towards truly global supply chainnetworks. As seen in upper part of Fig. 2, initially, the manufactur-ing units have been sharing the same brand names and supplying

Table 1The four phases of the value chain architecture at ABB.

Phase Objectives

The beginning 19891990. Forming a giant multi-national comautomation businesses.Access on lo

Restructuring. 19911996 Simplify the product/process structu

Dispersed worlds 19971999 Cost cutting in EuropeAdding preseon emerging markets.

Global supply chains. 2000present Simplication of network structureproduct offerings for different markged during the early 1990s in Europe, a new type of organizationmodel was suited better to the situation. The big change inelectrical machines took place in the late 1990s when the focusedfactory concept drove network design. The objective was tointroduce a common product portfolio, reduce production costsand apply common processes. The roadmap for this wassequenced accordingly. A common product portfolio is a prerequi-site for all improvements. This stage needs key decisions frommarketing and product design. Once common products are intro-duced, the manufacturing units can achieve the envisaged costreductions. Common process organization models follow theseand offer a similar type of interface for all customers globally. Eachpart of this transition period took about two years and the wholetransition took four years.

The consulting arm of the ABB Corporate Research Center (partof the Operations Development Group) received a request tosuggest a new and improved architecture and justify its new valuechain architecture with customer and nancial benets. Thesuggested new structure included several changes in productdesign to support global products and use of common focusedsuppliers. The new suggested structure as shown in lower part ofFig. 2. Here the customers, which are regional sales companies andABB project units are on the right hand side of the gure. Thesecompanies are purchasing products from regional factories, whichare producing in most of the cases the whole range of products forthe customers. The regional factories are assembling the productsfrom components provided by feeder factories, which supplyglobally. The manufacturing of low variety key components isfocused on key manufacturers within the ABB. From a valueperspective, the changes aim to centralized global product respon-sibilities and improved efciency by using global sourcing.

In the revised architecture, from the inventory point of view,most of the inventory is stored at the components level. Theregional factories are mostly assembly plants operating based oncustomer orders. The feeder factories are operating based onmake-to-order for larger project orders and assembly-to-orderfor the other demand. The key components in the networkupstream are made-to-stock due to a smaller number of variations.Fig. 2 illustrates the created structure. The new improved structureresulted lower inventory in value adding products and risk poolingeffect for standardized key components. The overall effect on protand costs was as expected. However, the structural changes

Actions

y in power andarkets

The merger.Acquisitions to get access on local markets.Matrix organization.

Focused factories.Stop overlapping products.

Centralized distribution warehouses.New factories inEast-Europe and Asia.

Global product responsibilities for factories.Focus onlead-time.

M. Holweg, P. Helo / Int. J. Production Economics 147 (2014) 230238 235affected some other performance parameters in the supply net-work as well: rstly, the risk managing aspect was changed due tosome key components now being made in a single factory.Secondly, in this type of network structure, the capacity uctua-tion changes the bottleneck location. In some cases, the regionalfactories may end up competing from the same time-slots of com-ponent suppliers. Additionally, lead-time improvements requirehigh reliability from the suppliers due to make-to-order principles.

In order to design the new supply network architecture and tojustify the nancial investment, ABB Corporate Research Centerbuilt a simulation model of each alternative. The customers,distribution warehouses, factories and suppliers were modeledby using a rapid modeling tool, and the internal consulting teamtogether with business managers developed structures whichcould be evaluated against the current network performance andagainst several future alternatives. As suggested by one of the leadconsultants the process of creating the next stage of network is notabout nding an optimized network structure based on a singlecriterion, but a feasible improved model, which can be phased intosmaller development steps. Simulations can suggest good input fordecision-making, but the actions need to be evaluated case by case.Another aspect of the VCA process was pointed out by a factorymanager the business objectives may change and the new proposedstructure needs to be robust for changes.

