CONFIGURING VALUE FOR COMPETITIVE ADVANTAGE: ON CHAINS, SHOPS, AND

Strategic Management Journal, Vol. 19, 413–437 (1998)

CONFIGURING VALUE FOR COMPETITIVEADVANTAGE: ON CHAINS, SHOPS, ANDNETWORKS

CHARLES B. STABELL* and ØYSTEIN D. FJELDSTADNorwegian School of Management, Sandvika, Norway

Building on Thompson’s (1967) typology of long-linked, intensive, and mediating technologies,this paper explores the idea that the value chain, the value shop, and the value network arethree distinct generic value configuration models required to understand and analyze firm-levelvalue creation logic across a broad range of industries and firms. While the long-linkedtechnology delivers value by transforming inputs into products, the intensive technology deliversvalue by resolving unique customer problems, and the mediating technology delivers value byenabling direct and indirect exchanges between customers. With the identification of alternativevalue creation technologies, value chain analysis is both sharpened and generalized into whatwe propose as a value configuration analysis approach to the diagnosis of competitive advantage.With the long-linked technology and the corresponding value chain configuration model asbenchmark, the paper reviews the distinctive logic and develops models of the value shop andthe value network in terms of primary activity categories, drivers of cost and value, andstrategic positioning options. 1998 John Wiley & Sons, Ltd.

Strat. Mgmt. J. Vol. 19, 413–437 (1998)

INTRODUCTION

Understanding how firms differ is a central chal-lenge for both the theory and the practice ofstrategic management (Nelson, 1991). In adynamic economic and institutional setting,changes in the dominant competitive logic offirms is of particular interest (Prahalad andHamel, 1994). Hence, a complete but parsimoni-ous typology of the alternative forms of valuecreation is a prerequisite for expressing andexploring how firms differ in a competitive sense.The purpose of this paper is to propose andexplore such a typology.

Porter’s value chain framework (1985) ispresently the accepted language for both rep-

Key words: value creation technologies; value con-figuration analysis; competitive positioning*Correspondence to: Charles Stabell, Norwegian Schoolof Management, Elias Smiths vei 15, P.O. Box 580,N-1301 Sandvika, Norway.

CCC 0143–2095/98/050413–25$17.50 Received 15 May 1995 1998 John Wiley & Sons, Ltd. Revised 4 November 1996; Final revision received 28 April 1997

resenting and analyzing the logic of firm-levelvalue creation. Although Porter’s industrialorganization (five-forces) competitive analysisframework (Porter, 1980) is challenged inresource-based critiques (Barney, 1991; Werner-felt, 1984), the value chain maintains its centralrole as a framework for the analysis of firm-levelcompetitive strengths and weaknesses.

Value chain analysis is a method for decom-posing the firm into strategically important activi-ties and understanding their impact on cost andvalue. According to Porter (1985, 1990), theoverall value-creating logic of the value chainwith its generic categories of activities is validin all industries. What activities are vital to agiven firm’s competitive advantage, however, isseen as industry dependent.

414 C. B. Stabell and Ø. D. Fjeldstad

Having supervised in-depth application of thevalue chain model to more than two dozen firmsfrom a variety of industries1 during the last 4years, we have experienced serious problems inapplying the value chain framework. Whereas theprimary activity typology of the value chainappears well suited to describing and understand-ing a traditional manufacturing company such aspresented in the familiar ‘Crown Cork and Seal’case (1977), the typology and underlying valuecreation logic are less suitable to the analysis ofactivities in a number of service industries.2 It isnot only difficult to assign and analyze activitiesin terms of the five generic primary value chaincategories, but the resulting chain often obscuresrather than illuminates the essence of value cre-ation.

Consider the insurance company. What isreceived, what is produced, and what is shipped?Few insurance executives would perceive unin-sured people as the raw material from which theyproduce insured people. Nor would a descriptionof an insurance company as a paper-transformingcompany, producing policies from blank paper,capture the value creation logic. This is not tosay that the logistics of handling paper and datain a large insurance company is a minor undertak-ing to those involved in it. Significant savingscan be realized by reengineering (Hammer, 1990;Schonberger, 1990) the document flow and con-siderable cost is incurred in operating theinsurance company’s computer systems. However,such a description hardly captures the essence ofvalue creation in an insurance company from astrategic point of view. The logic of many stra-tegically important activities such as reinsuranceto cover risk, actuarial calculations, and customerrelationship management are not well describedby a paper-flow-transformation-process perspec-tive.

Similar problems occur in the analysis ofbanks. Our experience is that value chain analysisfrequently results in either postulating deposits asthe ‘raw material’ that the bank’s primary activi-ties transform into loans, or postulating that allprimary banking activities collapse into a single

1 We have worked with, amongst others, insurance, banking,metal processing, telecommunication, health services, down-stream and upstream petroleum, engineering, and transpor-tation.2 For similar critiques see, for example, Løwendahl (1992);Armistead and Clark (1993).

1998 John Wiley & Sons, Ltd. Strat. Mgmt J., Vol19, 413–437 (1998)

major activity class: operations. In either case,the chain model cannot deal explicitly with bothlenders and borrowers as bank customers. Thevalue chain metaphor obscures the competitivelogic of banking by focusing attention on trans-action-processing unit costs, with little attentionto interest spread and risk management.

Slightly different problems occur when we tryto analyze in more detail critical support activitiessuch as technology development. Considerupstream petroleum exploration and field develop-ment. Value chain analysis directs too much atten-tion to unit costs, i.e., finding costs, developmentcosts, and production costs per barrel of oil.Although unit cost is a relevant performance mea-sure when we consider the complete life cycleof an oil field, it is less useful as a guide to theeconomics of exploration. Efficiency in explo-ration is subordinate to effectiveness. Upstreampetroleum is mainly the logic of extraordinaryvalue creation such as finding giant oil fields orcreating innovative field development conceptsthat alter the rules of commercial petroleum pro-duction. Value created seldom correlates withfinding costs. We need an analysis frameworkthat can handle the contingent nature of petroleumexploration and field development, where projectsoften require a custom approach and where mostexploration projects are not successful.

We suggest that the value chain is but oneof three generic value configurations. Based onThompson’s (1967) typology of long-linked,intensive and mediating technologies, we explorethe idea that the value chain models the activitiesof a long-linked technology, while thevalue shopmodels firms where value is created by mobilizingresources and activities to resolve a particularcustomer problem, and thevalue networkmodelsfirms that create value by facilitating a networkrelationship between their customers using amediating technology. Hospitals, professional ser-vice firms, and educational institutions areexamples of firms that rely on an intensive tech-nology. Examples of companies that create valueby facilitating exchange among their customersare telephone companies, transportation com-panies, insurance companies and banks.

Introducing three distinct value configurationsleads us to propose that value chain analysisneeds to be transformed into value configurationanalysis, which in turn helps us clarify criticalanalysis assumptions. Value configuration analysis

On Chains, Shops, and Networks 415

is defined as an approach to the analysis of firm-level competitive advantage based on a theory ofthree value creation technologies and logics. Thewell-known value chain ‘diagram’ serves both asan analytical tool for the analysis of value cre-ation in a specific firm and as a representationformat (Morecroft, 1992). The analysis serves asa means to develop an understanding of the cur-rent competitive position of the firm and howthis position can be both maintained and strength-ened; the firm is the unit of analysis. We presentalternative analytical representation and presen-tation formats that summarize the unique valuecreation logic of the intensive and mediatingtechnologies.

Table 1 summarizes the main differences forthe three value configurations that the remainderof the paper develops in more detail. Distinctivevalue creation technologies are the critical refer-ence point. The next section presents the maincharacteristics of the long-linked value creationlogic and the corresponding value chain con-figuration. The section also develops the key con-

Table 1. Overview of alternative value configurations

Chain Shop Network

Value creation logic Transformation of inputs (Re)solving customer Linking customersinto products problems

Primary technology Long-linked Intensive Mediating

Primary activity I Inbound logistics I Problem-finding and I Network promotion andcategories I Operations acquisition contract management

I Outbound logistics I Problem-solving I Service provisioningI Marketing I Choice I Infrastructure operationI Service I Execution

I Control/evaluation

Main interactivity Sequential Cyclical, spiralling Simultaneous, parallelrelationship logic

Primary activity I Pooled I Pooled I Pooledinterdependence I Sequential I Sequential I Reciprocal

I Reciprocal

Key cost drivers I Scale I ScaleI Capacity utilization I Capacity utilization

Key value drivers I Reputation I ScaleI Capacity utilization

Business value systemI Interlinked chains I Referred shops I Layered and interconnectedstructure networks


cepts and assumptions of value configurationanalysis. The next two sections develop the mainelements of the value shop and the value networkconfigurations. An initial review of the distinctivelogic of the alternative value creation technologiesis used to motivate our proposed value configur-ation representations and situate the discussion ofcost and value drivers. The review of the technol-ogies is primarily conceptual. Although it is basedon both our interpretation of relevant literatureand on consideration of illustrative examples, theassertions made should be viewed as propositionsand hypotheses in need of further research.

The three generic value creation technologieswith their associated distinctive value configur-ation models provide the foundation for a theoryand a framework for the analysis of firm-levelcompetitive advantage. The discussion sectiondevelops some of the implications of the proposedframework for strategic analysis and strategy withemphasis on the distinctive importance of drivers.The section also links the proposed configurationsto organizational design as strategy implemen-


tation and to competitive logic at the businessvalue system level. The paper concludes withsome implications for further research.

THE VALUE CHAIN

Porter’s work (1985) is the key reference onvalue chains and value configuration analysis forcompetitive advantage. Porter, however, does notuse the term value ‘configuration’ analysis asthe value chain is the sole value configurationconsidered. In our review of the value chainmodel we make explicit a number of argumentsthat are implicit in Porter’s value chain analysisframework.

