innovation, time, and territory: space and the business … · 2020. 12. 11. · space and the...

119

Business firms and the forms of organi-zation that they assume in pursuit of profitare inherently territorial (Walker 1988).Companies assume territorial attributes inthe ways in which they organize the differentactivities that go into creating and sellingproducts and the way they allocate theseactivities within their own enterprises andamong other firms with which they interact.Such managerial allocation of the varioussteps that are involved in the production andsale of goods creates linkages of economicactivity within and across firms. Fromthese intrafirm and interfirm linkagesemerge forms of business organization thatare marked by boundaries within and

between enterprises that reflect an organi-zational distribution of economic activity.This organizational distribution of economicactivity, built from the boundaries withinand between firms, spreads economicactivity territorially across space. In this way,business organization has territorialoutcomes. This article reveals how busi-ness firms reshape the territorial configu-ration of economies in creating new formsof organization as part of the innovationprocess.

Innovation and technological change playdecisive roles in the geographic transfor-mation of economic activity (Storper andWalker 1989; Angel 1994; Castells 1996).

Innovation, Time, and Territory:Space and the Business Organization of Dell Computer

Gary FieldsDepartment of Communication, University of California,

San Diego, CA [email protected]

Abstract: Businesses reshape the territorial configuration of economic activity bycreating new forms of organization as part of the innovation process. Focusing onthe case of Dell Computer, this article builds an argument about the geographicdevelopment of economies that is structured around four elements: (1) the firm,(2) the innovation process occurring within the firm, (3) the business organizationof the firm, and (4) the territory in which the firm operates and extracts profit. Intracing this route from the enterprise to territory, this article draws upon thenotion of “communications revolutions” as a catalyst for the innovative impulse offirms. In adapting to communications revolutions, firms, such as Dell, emerge asinnovators by learning to recalibrate the time increments that are expended duringthe steps of making and marketing products, thereby shifting competition from theproduct to the processes of capitalist circulation. This recalibration of time resultsin the creation of process-driven routines for making profit and a transformation inthe organizational arrangements through which firms implement such routines. Asfirms reinvent forms of organization to implement these time-based innovativeroutines, they alter the linkages between adjacent steps in their profit-making activity.By reorganizing these linkages and changing the nature of business organization,innovative firms reconfigure the territory in which they operate and accumulate.This reconfiguration of territory, however, is not the mechanical result of effi-ciency criteria. Firms use power over other firms to redeploy the location ofactivity in their production networks in an effort to achieve time economies andinnovative efficiencies.

Key words: innovation, territory, business organization, communications, DellComputer.

#2780—ECONOMIC GEOGRAPHY—VOL. 82 NO. 2—82201-fields

Economic Geography 82(2): 119–146, 2006.© 2006 Clark University. http://www.clarku.edu/econgeography

120 ECONOMIC GEOGRAPHY APRIL 2006

Nevertheless, while the relationship betweeninnovation and the spatial reorganizationof economies is well established in thegeographic literature, what is less clearlyspecified are the mechanisms by which inno-vation transforms the geography ofeconomies. Part of the problem involves thechoice of what is to be studied as a unit ofanalysis from the myriad approaches to thepuzzle of innovation and spatial change.Some researchers have insisted on estab-lishing the connection between innovationand territory through the study of specificplaces and the cultures of interactive learningthat flourish in such places (Saxenian 1994).Others with a similar perspective haveshifted the emphasis from specific placesto more generalized territorial units, mostnotably the region or the nation, inaccounting for the emergence of regional ornational systems of innovation (Asheimand Gertler 2005; Nelson 1993). A differentapproach connects innovation to the spatialtransformation of economies through thestudy of specific industries and the shiftinglocational patterns of industries as theydevelop new products and processes andseek territorial outlets for producing andselling in new ways (Angel and Engstrom1995). Still another approach focuses on theprocess of capital accumulation in whichterritorial change results from the creativelydestructive and innovative tendencies ofindustries as they seek new spaces for invest-ment, resources, markets, and growth(Storper and Walker 1989). A final group ofresearchers have uncovered the nexusbetween innovation and territorial changein forms of business organization—commodity chains and interfirm networks—that emerge in response to new technolo-gies and new relationships among leadproducer firms, suppliers, and buyers(Gereffi and Korzeniwicz 1994; Castells1996).

This article draws from each of these liter-atures, but uses an “actor-centered”approach to the issue of business organiza-tion in uncovering the sources of spatialchange in economies (Markusen 1994,

2003).1 It focuses on the competitive expe-rience of the business firm, reconfiguringterritory as an agent making strategic deci-sions about accumulation and innovation andacting on such choices. Its aim is tocontribute to a surprisingly limited theo-retical and empirical literature in economicgeography on the nexus between the firmand territory (see especially Maskell 2001;Dicken and Malmberg 2001; Taylor andAsheim 2001). Thus, this article situates thefirm at the center of changes in economicgeography, while examining the specificmechanisms that link the business enterpriseas an agent to innovation and territorialtransformation.

In developing this connection betweenbusiness enterprise and territory, thisstudy builds an argument that is structuredaround four elements: (1) the firm, (2) theinnovation process occurring within the firm,(3) the business organization of the firm, and(4) the territory in which the firm operatesand accumulates profit. It connects the firmto territorial development by focusing on theorganizational attributes of innovation andthe spatial attributes of business organiza-tion. To establish this connection among thefirm, innovation, business organization, andterritory, this study elevates as its centralprotagonist one of the most pioneeringenterprises of the current period, DellComputer Corporation.2


1 Markusen (2003) argued forcefully how inmuch of the economic geography literature,“abstract processes are the principal actors” andhow our understanding of spatial change has beencompromised by a move away from real actorsthat are involved in transforming economies,primarily firms and workers.

2 Following Schoenberger (1991), the studyused interviews with procurement and logisticsmanagers at Dell, some of whom allowed me touse their names, and interviews with logisticsmanagers at three of Dell’s key suppliers, noneof whom allowed me to use their names orcompany. Dell is extremely guarded about itsbusiness practices and, as a rule, does not partici-

VOL. 82 NO. 2 DELL’S BUSINESS ORGANIZATION 121

By focusing on Dell to trace this routefrom the enterprise to territory, this studydraws upon the catalytic role of “communi-cations revolutions”—new communicationstechnologies and their systems of use—as astarting point for the innovative impulse offirms (John 1994; Albion 1932). It linksthis phenomenon to territorial outcomes byexamining how new communications tech-nology enables firms to exploit economiesof time and speed as a source of innova-tion. It shows how such time-driven processinnovations compel the innovative firm toreorder the sequencing and allocation of thediscrete steps that constitute this innovativeactivity, both inside the enterprise and acrossother firms in the production network of theinnovator. What this article reveals is howthe innovative enterprise restructures theorganizational linkages within and betweenfirms to accommodate time- and speed-driven routines and how in building suchnew forms of organization, the innovativefirm reshapes territorial patterns ofeconomic activity. In this way, it givesgeographic meaning to what is commonlyreferred to as “the nature of the firm,” theproblem that was derived originally fromCoase’s (1937) article on how firms chooseforms of business organization, whileimbuing this problem with a spatial dimen-sion.