The experiences from ABB network architects suggest that thistransition period may be connected to the ability to change overproduct generations. This factor describes the clockspeed foreach business area (Fine, 2000a). In case of long life-cycleproducts, such as heavy machinery, this cycle could take muchmore time than in faster-paced electronics and related industrieswhere product changes and updates are expected more frequently.Also product complexity and system wide interactions to other

Fig. 2. Improved supply chain architecture with local mproducts might hinder the speed of transition. In order to describethe impact of the architecture change at Electrical Machines, wesummarize the product and manufacturing data series fromTable 2 below which shows one product line and its developmentsduring a period of ve years.

At the same time the volume of this family went up and theproductivity steadily increased. The concept of the transitionseems to follow a sand-cone model type of pattern. The rstpriority is to x product line, then introducing platforms andreduce costs, and nally it will be followed by process improve-ments. Each of these elds are well known in industrial engineer-ing literature, but the suggested sequential connection has notbeen proven. The last stagene-tuning the common processeswas referred in one interview as copypaste manufacturingfootprint, which describes the process interestingly in a globalmulti-factory scope.

4.3. Synthesis

The description of our case rm ABB highlights many impor-tant aspects of decision-making related to value chain architec-tures, and their respective performance. First and foremost we

arket presence and concentrated focused factories.

Table 2Product development projects in the Electrical machines business unit.

Year Product families Parts Product lines Comment

1 18 80,000 47 Local markets2 10 45,000 22 Platform development3 1 5000 12 Product design4 1 5000 12 Sourcing global5 1 5000 9 Global markets

note that the decisions taken have considered predominantlyoperational issues, such as manufacturing and sourcing footprints,control principles within the supply chain, as well as productvariety related matters. These decisions fall largely into thefundamental choices in strategy: make or buy (or outsourcing),the footprint decisions (where to manufacture), the sourcingfootprint, the way products are customized, and how risk ismanaged. We combine these ve factors into a 5-point model ofa value chain architecture, highlighted in the case of ABB's threestages, as shown in Table 3.

We can see how this global enterprise has shaped its valuechain architecture. The three stages of electrical machines businessarea may be described in three major stages. The fully localbusiness, with large geographical presence, has been transformedto locally present regional factories, while the cost effectiveness ismaintained at the focused key component factories. According tointerviews the roadmap seems to lead towards truly globalfactories, where modularized products are supporting postpone-ment of variety.

More generally, from the ABB case we can identify ve domainsof a value chain architecture:

1. Value provisionseparation of product centric feeder factoriesand regional assembly plants changed the value creationlocation and specializes the roles with the chain. Centralizationof suppliers towards global vendors affected the value position-ing along the value chain.

2. Operations footprint is driven by market focus areas com-bined with cost-efciency, lead-time and other operationalperformance metrics.

3. Risk management aspects relate to local/global sourcing interms of having reliable capacity and capability to supplycertain components within the ABB network. Global supplyincreases the risk but improves cost efciency under stableconditions.

4. Order fulllment strategy and customizationchanging thelocation of order-penetration point affects the number of productvariants offered to customer as well as internal efciency. Change

Table 3Three stages of the value chain architecture development at ABB.

Initial Revised Current

1. Value provision Strong local presence Cost efcient Towards global optimality

2. Operations footprint Local suppliersLocalized products for each country

Global key componentsRegional products

Global key componentsGlobal products

3. Risk management Local customers Local customersLower inventories.

Global product responsibilitiesFaster lead-time

4. Order fulllment strategy andpoint of customization

Make-to-stock Make-to-order in regional factories Automated engineer-to-order

5. Buffering mechanism Finished goods inventory Key-components Capacity buffering

Table 4The ve determinants of a value chain architecture (VCA).