Value creation logic

We propose that the value chain models along-linked technology (Thompson, 1967), where valueis created by transforming inputs into products.The product is the medium for transferring valuebetween the firm and its customers. Raw materialsand intermediate products are typically trans-ported to the production facility that transformsthe inputs into products which are shipped to cus-tomers.

Marketing serves two complementary purposes.The first is in the development and refinement ofthe chain by providing product specifications andvolume estimates. The second is to simulate therequired level of demand for the chain’s outputto ensure stable operation and capacity utilization.Post-purchase service is performed to ensureproper use of the product by the customer, toremedy defects or to increase the lifespan ofthe product.

Consider assembly line-based manufacturing asan example of a long-linked value creation tech-nology. The assembly line is designed to producestandard products at low cost per unit byexploiting cost economies of scale. The activitiesare buffered from short-term input or output fluc-tuations in adjacent activities by intermediate stor-age.

Interdependencies between activities are dealtwith through coordination. Thompson distin-guishes between pooled, sequential, and reciprocalinterdependence (1967: 54–55).All value creation


technologies have some degree of pooledinterdependence—to the extent that organizationalactivities share common resources. Some technol-ogies have pooled and sequential or pooled andreciprocal interdependence, and the most complextechnologies have pooled, sequential, and recipro-cal interdependence. In firms with a long-linkedvalue creation technology, the interdependenciesof the primary activities are also sequential where,for example, the outputs of inbound logistics arethe inputs to operations.

The value of products is a function of BuyerPurchasing Criteria (Porter, 1985: 141–143).Variation in Buyer Purchasing Criteria gives riseto selective adaptation of products or differen-tiation. Differentiated products can command ahigher price if they provide a better match withBuyer Purchasing Criteria. Customer value isdefined either by the cost reductions that theproduct can provide in the customer’s activitiesor by the performance improvements that thecustomer can gain by using the product. Porter’sgeneric strategies of cost or differentiation (1980)are aimed at improving either the cost or value ofa product relative to the average of the industry.

Technology development is performed to eitherreduce the cost of a product, particularly throughprocess improvements, or to raise the com-mandable price by improving the adaptation ofthe product to Buyer Purchasing Criteria.

Representation of value creation

The value chain analysis framework postulatesthat competitive advantage is understood by dis-aggregating the value creation process of the firminto discrete activities that contribute to the firm’srelative cost position and create a basis for differ-entiation. The basic assumption underlying thedisaggregation is that activities are the buildingblocks by which a firm creates a product that isvaluable to its customers. Different activities havedifferent economics and contribute differently tothe valuable characteristics of the product.

The activity disagregation must be complete inthe sense that it capturesall activities performedby the firm. To maintain a strategic and manage-able perspective on value creation, it is importantthat the activity disaggregation not be toodetailed, while still enabling one to identify thoseactivities that are strategically important. The heu-ristic proposed by Porter for disaggregating activi-


ties3 is that the resulting activities (1) have differ-ent economics, (2) have a high potential impacton differentiation (value), or (3) represent a sig-nificant or growing proportion of cost.

The value chain configuration is a two-levelgeneric taxonomy of value creation activities(Porter, 1985). Primary activities are directlyinvolved in creating and bringing value to thecustomer, whereas support activities enable andimprove the performance of the primary activities(for a similar two-level activity categorization seealso Kornai, 1971; de Chalvron and Curien, 1978;Stabell, 1982). The ‘support’ label underlines thatsupport activities only affect the value deliveredto customers to the extent that they affect theperformance of primary activities. Primary valuechain activities deal with physical products(Porter, 1985: 38).

Primary activities

The five generic primary activity categories ofthe value chain are (Porter, 1985: 39–40):

I Inbound logistics. Activities associated withreceiving, storing, and disseminating inputs tothe product.

I Operations. Activities associated with trans-forming inputs into the final product form.

I Outbound logistics. Activities associated withcollecting, storing, and physically distributingthe product to buyers.

I Marketing and sales. Activities associated withproviding a means by which buyers can pur-chase the product and inducing them to do so.

I Service. Activities associated with providingservice to enhance or maintain the value ofthe product.

The primary activity categories—particulary theinbound logistics–operation–outbound logisticssequence—are well suited to characterizing themain value creation process of a generic manufac-turing company. Casual empiricism suggests thatmanufacturing or process industry firms fre-quently use the value chain activity categoryvocabulary when defining and describing theiroperations. Marketing is included as a primary

3 Porter’s discussion of the appropriate level of disaggregationapplies to these individual activities. He is not thinking ofthe choice of generic activity categories.


activity category as these activities inform thecustomer of the relevant product characteristicsand ensure product availability on the market.Similarly, the inclusion of service as a primaryactivity category follows from the fact that servicecan be critical for the value realized by the cus-tomer.

The set of generic activity categories is a tem-plate for identifying critical value activities thatprovide a basis for understanding and developingcompetitive advantage from the perspective of thefirm as a whole.

The value chain configuration is not meant tomodel the actual flow of production. The valuechain activity focus can be used for identificationof strategic improvement needs or opportunities,but is not necessarily useful for specifying areengineering of business processes.

Generic activity categories are not the same asorganizational functions. Related activities froma competitive advantage perspective can span sev-eral organizational functions. A single functioncan similarly perform activities that need to bedistinct from a competitive advantage perspective.This is perhaps most apparent in the distinctionbetween primary and support activities.

A firm’s value chain is embedded in a systemof interlinked value chains (Porter, 1985: 34).This value system includes the value chain ofsuppliers of raw materials and components. Italso might include the value chain of distinctdistribution channels before the product becomespart of the buyer’s value chain. The overall sys-tem is thus a chain of sequentially interlinkedprimary activity chains that gradually transformraw materials into the finished product valued bythe buyer.

Support activities

The generic support activity categories of thevalue chain are:

I Procurement. Activities performed in the pur-chasing of inputs used in the value chain.

I Technology development. Activities that canbroadly be grouped into efforts to improveproduct and process.

I Human resource management. Activities ofrecruiting, hiring, training, developing, andcompensating personnel.

I Firm infrastructure. Activities of general man-


agement, planning, finance, accounting, legal,government affairs, and quality management.

The categories of support activities are notuniquely linked to the value creation logic of along-linked technology. The same categories ofsupport activities should therefore be relevant toother primary value creation logics. Porter doesnot argue explicitly for his categories of supportactivities, and the taxonomy appears to followpragmatically the traditional functional organi-zation of the firm, where support categories coverthose functions not included in the primaryactivity categories of the value chain configur-ation.

Value configuration diagram

Figure 1 shows the generic value chain diagram.The sequencing and arrow format of the diagramunderlines the sequential nature of the primaryvalue activities. The support activities in theupper half potentially apply to each and all ofthe categories of primary activities. The layerednature of the support activities are apparentlymeant to tell us that activities are performed inparallel with the primary activities. The marginat the end of the value chain arrow underlinesthat the chain activities are all cost elements thattogether produce the value delivered at the endof the chain.

For the analysis and diagnosis of a particularfirm’s competitive advantage, it is necessary toidentify the firm’s individual value activities usingthe generic value activity categories. Figure 2shows an example of the instantiated value chaindiagram for a copier manufacturer with primaryvalue activities (Porter, 1985).

Figure 1. The value chain diagram. Reprinted with the permission of The Free Press, a division of Simon &Schuster fromCompetitive Advantage; Creating and Sustaining Superior Performanceby Michael E. Porter.

Copyright 1985 by Michael E. Porter.


Diagnosis of competitive advantage

Allocating individual activities to generic catego-ries is an analytical choice with strategic impli-cations. The same applies to the choice of activi-ties that are considered for explicit enumeration.

Value chain analysis is often limited to andsummarized by the identification and discussionof strengths and weaknesses in terms of criticalvalue activities (Hax and Majluf, 1992). A moredetailed first-order analysis assigns costs andassets to the value activities.

Second-order analysis requires a closer look atthe structural drivers of activity cost and valuebehavior. The drivers are related to the scale andscope of the firm, linkages across activities, andenvironmental factors. Cost and value drivers areoften analyzed separately.

First-order analysis

The allocation of costs and assets to each activitycan be used to assess the activities that are themost important determinants of overall productcost. Comparing differences relative to competi-tors or other relevant benchmarks provides anindicator of competitive advantage and improve-ment potential.

Obtaining reliable and accurate cost and valuedata for value chain analysis is difficult (Hergertand Morris, 1989). Traditional accounting dataare most often not collected and reported in afashion consistent with the needs of value chainanalysis. As noted above, effective analysis fordiagnosis of competitive advantage requires notonly obtaining historical data, but also projectingtrends and comparing results with similar datafrom competitors.


Figure 2. Value chain diagram for a copier manufacturer. Reprinted with the permission of The Free Press, aDivision of Simon & Schuster fromCompetitive Advantage; Creating and Sustaining Superior Performanceby

Michael E. Porter. Copyright 1985 by Michael E. Porter.

Despite the inherent difficulties often encoun-tered, first-order analysis is useful for a numberof reasons. First, value configuration analysis isuseful because it promotes the right questions:what is the firm’s competitive position and howcan it be sustained or improved? Second, theawareness and commitment promoted bytheprocessof diagnosing competitive advantage isoften just as important as obtaining accurate esti-mates of costs and value. Third, the difficulty ofobtaining a good understanding of cost and valuebehavior for critical value activities is an indicatorof causal ambiguity and barriers to imitation (cf.for example, Reed and DeFillipi, 1990). Thisdifficulty underlines the potential competitiveadvantage that might be obtained from effectivevalue configuration analysis.

Drivers of cost and value

The cost behavior of value activities is determinedby structural factors that are defined as cost driv-ers. Identification of structural factors provides aheuristic for assessing the cost behavior and costeconomics of the value activities for a firm. Therelative importance and absolute magnitude ofcost drivers will vary from industry to industryand from firm to firm. Exploiting and shapingthese structural factors is a main source of com-petitive advantage.

Drivers are partly related to internal relation-ships, partly related to external factors, and partly


related to the relationship between internal andexternal factors.