Dell is a compelling case in the way itadapted to the communications revolutionof the Internet and learned to accumulateprofit differently. As Dell incorporated theInternet into its business routines, itascended from a decent-sized, but by nomeans dominant, firm in 1994 to the toprank of the personal computer (PC) industryby 2001. In the process, it helped shape anew geography of global profit making, influ-encing a range of other firms, both withinthe PC industry and outside it.

Three questions frame the story of Delland the route from the firm to territory inthis study: (1) How does technologicalchange in communications systems enablebusiness users of these systems to transformtheir strategies and operational routinesfor producing and selling? (2) How do thechanges in strategies and operationalroutines of firms stemming from newcommunications systems result in the trans-formation of business organizations throughwhich firms compete and seek profit? and(3) How do innovations in business organi-zation reshape the geography of economicactivity? With its focus on Dell to addressthese questions, this article uncovers howbusiness users of communications systemsreorient their operational routines and orga-nizational structure as the technology ofthese systems changes and how the spatialpatterning of economic activity gets reshapedas firms deploy new communications tech-nology to compete and accumulate profitdifferently.3

Time, organization, and territory becameintegrally linked in Dell’s innovative advanceto dominance in the industry. Owing to itsdependence on process innovation, Dellbecame a firm that was obsessed withcompressing time between the various stepsof making and selling PCs as the most crit-ical element in its business model. The time-driven business system that Dell devel-oped from Internet technology as part of thisascent is what compelled the PC maker toreconfigure the territorial arrangementsbetween its key operations to make the high-speed logistics of its business model viable.

At the same time, it was the organizationalpower that Dell exerted over firms in itsproduction network that enabled it toforce these other companies to organize their


pate in academic studies in which it will not mate-rially benefit. Consequently, these interviewswere secured through unofficial channels, as werethe interviews with suppliers.

3 Methodologically, this article is an intensivecase study. It reveals how a specific causalprocess—in this case the route from communi-cations to territory—is observable in the case itselfand how, in reflecting this process, the case isrepresentative of a broader, more generalized,trend (see especially Sayer 1992, 242–43).


activities in certain spatial patterns, so thatDell could profit from the innovative timeeconomies it had developed from theInternet. In implementing its time-drivenroutines, Dell, in effect, confronted a chal-lenge that firms have faced throughoutperiods of capitalist development. Dell wascompelled to reshape the landscape to accessthe benefits of a shift in the temporal dimen-sions of producing and selling, a tendencythat is often referred to as “spatial fix”(Harvey 1982; Schoenberger 2004). In whatis a sharp rebuke to the idea of Internet tech-nology simply collapsing time and space, thecase of Dell reveals how the Internet, farfrom providing the PC maker with the capa-bilities to escape the frictions of geog-raphy, made it more dependent than everon proximity between key nodes in itsnetwork.

The focus of this article on operations andorganization as the basis of economic terri-tory has important implications for the studyof industrial geography. With Dell as a refer-ence point, the article expands on notions ofindustrial geography that are derived froman emphasis on the locations of facilitiesalone. While not discounting the role of facil-ities in the spatial patterning of economicactivity, this study broadens ideas aboutthe construction of economic territory byfocusing on the operations of the firm andthe organizational connections betweenthem.

This article is organized in four sections.The first section presents a theoreticalframework for the argument that links thefirm to innovation, organizational change,and territorial transformation. The secondsection describes the PC industry prior toDell and how the competitive structure ofthis industry provided Dell with an oppor-tunity for process, as opposed to product,innovation. The third section examines howDell used the Internet to transform itsprocess-oriented business routines and orga-nization. The fourth section analyses theterritorial consequences of Dell’s innovativeadvance. Through an intensive case study,this article engages in a theoretical conver-sation about the territorial dimensions of the

innovation process and the geography ofprofit seeking and power that are at the foun-dation of contemporary capitalist develop-ment.

Theoretical FrameBusiness firms are actors that shape the

spatial development of economies. As aconsequence, “studying firms” in their roleas agents that make decisions about profitmaking is a logical imperative for under-standing how economies develop spatially(Markusen 1994, 2003). Numerousapproaches to the issue of the firm and itsrole in spatial outcomes characterize thegeographic literature.

Theorists, influenced by the insights ofAlfred Marshall (1890) on externaleconomies and industrial districts, haveargued that the firm is a place-basedentity. According to this view, the innova-tive behavior and competitive differentia-tion of firms and their role in driving terri-torial development are derived from theunique history, culture, and institutions ofthe locations in which the firms areembedded (Brusco 1982; Saxenian 1994;Herrigel 1996; Storper 1997; Scott 1998;Gertler 1995; Asheim and Gertler 2005).Others have studied the firm and its impacton the geography of economic activity withinthe context of broader organizational forms,such as the production network orcommodity chain (Porter 1985; Gereffiand Korzeniwicz 1994; Castells 1996;Sturgeon 2002; Gereffi, Humphrey, andSturgeon 2005). Perhaps most common isthe approach that positions firms withinindustries (Schoenberger 1986; Angel 1994;Angel and Engstrom 1995; Harvey 1975,1982; Storper and Walker 1989; Scott 2005).In these representations, what is decisivein driving the process of spatial formationand differentiation are historical, cultural,economic, and organizational forces that actupon and operate outside the business enter-prise.

A different orientation to studying firmsas drivers of territorial development focuseson what occurs inside the enterprise. This



approach takes as its starting point theprocess of organizational decision makingand the contingent activities that occurwithin the so-called black box of the firm(Rosenberg 1982, 1994; Nelson and Winter1982; Dosi 1988). Yet, this framework thatfocuses on the firm as an actor remainsuncommon in economic geography owingto the absence of microlevel theory on the“firm-territory nexus” (Markusen 2003;Maskell 2001; Dicken and Malmberg2001; Taylor and Asheim 2001).

Two critical insights emerge from theemphasis on the activities and choicesinternal to the enterprise that provide theanalytical bridges between the firm andthe territorial development of economies.The first focuses on the innovative nature ofthe firm. The second focuses on the orga-nizational nature of the innovation process.

Although innovation occupies a positionof centrality in the literature on the spatialdevelopment of economies, the role of theinnovative firm in such outcomes is exploredfar less systematically. Schumpeter (1939,1942) was arguably the first theorist toelevate the firm as the central actor andprotagonist of economic development. Yet,even he conceded that the activities occur-ring inside the firm that enable it to assumethe role of innovator remained unspecifiedin his work (Schumpeter 1947). Theoristswho have been influenced by Schumpeterhave argued that innovation is a process oflearning in which firms assume mastery overnew sets of organizational capabilities(Nelson and Winter 1982; Rosenberg1982; Dosi 1988, 1997; Lazonick and Mass1995; O’Sullivan 2000). Such an augmenta-tion of capabilities, in turn, becomesembedded within the enterprise as changesin organizational structure (Chandler1962, 1977). In this way, innovation is aprocess with organizational consequences.