Factor Key decisions

1. Value provision What does the company want to focus on in terms of value creation?The essential question here relates back to the core competence of the rm, which drives what it makes in-house, versus what is

ourcringnnels asent custlso strur of souult

limit exmany oma

orders?nects thntory loe make-ts (

y?f mome,

M. Holweg, P. Helo / Int. J. Production Economics 147 (2014) 230238236outsources.

2. Operational Footprint Where does the company operate and sFootprint decisions including manufactuused in classication of distribution chaelement adds possibilities to treat differand vendor-managed inventories, adds aof view the key decisions are and numbeare made between single sourcing and m

3. Risk management What mechanisms does the rm use toRisk management in a SCM context usesand the like. Dual versus single sourcing

4. Order fulllment and productcustomization

How does the rm respond to customerThe point and type of conguration condecoupling point and thus the focal inveoperations models ranging between purthe product variety offered to the marke

5. Buffering mechanism How does the rm deal with uncertaintUncertainty is an omnipresent feature oused as buffer: inventory, capacity, or tidern business, a central question thus concerns the question what resources areor a combination of these.Mate product architecture with process structure. This decision drives the ordercation within the network or as sometimes referred as pivot point. Theto-stock and highly customizable engineer-to-order dene lead-time as well asher, 1988).y beposure?f the standard approaches in risk management, such as diversication, poolingone consideration, as is having multiple factories in key markets.iplewell. In addition to manufacturing location issues the distribution channelomer groups in different ways. Some logistics models, such as merge-in-transitctural parameters that allow reduction of inventory. From an architectural pointrces for each component and the location of sourcing. In practice, the decisionssuppliers policy as well as between use of local and global suppliers.e fr, souom?rcing, and distribution The aspects of global, regional or local operations may be

From the case we have derived ve dimensions of a value chain

Bur

M. Holweg, P. Helo / Int. J. Production Economics 147 (2014) 230238 237Stephen C., de Kok, A.G. (Eds.), Handbook in Operations Research and Manage-ment Science.. North Holland Publishing Company.gess, K., Singh, P.J., Koroglu, R., 2006. Supply chain management: a structuredliterature review and implications for future research. International Journal ofBertarchitecture, shown in Table 4. It needs to be recognized that awide range of other variables could have been considered, yet weargue that these variables stand out as the most important ones tobe considered. Also, there seems to not be any logical conictbetween qualitative and quantitative descriptors, as neither one inisolation provides a comprehensive description of the system, andin fact both can be used in a very complimentary fashion. Forexample, when discussing the risks of a sourcing strategy of a rm,providing system redundancy and risk pooling measures canprovide the empirical evidence that underpins the qualitativedescription.

Performance measures and metrics continue to be a crucialpart of supply chain management (Gunasekaran et al., 2001). Theproposed classication however does not state that one architec-ture type outperforms any others. Instead it addresses one of themost interesting questions: which architecture should be used inwhich case? This alignment question is of crucial interest to bothacademia and practice, and we propose the above model forfurther testing and expansion, towards a search for clusters ofcommon practice across supply chains. In other words, wepropose for future research that there may be similarities withincertain industries or settings is likely to show similar features,which in turn allows for the identication of best practicescongurations that can be directly translated into hands-on adviceto industry.

In conclusion, we have proposed a ve-point framework fordening a value chain architecture at the rm levelwhichmerges the nancial value chain view with the operationalsupply chain considerations. First and foremost this can serveas a guideline in supply chain management research answeringthe question "Which supply chain conguration is right for me?".As the ABB case has shown, value chain architectures are evolvingduring the time, as competitive priorities change. Hence the purelyoperational or efciency drive in SCM is clearly an unnecessarilylimiting view. The systematic and strategic processes behind valuechain decisions need to combine both value creation and opera-tional points of view, and in our view can only be describedthrough a combination of qualitative and quantitative metrics. Wehave proposed a ve-point conceptual model that attempts tobridge this very gap.

References

Baldwin, C.Y., Clark, K.B., 1997. Managing in an age of modularity. Harvard BusinessReview 75 (5), 8493.

Barney, J.B., 2012. Purchasing, supply chain management and sustained competitiveadvantage: the relevance of resource-based theory. Journal of Supply ChainManagement 48 (2), 36.

Beamon, B.M., 1998. Supply chain design and analysis: models and methods.International Journal of Production Economics 55 (3), 281294.rand, W., 2003. Supply chain design: exibility considerations. In: Graves,of product design responsibilities and product design can improvehow postponement can be implemented.