Porter (1985) identifies 10 generic drivers:scale, capacity utilization, linkages, inter-relationships, vertical integration, location, timing,learning, policy decisions, and government regu-lations. All drivers of cost and value identifiedby Porter are potentially relevant. However, theirrelative importance and role might differ acrossfirms and, as we shall show, systematically acrossthe three alternative value creation logics. Thevalue chain model promotes a heavy focus oncosts and cost drivers (Porter, 1991).4 The maindrivers of value are the policy decisions that aremade by product and segment choices when thefirm is established or is repositioned.

For the generic value chain, the major driverof cost is scale. Associated with scale is thestructural importance of capacity utilization.Internal scope relates to the degree of verticalintegration forwards towards customers and back-wards into suppliers. Thompson (1967) arguesthat vertical integration is the primary means forchains to reduce control costs due to supply anddemand uncertainty.

Traditional economics of scale relate to both

4 Porter (1991) proposes three uses of the value chain model:(1) a template for understanding cost position, (2) a templatefor understanding product effects on cost position of buyer,and (3) a tool for analyzing the added costs that differentiatingmight imply.


economies of labor–capital substitution and learn-ing. The other main drivers relate to the eco-nomics of both internal and external scope.5

Scope and scale have diseconomies that followfrom the need for coordination due to nonperfectdecomposition (Simon, 1982) of the activities ofthe firm.

The primary activities of the long-linked tech-nology have both pooled and sequential inter-dependence. There are, therefore, potentially sig-nificant cost and value drivers in the form oflinkages across primary activities and with theprimary activities of suppliers and customers.

Strategic positioning options

The purpose of value configuration analysis isdiagnosis and improvement of competitive advan-tage. Competitive advantage is relative to existingand potential competitors. Competitors are definedby product and market segment scope. A thirddimension is scope in terms of value activities inthe business value system of interlinked firms.This is often referred to as degree of verticalintegration. Strategic positioning for competitiveadvantage is therefore an issue of choosing posi-tion in terms of product scope, market scope, andbusiness value system scope. We suggest that thestructure of the business system is a function ofthe underlying value configurations of the firm.Or stated differently, there are unique value sys-tem scope options relative to the different con-figurations.

The appropriate choice of position depends onthe drivers of cost and value. For firms with along-linked technology, relationships betweenscale, capacity utilization, market scope, anduncertainty in input and output markets are thecritical generic determinants of the appropriatestrategic position. The drivers shape the businessvalue system, the industry, and thereby also thecompetitive position. Competitive position willalso be a function of where the industry is in theproduct life cycle.

A position of competitive advantage cannot bechosen directly, but must rather be attained byappropriate actions in terms of scope and interms of attempts to modify the drivers of costand value.

5 Scope can be further divided into horizontal and verticalscope.


Sustainable competitive advantage is deter-mined by the nature of the sources of competitiveadvantage. These are in part captured by unique-ness and nonimitability of the drivers of cost andvalue that underly a position.

The logic of the value chain implies an analysisof competitive positioning based on variants ofcost leadership. That is, the value chain frame-work has most to say about how to achieve acost leadership position. The overall flow logicof the primary activities direct attention only tothose Buyer Purchasing Criteria associated withimproving the flow of the larger value systemthat includes buyers and suppliers.

THE VALUE SHOP

Value shops—a short form for ‘firms that can bemodeled as value shops’—rely on anintensivetechnology (Thompson, 1967) to solve a customeror client problem. Selection, combination, andorder of application of resources and activitiesvary according to the requirements of the problemat hand. Thus while the chain performs a fixedset of activities that enables it to produce astandard product in large numbers, the shopschedules activities and applies resources in afashion that is dimensioned and appropriate tothe needs of the client’s problem. The problemto be solved determines the ‘intensity’ of theshop’s activities.

Examples of firms that rely on an intensivetechnology are professional services, as foundin medicine, law, architecture, and engineering.Important functions or parts of firms can alsohave a value creation logic that is best understoodas a value shop, even though the primary activi-ties of the overall firm have a value creationlogic that is consistent with the product andtransformational logic of the value chain. Forexample, petroleum exploration and petroleumfield development6 in the upstream petroleumindustry (Jones, 1988) and more generally indus-trial product and process development (Clark andWheelwright, 1993) can be understood as basedon an intensive, problem-solving technology.These functions or units are most often rep-resented as support activities in a value chain

6 As opposed to petroleum field operation.


configuration. The value shop configuration cantherefore be used to model the value creationlogic of critical support activities.

We have called the value configuration of anintensive technology a ‘value shop.’ The ‘shop’label captures that a firm so configured is directedat a unique and delineated class of problems—in a fashion similar to the way the shop of amechanic repairs cars. The shop metaphor signalsthat assembly and matching of both problems andproblem-solving resources are important for theorganization and management of the value shop.

The shop metaphor also signals that organi-zations with intensive technologies often bothimprove performance and reduce costs byincorporating the object worked on, be it byhospitalizing patients, by performing education inthe classroom, or by providing consulting serviceson customer premises (Thompson, 1967: 43). Inupstream oil, the object incorporated is a modelof the basin, play, prospect, or field—most oftenin the form of maps, seismic sections, strati-graphic columns—but increasingly in a computer-supported medium and using a computer-supported representation.


Problems can be defined as differences betweenan existing state and an aspired or desired state(Simon, 1977). Problem-solving, and thus valuecreation in value shops, is the change from anexisting to a more desired state. In the case ofmedical services, the change is to cure the patientof a sickiness. In the case of the architect, thechange can be to raise a building or other struc-ture at a particular site. Problems involve situ-ations requiring remedial action and situationswhere there are improvement opportunities.

The intensive technology is thus directed atbringing about desired changes in some specificobject of interest to the client or customer. Inmany cases, the object is human, such as inhealth care and education. But the same valuecreating logic is found in firms where the objectis an artifact to be created or modified, such as asite, a system, or a knowledge state. In petroleumexploration, the object is a basin, play, and pros-pect with more or less uncertain petroleumresources that, when explored, might betransformed into fields with proven commercialreserves.


Consider the case of the patient who visits ageneral practitioner because he has a chest pain.The physician starts the consultation (Stoeckle,1987) by asking about the chief complaint—thesymptoms that brought the patient in for medicalcare, and involves also asking questions aboutthe patient’s relevant history. The physician thentypically performs a physical examination. Theexamination may uncover indications for a sus-pected hypothesis or may trigger a reformulationof the hypotheses. Diagnostic tests are used toconfirm or rule out suspected diagnoses. Some-times trial therapy also serves as a diagnostictest. No therapy, i.e. wait and see, might alsoserve to pinpoint relevant diagnoses. In somecases the physician concludes that the patientneeds to be referred to a specialist in cardiology.In the final stage of the consultation, the physicianmakes a treatment plan for the patient, specifyingthe treatments to administer and the procedurefor monitoring the patient’s progress. Monitoringprogress towards resolving the client’s chest painproblem might involve a house call or a patientvisit to the office of the physician.

The simple example of the medical consultationillustrates a number of key distinctive character-istics of value creation with an intensive tech-nology.

Value information asymmetry.A strong infor-mation asymmetry between the firm and its clientis perhaps the single most important attribute ofan intensive technology. The asymmetry is thereason that the patient approaches the generalpractitioner. The physician knows something thatthe patient thinks he needs. Equally importantfrom a value creation perspective, the firmdelivers value even by determining that the clienthas no problem. The medical doctor often mightdeliver value by merely attending to the client.All this is due to the fact that the client–patientis not able to determine if the service is corrector appropriate, even in cases where the outcomeis negative (Friedson, 1960; Karpik, 1989).

Configured to deal with unique cases.Clientproblems often involve more or less standardizedsolutions, but the value creation process isorganized to deal with unique cases. In manysituations, less specialized personnel could handlemost of the problems. However, the professional(e.g., the medical doctor) always needs to be


involved to be able to recognize and deal withthe limited number of cases that require theirexpertise, have been incorrectly diagnosed, orwhere the treatment is not performing as expected(Abbott, 1988). The patient expects the serviceof the professional and is motivated to followand trust the testing and treatment by referenceto the relevant expertise.

Cyclical, iterative and interruptable activi-ties. The flow of activities is not linear, butiterative between activities and cyclical acrossthe activity set. Diagnosis moves back and forthbetween hypotheses and new data collection thatconfirm, reject, or lead to a reformulation of thediagnosis. Treatment might initiate a newproblem-solving process to determine the mostappropriate way of administering the treatment(Simon’s, 1977, wheels-within-wheels meta-phor). A treatment can result in the resolution ofthe client’s problem, but can also initiate a newand perhaps a different sequence of activities. Theprocess is not only iterative, but also potentiallyinterruptable at all stages, either when the symp-toms are found to be a false alarm, when thereis no known solution, or when the problem needsto be referred to a specialist.

Significant sequential and reciprocal inter-dependence between activities.The iterative andcyclical nature of problem-solving in shops resultsin a high degree of both sequential and reciprocalinterdependence between activities. For example,appropriate definition of the problem to be solvedis vital for all other activities; feedback both fromtrying to generate a solution and fromimplementing a chosen solution might requireredefinition of the problem or search for alterna-tive solutions. The consequent high demands forcoordination across activities are often dealt withby assigning the problem to a single professionalwho follows the problem to resolution and byusing lateral integration mechanisms (Galbraith,1973) that facilitate information exchange whilemaintaining high professional commitment andresponsibility. In demanding cases that requirethe interplay of multiple disciplines and expertisein the development of innovative solutions, thecoordination needs are often addressed byassigning a full-time cross-functional team (Clarkand Wheelwright, 1993). In all cases the heavycoordination needs are addressed by reducing the


coordination across the problems and clients tosimple pooled interdependence.