Yet, to understand innovation as a changein the organizational structure of the firm,it is necessary to probe the fundamentalnature of business organization. For Coase(1937), business organization is the result ofdecisions by firms on how to manage andcoordinate the linkages among the different

steps in producing and selling a product orservice. Such decisions establish organiza-tional boundaries within and between firms.These boundaries become fixed, dependingon the extent to which firms absorb differentsteps internally and become integrated andthe extent to which firms leave differentactivities to contracting relationships withother companies, resulting in disintegratedmarket links between companies. For Coase,choices about whether to absorb sequen-tial steps of economic activity internally orcontract with other companies depend upona single variable—the transaction costs ofundertaking such activities internally orexternally. This model yields two basictypologies for business organization, one thatis integrated and hierarchical and the otherthat consists of highly specialized disinte-grated firms contracting through marketlinks (Coase 1937; Williamson 1975).

Critics have assailed the model inspiredby Coase for neglecting organizational forms“neither market nor hierarchy.” These criticsargue that networks of firms constitute aunique form of business organization existingoutside the continuum represented by verti-cally integrated firms and firms contractingthrough markets (Powell 1990; Castells 1996;Amin and Hausner 1997). Such intermediateorganizational forms reflect a more complexdivision of labor between firms (Gereffi andKorzenwicz 1994; Sturgeon 2002; Gereffi,Humphrey, and Sturgeon 2005). Othercritics have insisted that transaction costs asa singular explanation for business organi-zation is fundamentally misguided. For thesecritics, business organization results fromefforts to augment capabilities, not to econ-omize on costs (Chandler 1977; Lazonick1991, 2003; Langlois 2003).4 As firms assumemastery of new capabilities and remake theirorganizations, they reassess the allocationand coordination of the different steps in


4 In commenting on the work of Williamson(1975), Chandler (1988) conceded that changesin business organization may result in reductionsin transaction costs for the firm, but he arguedthat such reductions were by-products, notcatalysts, of organizational change.


producing and selling products (Chandler1988). From these reallocations of economicactivity and transformations in capabilitiesemerge new operational routines forproducing and selling products, and newforms of organization to implement them(Chandler 1977).

Although Chandler’s theory of businessorganization contains a highly determinedrelationship among strategy, capabilities, andstructure (McCraw 1988, 17), his empiricaldata revealed the route to organizationalchange to be the result of contingent deci-sions exercised by actors within the firm. Theinnovative organization is one that actsstrategically by using its own administra-tive power to control the allocation andsequencing of economic activity, a processof replacing market links in the economywith the managerialism of “the visible hand”(Chandler 1977). Such nonmarket forms ofpower and control constitute “the socialfoundations of the innovative enterprise”(Lazonick 1991, 2003). In this way, innova-tion and organizational change are socialprocesses. It is this social character ofinnovation and organizational change,emphasizing the power of the firm to control,that elevates firms as actors and conditionsthe trajectory of an innovative advance.

Toward a Geography of theInnovative Organization

By positioning the firm within this frame-work of innovation and organization, it ispossible to outline how enterprises such asDell transform their capabilities and orga-nizational structure and, in the process,reshape the territory of profit making. As aterritorial entity, the firm possesses two basicattributes. First, the firm as an embodimentof territory emerges from the stocks of facil-ities where it locates and manages work.Second, the firm as a territorial entityemerges from the flows of material andinformation between its own stocks of assetsand those of other firms with which itinteracts. The economic activity that isderived from the linkages between stocksand flows and the organizational connections

that are developed to coordinate and managethese activities create patterns in space andgive the firm its territorial attributes.

The organizational arrangements that areused to coordinate the activity linking facil-ities and flows reflect choices made by firmson the extent to which they internalize oper-ations and contract with other companies inproducing and selling. These decisionsdefine the organizational boundaries andstructure of the firm. In demarcating bound-aries between its own capabilities and thoseof other companies, the firm allocateseconomic activity organizationally and,through such allocation, distributeseconomic activity spatially. In this way, firmsare territorial in the way they strategicallychoose to organize the sequencing and fulfill-ment of their profit-seeking activities.

Among historical factors that affect theenvironment of strategic choice for profitmaking, one of the most decisive is the tech-nology in systems of communication and thebuild-out of such systems (Lee and Whitley2002, 235). When the technology for thesesystems changes and the infrastructurethat is based on such new technology is builtout, society experiences the beginning ofcommunications revolutions (John 1994).The Internet represents a historically specificinstance of this phenomenon. Nevertheless,communications revolutions, such as theInternet, do not result from new infra-structure alone (Abbate 1999; Fischer 1992).Users of the new infrastructure, especiallybusiness users deploying the new systems intheir business models, are what enablecommunications revolutions to spread andbecome generalized throughout theeconomy.

When put into use, new communica-tions technologies reshape the systems ofaccess and circulation by which users ofthese systems secure the resources—mate-rial, informational, and human—and theoutlets to customers that are necessary forproducing and selling. What these firmsexploit from communications revolutions arechanges in systems of access and circulationfor economic activity that is structuredaround relationships of time and space. In



the profit-driven economy, control overthe time and space relationships in thesystems of economic access and circulationis a centrally important strategic, operational,and organizational problem for the firm(Schoenberger 1997, 12). Systems ofaccess and circulation are what provide firmswith routes to the inputs and sales outletsthat are necessary for producing andselling goods and services. By enabling firmsto recalibrate the time and space relation-ships in systems of access and circulation, anew communications infrastructure createsopportunities for firms to reorganize howthey procure materials, combine materialsto produce goods and services, obtaincustomers, and market what they produceto customers.

Communications revolutions, in effect,are control revolutions (Beniger 1986;Chandler 1977).5 They change the envi-ronment of profit making by providing firmswith possibilities to control the time andspace relationships in economic activity innew ways. Such control alters how firmsreach customers, process orders, procureresources, and produce goods and services.These changes in systems of access, inturn, enable firms to recalibrate the pacingof economic activity and rearrange theseactivities spatially to process output by meansof the newly discovered economies of timeand speed.

Although bound together, time and spaceplay different roles in the reorganization ofbusiness activity. It is actually time thatconstitutes the basis of value and profit incapitalist production (Harvey 1996, 241).For this reason, the recalibration of time—altering the pacing of business processes anddriving down the increments of labor timeembedded in units of output—is the primarystrategic route that firms take to achievegreater levels of efficiency in businessactivity. Despite this primacy of time in theeconomic life of capitalist production,however, reconfigurations of space are insep-arable from, and a logical outcome of,

temporal changes in business activity(Harvey 1996, 240; Sahay 1997). Firms fixspace as they recalibrate the pacing ofeconomic activity and reassess shifts in theorganization of profit making (Harvey 1982;Schoenberger 2004). Consequently, in theinnovation process, time leads and spacefollows, but the two are linked through thestrategic choices made by the firm.