5. Buffering mechanism, which can be changed by dening thestrategic location of safety stocks and improving the efciencyof delivery performance within the factories close to customers.

5. Conclusion

As we have shown in our longitudinal case of ABB, value chainarchitectural decisions change over time. They adapt to external orcontextual changes in light with the evolving business strategy.Operations and Production Management 26 (7), 703729.Christopher, M., 1992. Logistics and Supply Chain Management. Strategies forReducing Costs and Improving Services. FT Prentice Hall, London.

Christopher, M., 1998. Logistics and Supply Chain Management. Pearson, Harlow.Christopher, M., Ryals, L., 1999. Supply chain strategy: its impact on shareholder

value. International Journal of Logistics Management 10 (1), 1.Cooper, M., Lambert, D.M., Pagh, J.D., 1997. Supply Chain Management: Strategies

for Reducing Costs and Improving Services. FT Prentice Hall, London.Dyer, J. H., Singh, H. (1998). The relational view: Cooperative strategy and sources of

interorganizational competitive advantage. Academy of management review,23(4), 660679.

Eisenhardt, K.M., 1989. Building theories form case study research. Academy ofManagement Review 14 (4), 532550.

Feitzinger, E., Lee, H.L., 1997. Mass customization at hewlett packard: the power ofpostponement. Harvard Business Review 75 (1), 116122.

Ferdows, K., Lewis, M.A., Machuca, J.A.D., 2004. Rapid-re fulllment. HarvardBusiness Review 82 (11), 104111.

Fine, C.H., 2000a. Clockspeed-based strategies for supply chain design. Productionand Operations Management 9 (3), 213.

Fine, C.H., 2000b. Clockspeedbased strategies for supply chain design1. Productionand Operations Management 9 (3), 213221.

Fine, C.H., Golany, B., Naseraldin, H., 2005. Modeling trade-offs in three-dimensional concurrent engineering: a goal programming approach. Journalof Operations Management 23, 389403.

Fisher, M.L., 1997. What is the right supply chain for your product? HarvardBusiness Review (2).

Fixson, S., 2005a. Product architecture assessment: a tool to link product, processand supply chain decisions. Journal of Operations Management 23, 345369.

Fixson, S.K., 2005b. Product architecture assessment: a tool to link product, process,and supply chain design decisions. Journal of Operations Management 23 (34),345369.

Flynn, B.B., Sakkakibara, S., Schroeder, R.G., Bates, K.A., 1990. Empirical researchmethods in operations management. Journal of Operations Management 9 (2),25284.

Forrester, J.W., 1958. Industrial dynamics: a major breakthrough for decisionmakers. Harvard Business Review 36 (4), 3766.

Fugate, Brian S., John T. Mentzer. "Dell's Supply Chain DNA."Supply Chain Manage-ment Review, V. 8, No. 7 (Oct. 2004), pp. 2024: ILL 8, no. 7 (2004).

Gardner, J.T., Cooper, M.C., 2003. Strategic supply chain mapping approaches.Journal of Business Logistics 24 (2), 3764.

Gibson, B.J., Mentzer, J.T., Cook, R.L., 2005. Supply chain management: the pursuit ofa consensus denition. Journal of Business Logistics 26 (2), 1725.

Gilmore, J.H., Pine, J., 1997. The four faces of mass customization. Harvard BusinessReview (1).

Gow, H.R., Oliver, L.D., Gow, N.G., 2002. Co-operating to compete in high velocityglobal markets: the strategic role of exible supply chain architectures. Journalon Chain and Network Science 2 (1), 1932.

Gunasekaran, A., Patel, C., Tirtiroglu, E., 2001. Performance measures and metrics ina supply chain environment. International journal of operations and productionManagement 21 (1/2), 7187.

Handeld, R.B., 1994. US global sourcing: patterns of development. InternationalJournal of Operations and Production Management 14 (6), 4051.

Harland, C., Zheng, J., Johnsen, T., Lamming, R., 2004. A conceptual model forresearching the creation and operation of supply networks. British Journal ofManagement 15, 121.