Multiple disciplines and specialities in spirallingactivity cycles. Offshore petroleum field devel-opment illustrates an intensive technology thatrequires the interplay of several different special-ities in the development of artifacts. Field devel-opment moves from discovery to realization ofthe field operation through the following stages:application for concession, exploration, feasibilitystudies, field development and planning, basicengineering as part of planning for field develop-ment, detailed engineering, and with fabrication,construction, hook-up and commissioning as real-ization of field development (Hallwood, 1990).This might appear to be a sequentially interlinkedset of activities. It is rather a refinement of abasic problem-solving cycle where each cycleimplements the solutions chosen by the previouscycle or each cycle is passed the new problemthat has resulted from the resolution of the initialproblem. The process also changes in terms ofthe object of interest: in upstream petroleum,from the basin and play to the prospect; once theprospect is identified as a potential field, itbecomes delineated into several components that,depending on their extent and form, need to bedeveloped with one or more platforms (wells).7

Problem-independent information acquisitionactivities. The professional often has a standardinformation acquisition procedure to make surethat the problem has been correctly framed. Anexample is the doctor who always takes thepatient’s temperature, inspects his throat, andknocks on his knee, irrespective of what thepatient presented as the symptoms or the natureof his illness. This standardization of informationacquisition activities provides both value and lim-its overall costs, in part as it provides the basisfor early anticipation of succeeding activities.

Leveraging expertise.Firms with an intensivetechnology are labor intensive with professionalsand specialists in the problem domain covered asthe core and frequently the largest component ofthe workforce. Scale of operation beyond the

7 Note that the object changes name as it moves through theshop (analogous to the change of problem space in the Newelland Simon (1972) representation of human problem-solving).


collection of independently performing pro-fessionals is achieved by leveraging experiencedsenior professionals with more junior and lessexperienced colleagues (Maister, 1993). Seniorpersonnel mentor and back up their less experi-enced colleagues (Dalton, Thompson, and Price,1977), while clients are assured that at all timesthey get the appropriate professional expertiseand problem-solving effort. The role of seniorpersonnel has to balance quality assurance anddirection with the need to make sure that theperforming professional takes responsibility witha motivation to perform a best effort using bestpractice.

Coperformance of support and primary activi-ties. Human resource management ofprofessionals—recruiting, developing, and retain-ing good professionals—is critical. However, thishuman resource activity is often performed aspart of doing professional work, in part becausethe managing professional is a performing pro-fessional (Lorsch and Mathias, 1987). In part, adistinct human resource activity is limited becausethe recruiting and retention capability of firmdepends primarily on the reputation and qualityof the problem domain professionals.

Similarly, marketing, procurement, and tech-nology development are seldom distinct activitiesin all but the largest firms with an intensivetechnology. These activities are dependent on andtherefore often carried out by the professionalsin the course of solving client problems.

Consider marketing. Defining the client’s prob-lem is also client acquisition. Marketing is largelyrelationship management (Eccles and Crane,1988; Cox et al., 1987) that involves referralsfrom customers and colleagues (Karpik, 1989).The professionals—or rather their reputation—isoften the critical marketing resource.

An important procurement activity is acquiringor accessing the technology of the professionor specialization. The performing professionalsaccomplish this as part of their efforts to keepup-to-date with the state-of-the-science and thestate-of-the-art of their profession.

Learning and innovative problem-solving is themodus operandi in firms with an intensive tech-nology. Choice of challenging customer problemsis a main means for technology development.

Referrals based on reputation and relation-


ship. Relations between firms with an intensivetechnology within the corresponding businessvalue system is either one of referral or of sub-contracting. In the case of referral, as when thegeneralist passes the client to a specialist, theresponsibility for the problem and client is oftenirrevocably transferred. In the case of subcon-tracting, as when an oil company subcontracts adrilling assignment to a service company, theprincipal firm retains problem ownership and con-trol. The resulting business value system is anetwork of relations and reputations (Friedson,1960; Karpik, 1989).


Firms that can be modeled as value shops aretypically populated by specialists and experts,often professionals, in the problem domaincovered. A profession by definition has a knowl-edge base, methodology, and language that areunique and that require long training to master(Abbott, 1988). Accordingly, the primary activi-ties of the value shop are often couched in termsand sequenced in a form that is unique to eachspeciality and profession. Therefore, a commonterminology for primary value shop activitiesabstracts the generic categories of problem-solving and decision-making activities (Pounds,1969; Simon, 1982; Stabell, 1983).

Primary activities

There are five generic categories of primary valueshop activities. Each category is divisible into anumber of distinct activities that depend on theparticular industry and firm strategy:

I Problem-finding and acquisition. Activitiesassociated with the recording, reviewing, andformulating of the problem to be solved andchoosing the overall approach to solving theproblem.

I Problem-solving. Activities associated with gen-erating and evaluating alternative solutions.

I Choice. Activities associated with choosingamong alternative problem solutions.

I Execution. Activities associated with communi-cating, organizing, and implementing thechosen solution.

I Control and evaluation. Activities associatedwith measuring and evaluating to what extent


implementation has solved the initial problemstatement.

Problem-finding and acquisition have much incommon with marketing in the value chain. Theclient owns the problem and in certain cases,such as health services and education, ‘embodies’the problem.8

Choice is an activity category that in mostcontexts is of limited importance in terms ofeffort and time, but is important from the pointof view of value. It also represents the interfacebetween different specialities and a major dis-continuity in the problem-solving cycle.

Interfirm relations across decision cycles in thebusiness value system are either by referral afterproblem-finding, referral after choice activities, orsubcontracting of execution activities. Theresulting wheels-within-wheels or spirallingactivity configurations define the vertical scopeof the business value system.

Support activities

As many support activities, such as humanresource management, are coperformed with theprimary activites, one might conclude that theyshould be removed from the value shop diagram.These functions may not be well taken care of—precisely because they are not distinct, but theyare crucial to competitive advantage.


Figure 3 is the generic value shop diagram. Thecyclic nature of the activity set is captured bythe circular layout of the primary activity categor-ies, where postexecution evaluation can be theproblem-finding activity of a new problem-solving cycle. The wheels-within-wheels natureof the activity set can be shown by expanding theexecution activity into problem-solving–choice–execution–evaluation activities. The spiralling na-ture of the activity set is obtained when a decisioncycle refers (and passes control to) a different ormore specialized shop that picks up a reformu-lated or reframed client problem.Figure 4 illustrates the instantiation of the pri-

8 The value configuration models activities, not who performsthem. A client may therefore be actively involved in theproblem-solving, e.g., a patient taking his own temperature.


Figure 3. The value shop diagram

mary activity categories of the value shop dia-gram to the general practioner shop. The medicalconsultation shop appears to be a diagnosis-focused shop. Treatment plans follow more orless directly from the diagnosis.9

Figure 5 presents the instantiation of a valueshop diagram for upstream petroleum explorationand field development that illustrates a spirallingactivity set with referral. Petroleum explorationis a search-focused shop, where the search forpetroleum is concluded once drilling proves theexistence of petroleum in commercial quantities.

Simplified for our purposes, problem-finding inpetroleum exploration is identifying an area withpotential hydrocarbon prospects,10 problem-solving is generating and evaluating prospects inthe area, choice is what, if any, prospects to drill,while execution is drilling the prospects, andevaluation is the review of the results of the drill-ing.

Petroleum field development is a design-focused shop. Problem-finding is initiated byreferral from exploration and we have chosento define appraisal as an element of the fielddevelopment activity set. Problem-solving in fielddevelopment is the generation and evaluation ofalternative development concepts, then choosingthe field development concept to be used, if any;i.e., is there a commercially viable developmentconcept for the discovery, execution is the actualdevelopment of the field to the point that thefacility is ready for production, and postexecutionevaluation tests that the field is ready for pro-duction.

9 i.e., in most cases there is limited problem-solving activity.10 A prospect is a potential hydrocarbon accumulation.


Figure 4. Value shop diagram for a general practitioner

Figure 5. Value shop diagram for a petroleum explorer (A) and field developer (B)

The petroleum exploration and field develop-ment example raises the issue of choosing anappropriate level of aggregation of activity categ-ories. We distinguish two versions of each majorvalue shop activity category. The reason is thatvalue and cost implications of activities appearto differ in exploration and field development,that exploration and field development rely onquite different disciplines and competences, andthat they seem to be shops with a differentproblem-solving logic.11

11 Search/classification vs. design shops.


Diagnosis of competitive advantage

In value shops, the evaluation of firm-level rela-tive value advantage is more difficult than theevaluation of cost. Relative cost of an activity andits relative value contribution are not necessarilyrelated (Porter, 1985: 121). Shop activitiesaccounting for a small percentage of total costcan have a major impact on value. For example,structural factors that affect early activities typi-cally have a significant impact on both the valueand cost of later activities due to spiralling com-mitments as major phases both implement and areconstrained by the choices made in earlier phases.


The challenge is to establish meaningful indi-cators of value in a situation where we areassessing the capability of the firm to addressfuture client or customer problems—problemsthat are potentially unique and may requirenovel solutions.

Consider the example of the firm that providesmedical treatment. First-order and second-orderactivity analysis would estimate both the relativevalue and cost component of each activity overthe set of patients per unit of time. Professionaltime is a key determinant of cost. In large prac-tices, or hospitals, the relative use of junior andsenior personnel in activity performance is animportant cost component, as is efficient use ofdiagnostic and treatment facilities. Value to clientis estimated by the success ratio of treatments.Value is also associated with the convenience tothe client, e.g. the number of tests used to arriveat a correct diagnosis and the length of treatment.This value is in part a function of the numberof cyclings through problem-finding (number oftimes new diagnosis produced). Activity analysiswould also need to develop cost and value esti-mates for different types of consultations (initialdiagnosis, follow-up, yearly check-ups) and fordifferent disease or client categories. As a generalrule, the value of activity is assessed by its impacton the definition of the succeeding activity in thedecision cycle.

Cost and value drivers

Value drivers as opposed to cost drivers are ofcritical importance in value shops. Competitiveadvantage follows from the fact that clients areprimarily looking for relatively certain solutionsto their problems, and not for services that havelow prices as their main attribute.