In this process of spatial fix, firms do notmechanically graft innovative routines andbusiness organization onto pliablegeographic landscapes. Firms reshape terri-tory as part of a social process. To developthe organizational capabilities that drive thereconfiguration of territory, the innovativefirm cooperates with and confronts resis-tance from actors inside the enterprise andfrom firms in its network that play roles inthe deployment of an innovative advance.These relationships inside the innovativefirm and with external actors shape howcapabilities get created and deployed, howorganizations get restructured, and howterritory ultimately gets reconfigured.Territorial transformation is therefore theresult of the search by firms for efficientroutines, mediated by social processes ofcooperation and coercion. In this way, land-scapes, as sites for innovation, are sociallyconstructed terrain.

Inside the firm, the pathway to recastingroutines leads through relationships ofconsent and coercion between managementand the workforce (Marglin 1974). Theextent to which employees, both managerialand nonmanagerial, embrace or resist acompany’s efforts to create new sets of orga-nizational capabilities within the enterpriseinfluences the trajectory and outcomes of aninnovative advance.

At the same time, the innovative enter-prise resorts to relationships of coopera-tion and force with other firms not only togain access to external capabilities, but alsoto spread capabilities to these other compa-nies in order to implement an innovativeadvance. Such external relationships, builtaround mobilizing and extending organiza-tional capabilities, are integral to the inno-vation process because the firm undertaking


5 For a different view of communications revo-lutions as control revolutions, see Yates (1989).


the innovation, no matter how self-reliant orintegrated, is embedded in interfirmnetworks. These interfirm value chains orcommodity chains (Porter 1985; Gereffi andKorzeniewicz 1994) reveal patterns of coop-eration, as well as conflict, as lead firmsand suppliers compete to control the deploy-ment path of innovation to secure more ofthe rents that are generated by an innova-tive advance. Such relationships affect inno-vative outcomes, including the timing andpacing of new routines; how new routinesare allocated organizationally; and the terri-torial patterns where routines are performedand raw materials, along with semifinishedand final products, circulate. In this way,socially constructed interactions among firmsmediate the changes in time and space atthe core of innovation.

To exploit innovative time economies andreconfigure space to accommodate theseeconomies of speed, the innovative firm usescombinations of cooperation and force withits network partners in pursuit of twobroad aims: to enforce locational choices onfirms with whom it interacts in its networkand to impose sets of operational impera-tives on these other companies in placeswhere they have “chosen” to establishfacilities. This interplay of cooperation anddomination enables the innovator to restruc-ture the operations in the network, reallo-cate the operations organizationally, andreconfigure the pattern of facilities and flowsin the network to align the timing,sequencing, and territorial spacing of oper-ations. Such uses of organizational powercomplement the capacity of the innovativefirm to deploy efficiency criteria alone inrealigning time and space in economicactivity and in developing innovative busi-ness practices. Consequently, the transfor-mation of time in business routines and thefixing of space to accommodate thesetemporal shifts is a story about efficiencyand power.

Territory for economic activity is thus theoutcome of objective and subjective consid-erations. Firms reconfigure territory onthe basis of efficiency criteria. Territory isalso reshaped by relations of consent and

conflict between innovators and actors, bothinternal and external to the innovative enter-prise. Although this article links firms toterritory by focusing on the spatial attributesof business organization, it seeks a role forfirms as agents of power in shaping businessorganization through open-ended negotia-tion with other firms on the impacts of inno-vation. Some firms achieve decisive levels ofcontrol in this process of negotiating theimpacts of innovation. Dell Computer is onesuch enterprise.

Dell and the PC IndustryDell Computer entered the PC industry

in 1984 essentially as a logistics firm(Fields 2004). It created a business modelfrom the process of building and selling itsproducts, exploiting the sphere of capitalistcirculation as a source of accumulation, andextracting profit from the gains of trade.Such gains refer to increments of new valuethat accrue to raw and semifinished mate-rials as they change location and assumedifferent attributes in circulating fromprocurement to production to final sale.6

Dell uncovered how to extract profit fromgains of trade in two principal ways: (1) byeliminating intermediaries in the route fromproducer to consumer and capturing thatportion of the gains from trade normallyaccruing to these actors and (2) bycompressing time between the various adja-cent steps in producing and selling PCs,primarily between the final production ofthe PC and final sale to the consumer,thereby cutting costs that are associated withthe time expended in essentially ware-housing the product in inventory as prepa-ration for final sale. That Dell was able tocraft such a logistics-oriented businesssystem focusing on circulation, rather thanproduction, as a source of profit stems


6 This concept differs from the idea in classicaltrade theory derived from Ricardo in which theparties to an exchange secure benefits—intheory—by producing according to the principleof comparative advantage.


from the historically conditioned attributesof the PC industry and role played byIBM.

Open Standards, Logistics, and Dell’sInnovative Advance

When IBM began producing personalcomputers in 1981, its decision to outsourcecomponents and to create an open productarchitecture had two decisive impacts on thePC industry that would affect Dell. First,open standards enabled an array of special-ized suppliers to emerge and exploit oppor-tunities for producing parts for the IBM PC.Second, this open architecture helped enablefirms to clone the PC. As a result, theindustry evolved along a path marked by theproliferation of disintegrated supplierfirms subcontracting to PC makers, who,in turn, became dependent on the externalcapabilities of these suppliers and embeddedin interfirm networks to build the product(Langlois 1992). By 1984, Dell was one ofroughly 100 clone makers that was able toproduce PCs by taking advantage of theexternal capabilities of others.

Although a standardized commodity, thePC was in a state of ongoing technologicalchange, driven primarily by the continuousimprovement in semiconductor and micro-processor technology. Such changescompelled PC makers to develop new prod-ucts at ever-shorter intervals (Dedrick andKraemer 1998, 73). At the same time,these advances enabled costs as a measureof performance to plummet, with the pricefor PCs typically declining by 20 percent to41 percent per year (Curry and Kenney1999, 12).

As a product in flux, the PC was suscep-tible to competitive pressures derived notonly from technology, but from two othervariables—price and its relationship totime (Curry and Kenney 1999). When a newproduct came to market, it was underconstant price pressure, its value shrinkingwith time in anticipation of the next newprocessing technology and application soft-ware. The system of selling through inter-mediaries, the dominant form of distribu-

tion by 1984, only worsened this problem,with a period of 9 to12 months often sepa-rating the procurement of parts and the saleof the final product (Steffens 1994, 175).7Consequently, the indirect selling channel,by delaying the time to market, exacerbateda fundamental weakness with the PC—thedownward pressure on prices stemmingfrom ongoing technical improvements thatimbued the PC with a perishable-like quality(Kraemer, Dedrick, and Yamashiro 1999, 3).This quality posed a challenge to PC makers:how to get the product to the customerfaster. Concurrently, the idea of capturingthat portion of the value taken by interme-diaries presented an equally compellingopportunity for PC firms. In effect, solvingthe twin problems of perishability andentrenched power created a potentialpathway of innovation that was orientedaround logistics and distribution exploitedby Dell.