Harland, C.M., 1996. Supply chain management: relationships, chains and net-works. British Journal of Management 7, S63S80.

Hayes, R.H., Wheelwright, S.C., 1979. Link manufacturing process and product lifecycles. Harvard Business Review 57 (1), 133140.

Hines, P., 1997. A comparative typology of intercompany networking. In: Cox, A.,Hines, Peter (Eds.), Advanced Supply Management. Earlsgate Press, London.

Hines, P., Rich, N., 1997. The seven value stream mapping tools. InternationalJournal of Operations and Production Management 17 (1), 4664.

Holweg, M., Pil, F.K., 2001. Successful build-to-order strategies start with thecustomer. MIT Sloan Management Review 43 (1), 7483.

Holweg, M., Pil, F.K., 2008. Theoretical perspectives on the coordination of supplychains. Journal of Operations Management 26 (3), 389406.

Kapuscinki, R., Zhang, R.Q., Carbonneau, P., Moore, R., Reeves, B., 2004. Inventorydecisions in Dell's supply chain. Interfaces 34 (3), 191205.

Lambert, D.M., Cooper, M.C., Pagh, J.D., 1998. Supply chain management: imple-mentation issues and research opportunities. International Journal of LogisticsManagement 9 (2), 119.

Lamming, R., Johnsen, T., Zheng, J., Harland, C., 2000. An initial classication ofsupply networks. International Journal of Operations and Production Manage-ment 20 (6), 675691.

Lee, H.L., Padmanabhan, V., Wang, S., 1997. The bullwhip effect in supply chains.Sloan Management Review 38 (3), 93102.

Mather, H., 1988. Competitive Manufacturing. Prentice Hall, Englewood Cliffs.Meixell, M.J., Gargeya, V.B., 2005. Global supply chain design: a literature review

and critique. Transportation Research Part E: Logistics and TransportationReview 41 (6), 531550.

Meredith, J., 1998. Building operations management theory through case and eldresearch. Journal of Operations Management 16 (4), 441454.

Metters, R., 1997. Quantifying the bullwhip effect in supply chains. Journal ofOperations Management 15 (2), 89100.

Monczka, R.M., Trent, R.J., 1991. Global sourcing: a development approach. Inter-

national Journal of Purchasing and Materials Management 27 (2), 28.

Normann, R., Ramirez, R., 1993. From value chain to value constellation: designinginteractive strategy. Harvard business review, 71(4), 6577.

Pil, F.K., Holweg, M., 2004. Linking product variety to order fulllment strategies.Interfaces 34 (5), 394403.

Pil, F.K., Holweg, M., 2006. From value chain to value grid. MIT Sloan ManagementReview 47 (4), 7280.

Porter, M.E., 1985. Competitive Advantage: Creating and Sustaining SuperiorPerformance. The Free Press, Macmillan Publishing, New York.

Rother, M., Shook, J., 1999. Learning to See. The Lean Enteprise Institute, Brookline.Rungtusanatham, M., Forza, C., 2005. Coordinating product design, process design,

and supply chain design decisions: Part A: topic motivation, performance

implications, and article review process. Journal of Operations Management 23(34), 257265.

Ulrich, K., 1995. The role of product architecture in the manufacturing rm.Research Policy 24 (3), 419440.

Voss, C., Tsikriktsis, N., Frohlich, M., 2002. Case research in operations management.International Journal of Operations and Production Management 22 (2),195219.

Walker, W.T., 2005. Supply Chain Architecture: A Blueprint for Networking theFlow of Material, Information and cash. CRC Press, Boca Raton.

M. Holweg, P. Helo / Int. J. Production Economics 147 (2014) 230238238

Defining value chain architectures: Linking strategic value creation to operational supply chain designIntroductionLiterature reviewA need for more definitions?Supply chain or value chain?Strategy or architecture?The 3D alignment modelSummary

MethodResearch designCase selectionData collectionValidity and limitations

Case findingsValue chain design phasesThe case of the electrical machine divisionSynthesis

ConclusionReferences