Successas it materializes in reputation andrelationships is the canonical value driver in firmswith an intensive technology. Success improvesaccess to both the best personnel and access tothe best clients, problems, or projects (Perrow,1961; Løwendahl, 1992, 1993). For example,architect clients wish to avoid risk and thereforeseek an architect with an established reputationin work of a similar nature (Winch and Schneider,1993). In upstream petroleum, outstanding repu-tation in exploration enables an exploration shopto recruit the best explorationists. It also improvesthe shop’s ability to bid successfully for the mostpromising acreage.


Reputation signals value (Porter, 1985: 139).Relevant examples of signals of value and qualityare demonstrated success such as winning anarchitectural competition or obtaining a Nobelprize, high-quality employees, publications inprestigious journals, and strong demand in theform of long queues and difficult access. Thevalue-signalling issue is very much akin to theissues raised in the economics of informationliterature, where market signalling is a means fora potential employee to reveal information ontheir performance potential (Spence, 1973). Whilethe information asymmetry in second-hand prod-uct markets leads to the fact that most oftenpoorest-quality cars (‘lemons’) are offered forsale (Akerlof, 1970), the same information asym-metry in professional services appears to lead toa premium price and high demand for highest-quality services.12

Demanding projects and clients provide a basisfor effective learning. Demanding projects thathave been successfully performed provide thebasis for building relationships and reputation.Success affects and is affected by the shop’sability to recruit, retain, and develop high-qualitypersonnel. High-quality personnel transcends theeffect of drivers such as linkages across activities,learning, and spillovers.

Consider linkages. As noted earlier, the sig-nificant activity interdependencies within a clientproject or problem lead to an organization ofwork where single professionals or teams of pro-fessonals are assigned responsibility for all activi-ties related to each client problem. Overall per-formance and thus value depend primarily on thequality of the individual professionals assigned toclient problems and projects.

Learning is an integral and explicit part of theproblem-solving cycle of the shop. Evaluationand postimplementation control is a means toimprove the shop’s ability to deal more effec-tively with the problem at hand; both throughbetter problem definition (problem redefinition),better alternatives, and better implementation.

Learning across projects and client problemsis a critical shop-level(as opposed to individualprofessional) linkage in the value shop. This

12 This difference in dynamics is related to the fact that whilethe asymmetry in the second-hand product market is relatedto time- and location-bound information, the value shop asym-metry is related to general and more universal information(Hayek, 1945).


interproblem learning is particularly important forinterrupted problem-solving cycles that appear tobe inconclusive, but provide rapid feedback ifmonitored and classified in a systematic fashion.

In general, the large number of very smallvalue shops13 suggests that there are limitedadvantages of scale and significant advantages oflocation in the value shop. This is in part becauseof the relative value of outstanding professionals,the costs of coordination of large groups ofspecialists and the need for effective communi-cation in problem-finding and problem-solving.Location advantages are related to access to cli-ents and access to knowledge in terms of person-nel or professional communities such as univer-sities or other firms (Porter, 1990).

Possible scale advantages are related to thescale of the client’s problem and the distributionof the client across multiple locations. Forexample, we see that scale and location providean important advantage for shops, e.g. large con-sulting firms, serving global clients.


Business value system scope and product scopeare heavily interrelated in the value shop. Bothproduct scope and business value system scopeare related to degree of specialization in problemsor solution technologies. High vertical integrationin business value system implies broad coverageof specializations and existence of generalists thatcan refer to appropriate specialists.

Choice of business value system scope willdepend in part on market size and in part on therate of change of the intensive technology. Thelarger the market for a speciality and the greaterthe rate of change in the intensive technology,the less vertically integrated the firm.

An additional unique strategic positioningoption in firms that can be modeled as valueshops is the degree of incorporation of the prob-lem object. Problem incorporation is primarily atool for reducing uncertainty,14 but is also ameans to increase communication betweenspecialists and a means for efficient and effective

13 Consider consulting and professional service firms, inde-pendent professionals.14 According to Thompson (1967) problem incorporation is apositioning alternative equivalent to scale in the network (tobalance potential random demand) and vertical integration (inorder to control supply and demand uncertainty) in the chain.


postimplementation (treatment) evaluation (inhospital or educational establishment). Problemincorporation is thus a tool for both cost reductionand value creation. Degree of problem incorpo-ration is related to the degree of business valuesystem scope. This is because benefiting fromstrong problem incorporation requires that theshop has access to the full range of relevantspecialists.

THE VALUE NETWORK

Value networks—a short form for ‘firms thatcan be modeled as value networks’—rely on amediating technology(Thompson, 1967) to linkclients or customers who are or wish to be inter-dependent. The mediating technology facilitatesexchange relationships among customers distrib-uted in space and time. The firm itself is not thenetwork. It provides a networking service.

Examples of firms that rely on a mediatingtechnology are telephone companies, retail banks,insurance companies, and postal services. Theterm value ‘network’ underlines that a criticaldeterminant of value to any particular customeris the set, or network, of customers that areconnected. Stated in communication terms, thevalue of a communication service depends onwhom it enables the customer to communicatewith.


Modern society is characterized by a complex setof actual and potential relationships betweenactors, people, and organizations. Linking, andthus value creation, in value networks is theorganization and facilitation of exchange betweencustomers. The linking can be direct as in atelephone service, linking two or more parties ina call, or indirect as in retail banking where onecustomer is not linked directly to another cus-tomer, but a group of customers is linked througha common pool of funds.

Mediators act as club managers.One can thinkof managing a mediating firm as managing aclub. The mediating firm admits members thatcomplement each other, and in some casesexclude those that don’t. The firm establishes,monitors, and terminates direct or indirect


relationships among members. Supplier–customerrelationships may exist between the members ofthe club, but to the mediating firm they are allcustomers. Depositors are not bank suppliers, theyare just as much bank customers as those bor-rowing money. By acting as an intermediary,bilateral interactions between the mediator and itscustomers are used to enable multilateral inter-actions between customers.

A set of customer contractscommit both thecustomer and the company operating the networkto a mutual set of obligations. Contracts arerequired to be able to service efficiently ondemand mediation requests, randomly distributedin time and space. Contracts specify price andmutual obligations of service provider and cus-tomer.

Service value is a function of positive networkdemand side externalities.Adding one morecustomer to a network directly affects the valueof the service to other customers (Katz andShapiro, 1985).

Positive network externalities introduce uniquestrategic challenges. A new service has relativelylow value to its first customers, whereas the coststypically are the highest in the introduction phase.This leads to distinct life cycle phases.

Value is derived from service, service capacity,and service opportunity. The customer mayreceive value from the value network withoutever actually invoking the mediation services. Forexample, a bank customer may pay for a creditaccount in order to secure access to funds, ifnecessary.

Mediators typically charge customers separatelyfor the linking opportunity and the actual use oflinking services in terms of activities performedand capacity utilized. A subscription fee impliesa commitment to servicing potential customerrequests for the mediation services. Some bankshave fixed monthly charges associated withaccounts in addition to per transaction charges.Interest spread is payment for funding and place-ment capacity utilized.

Mediation activities are performed simultaneouslyat multiple levels. A concurrent and layered setof activities is required to service efficiently arandom need for mediation services between alarge number of customers. Servicing individual


customer transactions clearly involves a sequen-tial set of activities as, for example, initiation,validity checking, and posting for an accountwithdrawal transaction in banking. Such a trans-action, however, is only possible within a networkof contracts with other customers and an infra-structure that hosts the mediation between them.Each requires a distinct set of activities withdifferent cost and value economics.

The simultaneous and layered performance ofactivites implies strong reciprocal as opposed tosequential interdependence between primaryactivities. Failure to synchronize activities maylead to a breakdown of the system. An insurancecompany with insufficient reinsurance infrastruc-ture may bankrupt on major accidents, whereas atelephone system may break down due to excep-tional communication events if it fails toreroute traffic.

Standards are critical for the coordination ofthis reciprocity or, as noted by Thompson, ‘stan-dardization makes possible the operation of themediating technology over time and throughspace by assuring each segment of the organi-zation that the other segments are operating incompatible ways’ (1967: 17).

Standardization facilitates matching and monitor-ing. Standardization enables the mediator tomatch compatible customers and to effectivelymaintain and monitor the interaction betweenthem. The retail bank uses standard customercategories to qualify loan applicants or to setterms for borrowers and depositors. Standardizedaccount numbers are further used to direct pay-ments to accounts appropriately and to monitorthe interaction by way of bookkeeping andaccounting.

Distinct life cycle phases of rollout and oper-ation. Customers may be willing to pay a pre-mium price for a new service. However, as thevalue of the service is dependent on who elseadopts it, it may be difficult to target these cus-tomers on an individual basis. Stated differently,the value of the service is managed by the rolloutprocess for the service. It follows that in manycases it is impossible to charge for service orrequired equipment in this initial phase, leadingto ‘give-away strategies’ that have been observedin areas such as cellular and videotext telecom-munications, browsers for the Internet and free


cash balances for initial users of electronic pay-ment cards. Following a successful rollout,mediators may be in a position to charge formembership, service, and equipment in a poten-tially long-term operations phase in which con-tracts, infrastructure, and service activities areperformed concurrently.

Layered and interconnected industry struc-ture. The business value system relationshipsbetween industry actors is not as suppliers andcustomers in an industry value chain, but assimultaneously coperforming levels of mediationservice. For example, network operators deliverthe infrastructure for service providers in telecom-munication, who in turn serve as the communi-cation infrastructure for payment services.Exchange relationships offered by a mediationservice can also extend beyond its immediatecustomers to customers of other mediation serviceproviders. This gives rise to a structure of inter-connected mediation networks.