Dell’s strategy as a new entrant was tochallenge the route of distribution and seekprofit in a more direct relationship to thefinal customer. This strategy created aninnovative advance that ended updisrupting the industry. By using newcommunication technologies, such as thefax, to establish a direct path to the enduser, Dell was able to capture that portionof the value accruing to the PC and takenby intermediaries as the product circulatedfrom assembly to final sale. This direct pathalso represented a strategy for offsettingthe decline in product value that occurredover time. In bypassing intermediaries, Dellsold PCs at prices closer to the value ofcomponents before their inevitable pricedecline associated with time. What Dellaccomplished with this business system wasa reduction in the time that the product satin the sales channel as inventory, therebyeliminating inventory-carrying costs and


7 Even by 1987, with Dell and Gatewayselling direct, computer dealers still accountedfor 56 percent of the total shipments, while thevarious indirect channels together accounted for80 percent to 90 percent of all PC sales (Steffens1994, 260).


providing the firm with a competitiveadvantage over its competitors (Kapuscinskiet al. 2004). It was a strategy for producingPCs—profiting not from production, butfrom the logistics of making and marketingthe product.

As it forged this direct l ink to thecustomer, Dell conceived of anotherinnovation that played a decisive role inits business system. By communicatingdirectly with the end user, Dell sought todeliver a product touted as “customized.”From a menu of fundamentally standard-ized modular parts, Dell assembled slightlydifferentiated systems and sold thesesystems to end users as custom products bycommunicating with them directly,bypassing the intermediaries that domi-nated the industry.

The Early Dell Geography

The geography of this custom direct busi-ness system derived most decisively from thedecision by Dell to perform the assemblywork itself in an Austin, Texas, location whilecontracting for components from suppliers,mostly from East Asia, located long distancesfrom this assembly site. Operationally, thisgeography articulated a pattern of spatiallyextended linkages between two major placeswith flows of materials moving to andconcentrating in one city-region where Dellconfigured these supplies into finished PCs.At the same time, however, Dell’s businessmodel of creating custom-built computerscompelled the firm to establish a distinct setof proximity relationships in which it main-tained numerous warehouses in the Austinarea where it stored components to pullthem into the assembly process as it receivedorders. Consequently, Dell’s operationscreated a spatial pattern built upon twodistinct types of connections, one consistingof long-distance linkages between East Asiaand Austin that delivered components nearthe site of assembly, where they were stagedin warehouses and readied for the assemblyprocess, and the other consisting of linkagesof proximity between the parts warehouses

and Dell’s Austin assembly complex (seeFigure 1).8

Organizationally, the linkages betweenDell and its network partners that movedcomponent parts geographically fromsuppliers’ factories to the Austin warehousesoccurred through markets. Market links arewhat enabled Dell to secure componentsfrom suppliers, while market relationsbetween suppliers and Dell provided thecoordination mechanism for components tochange location in moving geographicallyfrom suppliers’ factories in East Asia to Dell’scomponent warehouses. Michael Delldescribed this process of contracting andcoordination as one built upon “traditionalbid-buy relationships” (Dell and Fredman1999, 180). Consequently, in the geographyof Dell’s business system, an interfirmnetwork organization of separate companiesinteracting through markets moved suppliesfrom one location to another, creating aspatial pattern of long-distance linkagesconnecting East Asia to Austin. In Austin,Dell assumed organizational responsibilityfor the relationship of proximity in movingparts from its warehouses into its ownassembly complex.

Although the interfirm structure of thisnetwork has remained intact until thepresent, Dell was forced to recast themarket-oriented relationships in this busi-ness organization when the time came todeploy the Internet in its business. As Dellused the Internet to uncover new path-ways for controlling time in its business


8 Unlike just-in-time practices in the autoindustry, in which numerous—though not all—suppliers established facilities close to the loca-tions of assembly, Dell’s network reflected thisspatially extended separation between the sourcesof supply and the location of assembly (Angel andEngstrom 1995). Until the early 1990s, Dellwas successful in storing a relatively small inven-tory of components in these warehouses,compared to the levels of its competitors. As aresult, Dell had an advantage over other firmsnot only because of its low inventory of finishedproducts, but also because of the low inventoryof components that it maintained.


system, it was exposed as never before toelevated levels of risk from the frictions ofgeography in trying to fulfill the imperativesof its just-in-time business model. Itsresponse reflected an innovative approachto the challenges of risk and organization,time, and territory.

Dell and the InternetWhen Internet communication emerged

as a commercially viable sales channel in1994–1996, Dell had an advantage overother PC firms in adapting to the newtechnology. Internet selling represented thesame logic of disintermediation that was atthe core of Dell’s business model. As a result,

Dell was the first PC firm to configure andsell its products over the Internet in 1996.Online selling, however, was only an initialset of innovative routines at Dell stemmingfrom the deployment of the Internet in itsbusiness system.

A more profound set of changes resultedfrom Dell extending the Internet intoprocurement and assembly. By deployingthe Internet in these processes, Dell aimedto reduce the time expended in procurementand assembly for information exchange, bothinternally and with suppliers. It accom-plished this aim by creating a more tightlyintegrated, Internet-driven closed-loop formof organization linking order cycles toprocurement and assembly cycles. In this


Figure 1. Dell production network, 1990.


sense, organizational change, orientedaround Internet communication, was inte-gral to an acceleration in the pacing ofeconomic activity. As a practical matter, thisorganizational and temporal shift wasembodied in new routines that Dell devel-oped for its system of material balancing,the way in which it allocates the flows ofmaterial and information over timethroughout the different steps of the PCvalue chain. In transforming these routines,Dell aimed to shrink the increments of timeboth within and between the different oper-ations. At the same time, this innovativeadvance led to the creation of a new form oforganization. Thus, while the system ofmaterial balancing and the logistics oflowering inventory served as the focus ofDell’s innovation, the PC maker wasengaged in a broadly singular strategic objec-tive. Dell committed itself to learning howto use the Internet for controlling thelogistics of time in a fundamentally new andinnovative way.9

Logistics and Time

When Dell entered the PC industry in1984 by recalibrating what had emerged asa decisive competitive variable in the PCindustry—the logistics of time—thecompany was both a pioneer and an inher-itor of an already-established tradition inmanufacturing. In creating a just-in-timeproduction system with little inventory, Delldeveloped a novel approach in the PCindustry, but its strategy echoed the effortdeveloped decades earlier by Toyota in automanufacturing.10 The element of the Toyota


Table 1

Rank of PC Firms, 1994 and 2001 (by World Market Share)

1994 2001Rank Firm % Share Rank Firm % Share

01 Compaq 10.0 01 Dell 14.202 Apple 08.3 02 Compaq 11.103 IBM 08.2 03 IBM 07.204 Packard Bell 05.2 04 Hewlett Packard 06.905 NEC 04.1 05 NEC 05.006 Hewlett-Packard 04.0 06 Apple 04.807 Acer 03.0 07 Siemens 03.408 Toshiba 03.0 08 Gateway 02.809 Fujitsu 03.0 09 Toshiba 02.210 Dell 02.4 10 Fujitsu 02.0

Source: Fields 2004, 184, 214.

four years later, it emerged in the top 10 (seeTable 1). As Dell’s expansion continued apace, itbecame more difficult for the firm to balance thetwo key elements of its business system—customized just-in-time production and little, ifany component inventory—that it had success-fully integrated in the early period. As a result,Dell lost $35.8 million on its sales of $3 billionin 1994, its first—and only—annual loss. Indeed,Dell conceded that its losses stemmed fromexcess procurement inventories. As Michael Delladmitted in his at-times revealing book, “wehad quickly become known as the companywith the inventory problem.” He went on toconcede how, by the early 1990s, Dell was “lastplace in inventory management” (Dell andFredman 1999, 37).

10 On just-in-time systems, see Linge (1991),Mair (1992), Kenny and Florida (1993), and Klier(2000). Where Dell departs from the just-in-time

9 What motivated Dell was a series of setbacks,following a period of extraordinary growth, thatchallenged its logistics-oriented businessmodel. Despite predictions that its direct saleswould not be competitive with the distributionnetworks of other PC makers, by 1990, Dellhad become the twentieth largest PC firm, while


system that Dell imitated most closely wascontinuous flow assembly without thebuildup of inventory. For Toyota, contin-uous flow, absent inventory, compressed thetime cycles between procurement andassembly, enabling the automaker toproduce in high volumes and reduce thecosts related to the buildup of time in eachcomponent.

While control over time is critical in virtu-ally all types of economic activities, suchcontrol was critical in the PC industry andemerged as the fundamental impulse ofinnovation at Dell. In an environment inwhich the prices of components and theproduct itself were subject to constant deval-uation with the passage of time in anticipa-tion of the next round of technical innova-tion, terms of competition had shiftedfrom the product to the production anddistribution process. In these circumstances,the logistics of time compression betweenadjacent steps from PC production to finalsale emerged as a potentially disruptive busi-ness model in the industry and a pathway ofinnovation that differentiated PC firms likeDell.

By forging a direct path to the customer,Dell was enormously successful in reducingthe time that the product was held in inven-tory between the final assembly and the saleto the user. For most PC firms, this time lagin distributing the product, often measuredin months, had a decisive impact on thecompanies’ performance. Compaq, one ofDell’s competitors during this period, typi-cally had two months of final product ininventory. In contrast, Dell maintained afinal product inventory that was measurable,at most, in days. This reduction in time fromthe assembly of the final product to final

marketing was one of Dell’s most criticalcompetitive assets during these early years.Consequently, what Dell pioneered throughdirect selling was not only a differentmarketing model, but also a new standardin the industry for the relationship betweentime and profit.

When Dell experienced difficultiesmanaging its growth by the early 1990s andbegan to resort to the buildup of an inven-tory of components, it was essentiallyrecasting the relationship between time andthe management of risk in its businesssystem. Arguably, the greatest risk that Dellfaces as a high-volume manufacturer isdisruption in the supply of raw materials. Insuch firms, the inventory of raw materialsis an offset against the risk of disruptionsin supply. If supplies become unavailable,the inventory is able to even out such imbal-ances in supply and demand. This offsetagainst risk, however, incurs costs. Thesecosts, in turn, are related to time. The longermaterials are held in inventory, the greateris the cost incurred as an offset to the riskof disruptions in supplies. By the early1990s, as Dell succumbed to the pressureof having to hold more inventory as an offsetto the risk of such disruptions, it sacrificedwhat was arguably its greatest competitiveasset. The firm effectively lost its ability tominimize the buildup of time embedded inits system of balancing demand for finishedPCs (orders) and supplies of components tobuild them (inventory). For Dell, riskmanagement through the buildup of inven-tory had overwhelmed the advantages it hadachieved over competitors by controllingtime in its system of balancing supply anddemand.

Broken down into its most basic element,supply-and-demand balancing involves twotypes of time management: informationaland logistical. Informational time manage-ment for balancing the supply and demandof components consists of coordinating themyriad transactions at the core of theprocurement process. Transacting, in turn,is fundamentally a process of informationgathering to justify and initiate a purchaseand sale and the exchange of information


practices in the auto industry is precisely on theissue of proximity to suppliers. Glasmeier andMcClusky (1987) revealed that the auto industrywas able to maintain the integrity of just-in-time supply systems even when suppliers werelocated eight hours’ or two days’ drive fromassembly facilities. As I show later, Dell and itssources of supply operate in much closer prox-imity and on much tighter delivery schedules.


with the other party to consummate theterms of the transaction. As economic activ-ities, information gathering and informationexchange are processes that involve expen-ditures of time that incur costs based on thetime that is expended.

Owing to the interfirm structure of theindustry, procurement is a transaction-inten-sive set of activities between assemblers andsuppliers in which time plays a critical role.In its role as assembler, Dell is positionedat the center of two sets of transactionsunderlying the procurement process: (1) thetransactions from customers ordering PCsand (2) the transactions with suppliers toacquire the components for such orders. Asa buyer of components from suppliers andthe source for orders from customers, Delloversees an information- and communica-tion-intensive enterprise for coordinating theexchanges that are needed for the procure-ment of parts. In an effort to innovate thisprocess and balance the demand for andsupply of components in close to real timewithout holding components in inventory,the PC maker has sought to develop routinesfor compressing two time horizons. First, indeploying the Internet as the communica-tion infrastructure for its interactions withcustomers and suppliers, Dell aimed toreduce the average time embedded in thetransactions for order intake and procure-ment. Second, as the firm reduced the timeincrements embedded in its transactionalprocesses with customers and suppliers, itsought ways to compress the time lag sepa-rating these two sets of transactional activi-ties. In practical terms, balancing the supplyand demand of materials without inventoryin a high-volume environment compelledDell to attack the increments of time thatwere expended in initiating and consum-mating the two sets of exchanges whilealigning the two sets of transactions closertogether in time. To accomplish this aim,Dell not only had to learn how to use theInternet to develop new transactionalroutines, but also had to build an organiza-

tion with the capabilities of implementingthese informational time economies.11

In contrast to the time expended inorganizing information flows for transactions,logistical time management involves therecalibration of the time expended inexecuting the movement of materials fromsupplier factories to Dell’s assembly sites.The objective is to reduce these timeincrements so that materials do not sit in anyone place as inventory as they circulatebetween firms and locations. Although thistask is also dependent on the gatheringand exchange of information between sepa-