In telecommunication the distinction is madebetween access networks typically organizinglocal networks of end users and carrier networksproviding communication between local networks.Communication between local subscribers ishandled by the local switch, while internationalphone calls are routed from the local switchthrough a complex web of multiple lines andswitches to connect with a subscriber at a localswitch in another country. As a result, a largenumber of telephone companies specializing inlocal, regional, and international traffic areinvolved in coproducing the phone call.

In summary, the business value system in amediation industry is potentially a set of copro-ducing, layered and interconnected networks thatenhance the range and reach of the services pro-vided.


The object of mediation distinguishes mediators.There are, however, strong similarities betweenthe activities of various value networks even ifthe nomenclature used to describe them differsfrom industry to industry.

Primary activities

The primary activity description is inspired by


that used in telecommunication because telecom-munication is a rather generic form of mediationand because explicit activity decomposition mod-els are well established both at the micro levelof peer-to-peer communication and at the industrylevel in delineating industry actors.

The primary activities of the value network areas follows:

I Network promotion and contract managementconsists of activities associated with invitingpotential customers to join the network, se-lection of customers that are allowed to joinand the initialization, management, and termi-nation of contracts governing service pro-visioning and charging.

I Service provisioning consists of activitiesassociated with establishing, maintaining, andterminating links between customers and billingfor value received. The links can be synchron-ous as in telephone service, or asynchronous asin electronic mail service or banking. Billingrequires measuring customers’ use of networkcapacity both in volume and time.

I Network infrastructure operationconsists ofactivities associated with maintaining and run-ning a physical and information infrastructure.The activities keep the network in an alertstatus, ready to service customer requests.

Contracts and contracting activities vary acrossnetworks. Greater commitment between mediatorand customer leads to more extensive contractsand contracting process. Qualification for a houseloan is more extensive than for a telephone ser-vice. Network promotion differs from sales andmarketing in the value chain in that selection ofcustomers is as important as attraction.

Service provisioning depends on the nature ofthe mediation. Establishing a network link, be ita bank transaction or a telephone call, requiressome form of feasibility check which includesclarifying the nature of the transaction, avail-ability of linking possibilities and the eligibilityof the customer in making the link.

The specific network infrastructure operationactivities depend on the nature of infrastructureused. In telephone and other public utility com-panies the key infrastructure is switches or distri-bution centers; in banks and insurance companiesit is the branch offices and financial assets; whilein transportation and distribution companies it isthe vehicles and warehouses.


Support activities

Among the support activities of the value net-work, two distinct, but related technology devel-opment activies are of special interest: networkinfrastructure development and service develop-ment. Network infrastructure developmentincludes activities associated with the design,development, and implementation of networkinfrastructure. Service development includeseverything from the modification of a large setof possible customer contract terms, e.g. interestand time schedules in a bank, to the developmentof brand new services, e.g., voice mail servicesin a telephone company. It also includes modifi-cations to the company—customer interfacethrough modifications of procedures, forms, andself-service computer interfaces.

Procurement is heavily linked to network infra-structure and service development, and is oftenspecialized for these activities. Similarly, humanresource management is often quite different forinfrastructure development and service develop-ment, relative to primary activities.

Firm infrastructure, i.e., general management,financing, and management information systems,should not be confused with the value networkinfrastructure. The former facilitates operating thecompany, while the latter is at the heart of valuecreation for customers.


Figure 6 shows the generic value network dia-gram. The three primary activity categories over-lap in order to underline the concurrent inter-activity relationship across primary activity

Figure 6. The value network diagram


categories.15 The lack of direction of value cre-ation where no arrow identifies the final customerunderlines that the work creates value bymediating between customers.

Figure 7 shows an example of the instantiatedvalue network diagram for a retail banking firm.Network promotion and contract managementinclude promotion of sale of services, risk evalu-ation, contracting, and monitoring of contracts.The bank both attracts and selects among cus-tomers. Contracts govern explicitly the relation-ship between the customers and the bank andimplicitly between the customers. The contractsof a retail bank govern the implicit exchangerelationship between the customers. The retailbank is the agent of its customers and as such itassumes the financial risk involvement in theexchange relationship between depositors andlenders.16

Service provisioning includes deposit, with-drawal, funds transfer, maintaining account bal-ances, and interest calculation. These activities aregoverned by the contracts managed by networkpromotion and contract management activities.Breach of the contract agreements, e.g., by over-draft of a savings account or default on loan

Figure 7. Value network diagram for a retail bank

15 Interaction is often modeled as taking place at several levelsor in layers to capture the simultaneous performance ofactivities (Tanenbaum, 1981; Alderferer, 1987).16 Note that in the case of an investment bank the contractsmay be established directly between customers, and the bank’smain involvement is in establishing the contract. This, how-ever, may require financial exposure for the bank, as exem-plified, by the billion-dollar bridge loan extended by FirstBoston in the restructuring of ‘Union Carbide Deal’ Case(1988).


payments, reciprocally impacts the contract activi-ties and may lead to contract modifications or ter-mination.

Infrastructure operation consists in part ofoperating a physical infrastructure for storing andtransmitting funds. This includes branch offices,automated tellers, and IT systems. However, avital part of the infrastructure is also the financialinfrastructure that provides mediative capacity.This consists of maintaining liquid assets andlinks with other sources of liquid assets suchas central banks and the money market throughmanaging the bank’s rating.

Diagnosis competitive advantage

Drivers of cost and value

As mediating firms offer value to their customersboth through the access option and the actual useof services, cost and value must be associatedwith both.

Scale and composition.Scale is a potentialdriver of both cost and value in the value net-work. Value network services are characterizedby demand-side economies of scale resulting frompositive network externalities (Katz and Shapiro,1985). The value of the service to existing cus-tomers increases with each new customer addedto the network. Positive externalities exist for avariety of products. Examples are micropro-cessors, consumer electronics, and software(Wade, 1995). Mediation services offered byvalue networks represent the extreme casebecause the dependency among customers is themain product delivered. Stated differently, invalue networks, the other customers are the keypart of the product. The services of a valuenetwork mainly deliver the customers’ opportuni-ties to exercise those dependencies. Size andcomposition of the customer base are thereforethe critical driver of value in the value network.

Consider an insurance company. If the cus-tomer network is unbalanced, in the sense thata subset of customers systematically receive asdisproportionate part of the claims, then the costof insurance will either be too high for the restof the customers or the insurance company’sprofits will be below industry average.

Insurance companies may try to attract specialcustomer groups to achieve their targeted network


composition, because individual risk assessmentand pricing are sometimes too costly. In fact,some insurance companies make it their explicitstrategy to attract only ‘low-risk’ customers byproviding special terms to good drivers or lifeinsurance to employees of companies in lowhealth hazard industries. It is possible to operatean insurance company for almost any kind ofrisk group, given that they are willing to pay therequired premiums. But the network compositiongiven a particular customer segment is key toboth pricing and profitability within that segment.

Telephone service represents an extreme caseof positive network externalities. Each new cus-tomer added to the network allows for one morepossible connection. For trade exchanges scalecontributes directly to value by increasing marketliquidity (Domowitz, 1995). The externalities areless obvious in deposit banking where althoughscale allows risk sharing or deposit insurance, italso provides variety in risk–interest options.When network externalities are present, the valueof the service provided is affected by the charac-teristics of customers that join the network(Bental and Speigel, 1995).

Scale is also important to the extent that itaffects accessibility. A geographically extendednetwork requires an extended infrastructure. Thiscontributes to an additional size effect on valuebecause the number of access points available tothe customer increases (Domowitz, 1995). Thuswhile the externality effect of scale increasesdirectly the value of the network to the customer,the size effect in the form of increased accessi-bility affects the customer’s cost of using themediation service.

Scale advantages may, however, not be observ-able at the level of individual firms (Forestieri,1993). A single bank, for example, by the natureof banking, extends its network through otherbanks using strategic alliances or correspondentarrangements and the money market. Inter-network alliances or agreements directly affectthe value of the individual customer’s networkmembership.

Common industry standards are a prerequisitefor inter-network connections. The evolution anddiffusion of standards are therefore critical in theexploitation of demand side scale economies.

Capacity utilization. Capacity utilization isclosely related to scale. As in the value chain,


capacity utilization reduces unit costs. High-capacity utilization, however, may also reduceservice levels.

Consider the case of heavy load for a com-munication service: it becomes difficult to get aline. Applied to banking load increases the prob-ability that a bank is not able to service itscontracts; in the extreme case load increases theprobability of a bank run. Hence, in the valuenetwork, capacity utilization is both a cost andvalue driver, while it is primarily a cost driverin the chain.

Linkages. In the value network there is recipro-cal interdependence across primary activity cate-gories due to the need for synchronization anddimensioning of simultaneous activities.Important linkages arise from this reciprocity.Both geographical coverage and capacity mustreflect the composition of customers who aremembers of the network. Adjustments are doneon a continuous basis. A case in point frombanking is the pool of funds. This pool has tobe dimensioned to reflect the liquidity demandpatterns of the customer network. Service pro-visioning capacity must be coordinated with cus-tomer recruitment and diffusion of new services.The pool of funds thus affects the nature of theservice that can be provided and the costs for thebank. Similarly, the switching and line capacity ofa telephone company must reflect the customerbase.

Learning. The key areas for learning in primaryactivities is in membership selection and servicemonitoring. The two are reciprocally related ascan be illustrated from banking. Credit qualifi-cation provides information for monitoring activi-ties, while monitoring can assure both improvedcontract execution and information for theimprovement of credit qualification activites. Inaddition, interfirm learning (spillovers) is criticalin the diffusion of standards as the ability tointerconnect value networks increases size andhence value.


Unique strategic positioning options in terms ofvalue system scope in mediative industries needsto consider both vertical and horizontal inte-


gration. These options mirror the layered andinterconnected nature of the corresponding busi-ness value system.

Vertical scope. A mediation exchange requiresmultiple levels of coproducing mediation activ-ites. The activities of one mediator build on theactivities of another. Vertical scope in mediationindustries describes to what extent a firm controlsall levels of coproducing activities required tocomplete mediation exchanges. Choice of scopedepends on whether suitable lower-levelmediation services covering the relevant cus-tomers are available.