11 Even Marx, despite his emphasis on profitderiving from labor expended in production,made trenchant observations on the role oftime and transaction costs in profit making. InVolume 2 of Capital, Marx wrote how the circuitof capital, whereby money is turned intocommodities and commodities into money,involves not only production, but also acts ofpurchase and sale in which the capitalist func-tions as both the buyer and the seller. The timeduring which these transactions take place andduring which products move from one step to thenext constitute what Marx acknowledged as“genuine costs of circulation” for the capitalist.“Just as the time of circulation of capital is a neces-sary segment of its reproduction time, so too isthe time in which the capitalist buys and sells andscours the market also necessary in which he func-tions as a capitalist.|.|.|. It is part of his businesshours” (Marx 1974 [1885], 132). In this way, Marxregarded the circulation of commodities—theirphysical movement in changing location as theytransition from raw materials to finished prod-ucts—and the time expended in circulation asintegral to the process of adding value. For Dell,compressions in the time increments embeddedin the transactions that are central to the procure-ment process enable the firm to shorten the cycleswhereby capital is turned into finished commodi-ties and back into money with an increment ofprofit, thereby accelerating the number ofcircuits—turnovers—that the capital of the firmis able to make in a fixed period. Again, Marx wasclear that turnover time is equal to productiontime plus time of circulation. With Dell, time andturnover are part of the same circulation-basedbusiness system.


rate firms, it is the process of movingmaterials from one step in the process toanother and one location to another—andthe time it takes to execute such move-ment—that differentiates it from the firsttype of time management.

Although informational and logistical timemanagement differ, they share the sameobjective. Both seek to reduce costs bydiminishing the time increments that areembedded in adjacent steps linking procure-ment to production and to the final sale.From an operational standpoint, theoutcome from this recalibration of time issimilar in both instances. Informational andlogistical time management aim to increasethe number of times that firms are able toturn inventory—turnover cycles—during afixed period. From a financial standpoint,this recalibration of time enables money tobe turned into commodities and back intomoney with an increment of profit—at afaster pace. Such acceleration in the turnoverof capital and commodities is a basic objec-tive of firms in seeking profit (Harvey1996, 241).

In the Internet, Dell uncovered a pathto two critical innovations for recalibratingtime and rebalancing the supply and demandof materials in its demand-pull businesssystem, which it refers to as global supplyplanning and demand fulfillment. Globalsupply planning involves the exchange ofinformation among Dell, its customers,and its suppliers for the consummation oftransactions in the procurement process.Demand fulfillment deals with the logisticsof executing the movement of materials fromprocurement to assembly and to final

marketing. These innovations represent thebreakthroughs by Dell to control time inits system of material balancing bydecreasing the average levels of inventory,measurable most visibly since 1994 when thefirm conceded that its procurement prac-tices were among the worst in the PCindustry and committed itself to a programof change (see Table 2).

Innovation: Global Supply Planning

Global supply planning represents aneffort by Dell to recalibrate the time hori-zons in one of its most time-driven activi-ties, the process of procurement with itscomponent suppliers. What Dell seeks fromthis innovation is a reduction in the time thatis needed for communicating andexchanging information with suppliers aspart of the transacting process to securecomponents and verify the terms of theexchange. Global supply planning uses bothtechnology, in the form of the Internet, andorganizational change, in the form of adifferent set of relationships between Delland its supply-chain partners, to accomplishthis aim. The rationale for global supply plan-ning is derived from Dell’s recognition thatprocurement is basically a transactingprocess that is driven fundamentally byexpenditures of time in the collection ofinformation and the exchange of commu-nication.

At the core of global supply planning is aset of routines for modulating procure-ment flows by balancing four differenttime horizons in the procurement process:(1) 1-year periods for which Dell generates


Table 2

Days’ Supply of Inventory at Dell

1994 1995 1996 1997 1998 1999 2000 2001 2002

Number of days’ supply ofinventory at Dell Computer 32 21 16 13 8 6 5 4 3

Sources: Dell Computer Corporation, 10-K reports (1996, 1999, 2000); Cihra (1998); Edwards and Park (2002);and an interview with Stephen Cook, Dell senior process engineering manager, 25 April 2002.


a Master Demand Plan for componentsbased on historical data and informationfrom its largest customers; (2) the 2- to 3-month lead times required by suppliers tosecure their own supplies and fabricatecomponents; (3) the 7- to 30-day periodfor transporting supplies to Dell’s facto-ries; and (4) the just-in-time requirementsof Dell to postpone the delivery of compo-nents until orders are actually received,enabling the PC maker to “pull” componentsinto its build cycle supposedly without inven-tory (Kapuscinski et al. 2004). The innova-tive character of the process is derived fromthe use of the Internet to automate andcontrol the exchange of information betweenDell and suppliers to accomplish Dell’saim of balancing the demand for and supplyof components across and within thesefour time frames.

Operationally, suppliers commit to thematerial requirements of the one-year planand coordinate their own two- to three-month production schedules, along withtransport time, with the plan’s general para-meters. As orders are received andprocessed, however, Dell alerts parts vendorscontinuously through Internet-basedcommunication of changes—exceptions—inthese requirements for materials. When anexception exists and a change in the flow ofmaterials is required, Dell triggers a requi-sition and automatically sends it through theweb to the supplier. The supplier thencommits to the order or, depending uponthe circumstance, to a postponement inthe flow of supplies. By late 2001, 90 percentof Dell’s purchases of components wereoccurring through these Internet-basedinteractions with parts vendors, effectivelyrecasting the informational routines in Dell’sprocurement process (interview with a vicepresident, 20 June 2001).

Yet, in remaking these routines, Dellconfronted an organizational barrier that ithad to overcome and transform. For globalsupply planning to function, suppliers hadto operate on the same systems of informa-tion sharing and exchange as did Dell. Thistechnical imperative forced Dell to shift thenature of its relationship to its suppliers—

as well as the number of these suppliers—and restructure the form of business orga-nization at the foundation of its procurementand production network. To supply Dell,parts vendors had to develop the same infor-mation-sharing capabilities and the samecommunications platform as Dell. Thisrequirement compelled suppliers to makeinvestments in their own information andcommunications systems that were compat-ible with Dell’s system (developed jointly byDell and the supply-chain software firm ofi2) as a precondition to supplying Dell.12

While it is not uncommon for vendors toconform to often-onerous requirements thatare imposed upon them by their customers,Dell’s Internet-based procurement routinesimposed a new type of organizational imper-ative on the supply base. Not only did partsvendors have to upgrade their informationsystems and accept the responsibilities formodulating the delivery of supplies in accor-dance with the material-balancingconstraints of the global supply planningsystem, but in accepting these responsibili-ties, they had to become more organiza-tionally integrated with the procurementplanning routines that Dell was recasting.These routines, in turn, were operationalonly through a more integrated andcontrolled organizational relationshipbetween Dell and the supply base. Dell, ineffect, assumed the task of remaking itsprocurement routines by imposing a specificset of technical requirements on its suppliersand spreading a new set of capabilitiesamong separate firms. In the process, itimposed a form of organization upon itssupply base that resembled many of theattributes of integrated firms. Organizational


12 In reference to these requirements onsuppliers, CEO Michael Dell was unambiguousin describing the market power of his companyand its effects. “Dell in the U.S. is 50 percentlarger than its nearest competitor and growingfour times as fast,” he said. “Suppliers have achoice: Supply Dell, or lose market share. Let’sface reality. If my largest customer had a newrequirement, I’d listen to them” (quoted inPerman 2001).


change was thus the route to technologicalinnovation.