Consider an electronic payment clearance ser-vice. It requires the operation of a communicationinfrastructure by which transactions can be carriedand a transaction processing infrastructure bywhich transactions can be cleared. A service pro-vider may choose to operate both or the providermay choose to base its service on the communi-cation service of, for example, a telecommuni-cation company. The choice is one of verticalvalue system scope.

In the case where the provider of the paymentmediation service chooses to use the communi-cation service of the telephone company, the twocompanies are coproducing the service. The cus-tomers, e.g., banks or end users, may be cus-tomers of both companies and may be payingseparately for the communication and clearingcomponents of the service to the telephone com-pany and the electronic payment service pro-vider respectively.

Horizontal scope. A firm that delivers amediation service can extend its customer seg-ment scope either by increasing its own customerbase or by exchange agreements with othermediating firms that extend the set of exchangesthat the firms provide for their customers. At oneextreme the firm may limit exchanges to thosethat can be completed within its own customer(contract) base and hence be fully horizontallyintegrated relative to this market segment. At theother extreme the firm may specialize on oneside of the exchange, e.g., origination of loansthat are securitized and exchanged in a market(Crane et al., 1996).

Exchange relationships offered by a mediationservice provider and extend beyond its immediatecustomers to customers of other mediation service


providers gives rise to a structure of inter-connected mediation networks.

DISCUSSION

We have proposed three alternative value con-figurations as a foundation for a theory of valueconfiguring for competitive advantage. The theoryis an extension of Michael Porter’s original valuechain framework (1985) and uses Thompson’stypology of long-linked, intensive and mediatingtechnologies (1967). Although the set of threedistinctive value creation technologies is the criti-cal reference point, the focus of our contributionin this paper is on the development of the rep-resentation, logic, and strategic implications ofthe corresponding value configurations.

The unique characteristics of each value con-figuration are summarized in Table 1. Here wewill expand on the configurations by contrastingthem.

We first note that all three configurations havein common a focus on critical value activities,the distinction between primary and supportactivities, and the analysis of cost and valuedrivers as a means to translate a value configu-ration analysis into a competitive strategy. Theprimary activity categories capture the main dif-ferences between the configurations, while the setof support activity categories is not a distinctiveattribute of the three alternative value configu-rations.

The long-linked technology transforms objectsaccording to a predefined set and sequence ofactivities. The intensive technology solves prob-lems by a custom combination of activities. Themediative technology is provided by a standardcombination of activities at multiple levels. Thedistinctive dimensions are thus to what extent thetechnologies rely on standard or custom combi-nations of activities and to what extent the valuecreation logic is transformative or mediative. Acustomized mediative service has elements ofboth the shop and network. Firm strategy mightinvolve choosing what value creation logic toemphasize.

Coordination of distinct value creation logicsis the concern of organizational design andadministrative theory. Thompson’s (1967) pri-mary concern was how organizations dealt withuncertainty. He proposed three distinct approaches


to deal with uncertainty in what we have labeledrespectively chains, shops, and networks: by verti-cal extension up and down the chain, by incorpo-ration of the problem in the shop, and by increas-ing the size of network served. All three providea means to deal with uncertainty and buffer thecore technology of the organization from environ-mental uncertainties.

A key concept of contingency theory (see, forexample, Lawrence and Lorsch, 1967) is thatdifferent kinds of businesses need to be manageddifferently. According to Porter (1985: 23), themain implication of value chain analysis is theneed for different degrees of differentiation andintegration across activites and functions. Thedistinction between chain, shop and networkextends this argument by suggesting that thereare distinct configurations based on the logic ofcoodination (Mintzberg, 1979).

Although it is beyond the scope of this paperto develop the ideas in detail, referring to Mintz-berg’s design alternatives (1979), we suggest thatwhile the value chain requires a machine bureauc-racy organization of primary activities, the valueshop is organized according to either the pro-fessional bureacuracy or the operational adhoc-racy. The value network is often organizedaccording to an administrative adhocracy, partic-ularly when the technology of the infrastructureis complex and requires highly specialized devel-opment activites such as is the case with moderntelecommunications.

The link to alternative organizational designsand configurations reinforces the notion that thealternative value configurations are quite distinct.The link to the organizational design literature isalso potentially important in that it integratesstrategy and structure—by providing a commonframework for the development of a competitivestrategy and the organizational structure requiredto implement the strategy.

Differences in value creation logic reflect dif-ferent economics. These relate to three importantand distinct traditions in the study of the eco-nomics of the firm: the cost economics of scaleliterature for the value chain, the positive networkvalue externalities literature for the value network,and the value-signalling literature for the valueshop. As shown in Table 1, the logics differ interms of a cost or value focus. While the chainhas a cost orientation, the shop is orientedtowards value. The value network—where synch-


ronization of simultaneous, parallel primary activ-ites is the foundation of value creation—needsto balance cost and value as scale and capacityutilization are drivers of both.

Thus there are distinct scale logics. Scale is acost driver in the chain. Scale is a cost and valuedriver in the network. In the shop scale primarilyaffects value to the extent that it signals success.

The distinct economics can be further illus-trated by the role of queues and the nature ofcontracts across the three configurations. Queues,in terms of input and output buffers, assure highcapacity utilization in the chain. Queues representpotential service and value degradation in thenetwork when they lead to excessive capacityutilization, whereas ‘external’ queues signalpotential service quality in the value shop.

Simplified, contracts in the chain are linked tothe actual exchange of products, are often implicitand are largely governed by general rules formarket-based transactions. Contracts in the net-work are explicit and govern both access to themediation service and the actual use of the ser-vice. Contracts in the shop are mainly implicit,but often policed and enforced by a professiongiven the information asymmetry in the relation-ship between the firm and the client.

Even though our analysis appears to provideextensive support for both the strategic impor-tance and the analytical significance of the threevalue creation logics, the issue of alternativesto the proposed primary activity decompositionremains. What are possible alternative genericdecompositions?

We have already suggested that instantiationof the generic models raises the issues of distinctrepertoires of primary activity configurationswithin the three main alterantive value creationlogics. This is obviously an issue for futureresearch. We can, however, comment on to whatextent one might capture the alternative valuecreation logics with the generic value chain rep-resentation.

Recollect that problems in applying the valuechain motivated the development of the alterna-tive representations. For certain types of intensivetechnology, such as oil exploration and productdevelopment, one alternative would be to rep-resent shop as an ‘information refinery’ that trans-forms data into client solutions (discoveries andnew products). This representation loses the prob-lem-contingent characteristics of value creation,


is most suited to capture sequential flow-dependent cost economics, and sees value asprimarily a function of data quality and trans-formative quality of the process.

Applying the value chain representation tomediation, we either need to introduce an artificialdistinction between clients as being either supliersor customers, or end up with all activities inoperations. Stated differently, the proposed pri-mary activity categories of the value network area decomposition of what would be defined asoperations in an equivalent value chain represen-tation. The value network activity decompositionavoids the artificial distinction between clientroles.

Most firms are not pure instances of a singledistinct value configuration. A single firm mayemploy more than one technology and hencehave more than one configuration. A particulartelephone company (consider the old AT&T)employs a mediating technology for operationsof its telephone services and a long-linked tech-nology in the production of equipment. A manu-facturing company employs a long-linked tech-nology to produce its products whereas productdevelopment relies on an intensive technology.The recent argument of Normann and Ramirez(1993) on value constellations and interactivestrategies can perhaps be understood as an argu-ment for combining several value-creating logics,where the production of goods (chain) is sup-plemented by both assisting customers in theirproblem-solving (shop) activities and by value-adding new services and products on the distri-bution infrastructure (value network) developedby, for example, a firm such as IKEA. In short,hybrids might have distinct logics in primaryactivities and support activities. Hybrids mightalso have an overall primary activity logic, butwhere decomposition of activities requires theapplication of other value creation logics. Thechallenges of effective integration and coordi-nation across different value creation logicspresent unique opportunites for competitiveadvantage (Lawrence and Lorsch, 1967).

The business value systems reflect the activityinterrelationships and drivers of the respectiveunderlying value configurations. Value chainsform sequentially interrelated value systems ofsuppliers, producers, and distributors, each addingvalue to the output from the preceding chain.Value shops are linked and referred in a wheels-


outside-wheels relationship to specialized prob-lem-solving and implementation activities. Valuenetworks form coproducing layers of mediatorswhere one network may use a lower-level net-work as a subnetwork. In addition value networksform horizontal interconnected value systems ofsimilar firms that extend the scope of the networkby virtual mergers to gain mutual benefits fromnetwork externalities. The resulting scope is equi-valent to the horizontal union of the verticalintersection of customer contracts. Most businessvalue systems include firms representing all valuecreation logics.

Our analysis suggests that there are distinctstrategic business system-level implications of thedifferent firm-level value configurations whenthey aggregate and interact in the business valuesystem. Although this is an interesting issue forfurther research, we already noted that for bothshops and networks the business value systemmust include competitive and cooperativerelationships. The simultaneous, coproducing na-ture of a system value of networks requires com-mon standards. Similarly, the knowledge stan-dards of professions facilitate relationshipsbetween firms in a system of value shops.

CONCLUDING COMMENTS

Strategy can be defined as the art of creatingvalue (Normann and Ramirez, 1993). Althoughwe agree that strategy is and will largely remainan art, our project is directed at contributing tothe development of the science of value creation.

By introducing two additional value configur-ations, we have aruged that the concepts promotedin value chain analysis are adaptable beyond thetraditional manufacturing context to which itsdescription and sequencing of activities are bestsuited. We are also suggesting that choice of(emphasis of) value configuration is an additionaldimension or third option beyond Porter’s twobasic strategies of cost advantage and differen-tiation (1980, 1985).

The notion of alternative value configurationsis in part motivated by the problems of applyingvalue chain analysis, both as an effective concep-tual tool and as a means to benchmark andimprove a firm’s competitive position. As notedby Porter, problems of assessing costs imply alsoa potential opportunity for competitive advantage.


The challenge for further research is to demon-strate that the three configurations give a betterhandle on competitive cost and value analyses.Equally important, such research on the use andusefulness of the proposed value configurationmodels needs to be supplemented and based onresearch that develops the structural properties ofalternative value creation technologies. Althougha vast amount of research exists that might beused to illuminate these issues, there is probablya need for a new look and additional empiricalwork that explores the issues from a firm-levelstrategy and strategic perspective, both in termsof pure forms and more hybrid forms. Finally,there are perhaps equally interesting challengesin considering the implications at the businessvalue system level.

ACKNOWLEDGEMENTS

We thank Bente Løwendahl and two anonymousreferees for valuable comments and suggestionson earlier drafts of this paper. We also thank theIBM Client Executives class that challenged usand the established tools of strategic analysis. ThePETROPOL Program of the Norwegian ResearchCouncil provided support for the research.

REFERENCES

Abbott, A. (1988). The System of Professions: AnEssay of the Division of Expert Labor. Universityof Chicago Press, Chicago, IL.

Akerlof, G. A. (1970). ‘The market for lemons: Quali-tative uncertainty and the market mechanism’,Quar-terly Journal of Economics, 84, pp. 488–500.

Alderferer, C. P. (1987). ‘An intergroup perspective ongroup dynamics’. In J. Lorsch (ed.),Handbook ofOrganizational Behavior. Prentice-Hall, EnglewoodCliffs, NJ, pp. 190–222.

Armistead, C. G. and G. Clark (1993). ‘Resourceactivity mapping: The value chain in service oper-ation strategy’, Service Industries Journal, 13(4),pp. 221–239.

Barney, J. B. (1991). ‘Firm resources and sustainedcompetitive advantage’,Journal of Management, 17,pp. 19–120.

Bental, B. and M. Spiegel (1995). ‘Network compe-tition, product quality, and market coverage in thepresence of network externalities’,Journal of Indus-trial Economics, 43(2), pp. 197–208.

Clark, K. B. and S. C. Wheelwright (1993).ManagingNew Product and Process Development. Free Press,New York.


Cox, W., N. F. Hartung, H. Hochberg, B. J. Lewis, D.H. Maister, R. F. Mattox and P. A. Piven (1987).Success Strategies for Design Professionals: Super-Positioning for Architecture and Engineering Firms.McGraw-Hill, New York.

Crane, D. B., R. C. Merton, K. A. Groot, Z. Bodie,S. P. Mason, E. R. Sirri, A. F. Perold and P. Tufano(1996). The Global Financial System: A FunctionalPerspective. Harvard Business School Press, Bos-ton, MA.

Crown Cork and Seal Company, Inc. (1977). HBS Case9-378-024, Harvard Business School, Boston, MA.

Dalton, G. W., P. H. Thompson and R. L. Price (1977).‘The four stages of professional careers: A newlook at performance by professionals’,OrganizationDynamics, 6(1), p. 19–42.

de Chalvron, J. G. and J. G. Curien (1978). ‘Infor-mation, energy and labor force’. In M. C. J. Elton,W. A. Lucas and D. W. Conrath (eds.),EvaluatingNew Telecommunication Services. Plenum Press,New York, pp. 224–246.

Domowitz, I. (1995). ‘Electronic derivatives exchanges:Implicit mergers, network externalities and stan-dardization’, Quarterly Review of Economics andFinance, 35(2), pp. 163–175.

Eccles, R. G. and D. B. Crane (1988).Doing Deals:Investment Banks at Work. Harvard Business SchoolPress, Boston, MA.

Forestieri, G. (1993). ‘Economies of scale and scopein the financial services industry: A review of recentliterature’. Financial Conglomerates. OECD, Paris,pp. 63–124.

Friedson, E. (1960). ‘Client control and medical prac-tice’, American Journal of Sociology, 65, pp. 374–382.

Galbraith, J. (1973).Designing Complex Organizations.Addison-Wesley, Reading, MA.

Hallwood, C. P. (1990).Transaction Costs and Tradebetween Multinational Corporations: A Study of Off-shore Oil Production. Unwin Hyman, London.

Hammer, M. (1990). ‘Reengineering works: Don’t auto-mate, obliterate’,Harvard Business Review, 68(4),pp. 104–112.

Hax, A. C. and N. S. Majluf (1992).The StrategyConcept and Process: A Pragmatic Approach. Pren-tice-Hall, Englewood Cliffs, NJ.

Hayek, F. (1945). ‘The use of knoweledge in society’,American Economic Review, 35, pp. 519–530.

Hergert, M. and D. Morris (1989). ‘Accounting data forvalue chain analysis,Strategic Management Journal,10(2), pp. 175–188.

Jones, P. E. (1988).Oil: A Practical Guide to theEconomics of World Petroleum. Woodhead, Cam-bridge, UK.

Karpik, L. (1989). ‘L’economie de la qualite´’, ReviewFrancaise de Sociologie, 30(2), pp. 187–210.

Katz, M. and C. Shapiro (1985). ‘Network externalities,competition, and compatibility’,American EconomicReview, 75, pp. 424–440.

Kornai, J. (1971).Anti-Equilibrium. North Holland,Amsterdam.

Lawrence, P. R. and J. W. Lorsch (1967).Organizationand Environment: Managing Differentiation andIntegration. Irwin, Homewood, IL.


Lorsch, J. W. and P. F. Mathias (1987). ‘When pro-fessionals have to manage’,Harvard BusinessReview, 65(4), pp. 78–83.

Løwendal, B. R. (1992). ‘Global strategies for pro-fessional service firms’, unpublished Ph.D. disser-tation, The Wharton School, University of Pennsyl-vania, Philadelphia, PA.

Løwendahl, B. R. (1993). ‘Strategic management ofprofessional business service firms: Three genericstrategies’, working paper, Norwegian School ofManagement, 1993/32.

Maister, D. H. (1993).Managing the Professional Ser-vice Firm. Free Press, New York.

Mintzberg, H. (1979).The Structuring of Organiza-tions. Prentice-Hall, Englewood Cliffs, NJ.

Morecroft, J. D. W. (1992). ‘Executive knowledge,models and learning’European Journal of Oper-ational Research, 59, pp. 9–27.

Nelson, R. R. (1991). ‘Why do firms differ, and howdoes it matter?’,Strategic Management Journal,Winter Special Issue,12, pp. 61–74.

Newell, A. and H. Simon (1972).Human ProblemSolving. Prentice-Hall, Englewood Cliffs, NJ.

Normann, R. and R. Rairez (1993). ‘From value chainto value constellation: Designing interactive strat-egy’, Harvard Business Review, 71 (4), pp. 65–77.

Perrow, C. (1961). ‘Organizational prestige: Some func-tions and dysfunctions’,American Journal of Soci-ology, 66, pp. 335–341.

Porter, M. (1980).Competitive Strategy: Techniquesfor Analyzing Industries and Competitors. FreePress, New York.

Porter, M. (1985).Competitive Advantage: Creatingand Sustaining Superior Performance. Free Press,New York.

Porter, M. (1990). The Competitive Advantage ofNations. Free Press, New York.

Porter, M. (1991). ‘Towards a dynamic theory of strat-egy’, Strategic Management Journal, Winter SpecialIssue,12, pp. 95–117.

Pounds, W. F. (1969). ‘The process of problem find-ing’, Industrial Management Review, 11(1), pp. 1–19.

Prahalad, C. K. and G. Hamel (1994). ‘Strategy as afield of study: Why search for a new paradigm?’,Strategic Management Journal, Summer SpecialIssue,15, pp. 5–16.

Reed, R and R. J. DeFillipi (1990). ‘Causal ambiguity,barriers to imitation and sustainable competitiveadvantage’,Academy of Management Review, 15(1),pp. 88–102.

Schonberger, R. J. (1990).Building a Chain of Cus-tomers. Hutchinson Business Books, London.

Simon, H. (1977).The New Science of ManagementDecision. Prentice-Hall, Englewood Cliffs, NJ.

Simon, H. (1982).The Sciences of the Artificial(2nded.). MIT Press, Cambridge, MA.

Spence A. M. (1973).Market Signalling: InformationTransfer in Hiring and Related Processes. HarvardUniversity Press, Cambridge, MA.

Stabell, C. B. (1982). ‘Office productivity: A microeco-nomic framework for empirical research’,Office:Technology and People, 1(1), pp. 91–106.

Stabell, C. B. (1983). ‘Putting the D back into decision


support systems’. In J. Bennett (ed.),BuildingDecision Support Systems. Addison-Wesley, Read-ing, MA, pp. 221–260.

Stoeckle, J. D. (1987).Encounters between Patientsand Doctors. MIT Press, Cambridge, MA.

Tanenbaum, A. (1981).Computer Networks. Prentice-Hall, Englewood Cliffs, NJ.

Thompson, J. D. (1967).Organizations in Action.McGraw-Hill, New York.

Union Carbide Deal (1988). HBS Case 9-288-065,Harvard Business School, Boston, MA.

Wade, J. (1995). ‘Dynamics of organizational com-


munities and technological bandwagons: An empiri-cal investigation of community evolution in themicroprocessor market’, Strategic ManagementJournal, Summer Special Issue,16, pp. 111–133.

Wernerfelt, B. (1984). ‘A resource-based view of thefirm’, Strategic Management Journal, 5(2),pp. 171–180.

Winch, G. and E. Schneider (1993). ‘Managing theknowledge-based organization: The case of architec-tural practice’, Journal of Management Studies,30(6), pp. 922–937.

CONFIGURING VALUE FOR COMPETITIVE ADVANTAGE: ON CHAINS, SHOPS, AND

Documents

Transcript of CONFIGURING VALUE FOR COMPETITIVE ADVANTAGE: ON CHAINS, SHOPS, AND