As it forced suppliers to become morefunctionally integrated with its global supplyplanning routines, Dell was compelled toreduce the number of its primary compo-nent vendors from more than 200 in 1994,to roughly 35 by 2002. The reason for thiscontraction stems from the fact that, indeveloping the capabilities to implement theroutines of global supply planning andspreading these capabilities to other firms,Dell created an organizationally specificasset. This organizationally embedded asset,in turn, is not easily duplicated by switchingto alternative suppliers through marketmechanisms. Indeed, any such shift wouldforce Dell to incur costs and bear risk.Consequently, far from using the Internetto expand the number of its suppliers in amarket-oriented bid and auction system,Dell is using the Internet and its routinesfor global supply planning to establish anintegrated and proprietary system ofcontrolled interactions with suppliers. ThePC maker refers to the form of organizationcreated by this interplay of technical effi-ciency and organizational necessity as virtualintegration.

Virtual integration combines the organi-zational structure of interfirm networks andauthority relations that are typically ascribedto vertically integrated companies. Suchorganization takes advantage of capabilitieslying outside the boundaries of Dell. Atthe same time, in assessing how this inter-firm organization actually functions, it is crit-ical to recognize Dell’s capacity to exertcontrol over the other actors in this network.Dell forces suppliers to conform to specificoperational and organizational imperativesto implement routines that it has developedto become more competitive. In effect,virtual integration relies on administeredrelationships involving power and control,rather than on market relationships, tospread capabilities among the different firmsin the network. This model of virtual inte-gration, emerging from the control imper-atives of global supply planning, is the orga-nizational precondition for the second of the

two Internet-driven innovations, demandfulfillment.

Innovation: Demand Fulfillment

Demand fulfillment, the second of thetwo Internet-driven innovations at Dell,consists of a process for transferring compo-nents from supplier factories into assemblysites for configuration to finished computersystems. Whereas global supply planningcreates the informational routines for deter-mining the requirements of components tobe pulled into a system of just-in-timeproduction and the transactional routinesfor consummating the exchanges withsuppliers to secure access to these compo-nents, demand fulfillment creates thelogistical routines for executing the move-ment of parts from one stage of the valuechain to the next and from one location toanother. Similar to global supply planning,demand fulfillment uses ongoing Internet-based communication between Dell andits suppliers but incorporates a crucialthird actor in the process—third-party logis-tics providers (3PLs). These 3PLs functionas intermediaries that are responsible forcoordinating the transfer of parts fromsuppliers to Dell and for managing thestorage and staging of these parts in supplylogistics centers (SLCs) so that componentscan be pulled into Dell’s assembly processon a just-in-time basis.

In contrast to the sales channel throughwhich Dell bypassed intermediaries to reachthe customer directly, Dell relies on theintermediary of the 3PL in the procurementchannel to perform a critical step in movingparts from locations of supply to locations ofassembly. These 3PLs and the SLCs inwhich they operate are part of a recentphenomenon in the organization andmanagement of supply-chain systems knownas vendor-managed inventory (VMI) inwhich Dell is a pioneer. The aim of VMI isto shift the costs and responsibilities forsupplying components on a just-in-time basisto supply firms. The 3PLs have found a rolein these VMI systems as managers of theSLCs, where parts are staged for transport



on the final leg of a trip to the factory floorfor final assembly. It is suppliers that nego-tiate contracts for the operation and manage-ment of SLCs directly with 3PLs and paywhat is called “pallet in/out charges” to 3PLsfor storing the inventory of components(interview with Stephen Cook, seniorprocess engineering manager, 12 June 2001).While Dell monitors the ability of 3PL-managed SLCs to provide the requiredparts-transfer services to Dell factories, theoperation of SLCs is conducted indepen-dently of Dell (interview with Gregory Kelly,senior manager of materials and logistics,Dell Nashville, 4 May 2001). This organi-zational arrangement is a way for Dell tocontrol aspects of the supply chain withoutassuming the formal ownership responsibil-ities or costs of these activities.13

While the largely planning-orientedroutines of global supply planning and thelogistical routines of demand fulfillment havedifferent functions, both innovations sharea fundamental objective—compressing timein the processes that go into producing andselling a product. As the logistical elementin Dell’s system of just-in-time production,demand fulfillment is actually a postpone-ment system. Its aim is to delay the deliveryof parts to Dell’s factory sites until partsare ready to be used in the assembly processso as to preclude the buildup of an inven-tory of components. Operationally, demandfulfillment establishes routines wherebyparts that are produced in the factories ofsuppliers are delivered to Dell’s assemblysites at two-hour intervals and all “pulled”into assembly, where they are configuredinto finished products. Customers’ ordersthat are cleared for assembly and processedin two-hour blocks are what provide theinformational input that initiates these mate-rial pulls. Only components that are neededto fill orders for the two-hour cycle inquestion are delivered to Dell’s receivingdock, where they are unloaded and allocated

to various workstations where Dell assemblyworkers configure a finished computer froma kit of parts every two to three minutes.Dell considers demand fulfillment, withthese two-hour material pulls triggeredthrough ongoing Internet informationexchange, to be the most innovative capa-bility in its system of material balancing andjust-in-time production (interview withLance St. Clair, director of supply chain andmaterials management systems, Dell, 10January 2002).

Although technology in the form ofInternet communication provides a criticalplatform for enabling components to circu-late from suppliers to Dell in two-hour inter-vals, the process of demand fulfillment relieson a necessary organizational component.Virtual integration among Dell, its compo-nent vendors, and 3PLs is the organizationalprecondition for the two-hour material pullsat the core of this innovation. For Dell, suchtime-driven levels of coordination are notpossible through market relationships amongfirms. In this sense, demand fulfillment,much like global supply planning, is depen-dent on a fundamental change in businessorganization in which the boundariesbetween firms remain formally separated,but the functional relationship among theseformally separate companies is one of inte-gration.

What differentiates demand fulfillmentas an innovation is the decisive role of“spatial fix” used by Dell for the two-hourmaterial pulls. The PC maker needs thistransformation in the territorial arrangementof economic activity to exploit its owninnovative economies of time and speed.Dell, in effect, has been able to remake itstime-driven system of balancing materialson the basis of a spatial strategy. The resultis a geography of innovation—the creationof a new industrial space—fusing impera-tives of time, organization, and territory. Thisgeography is critical for understandingDell as an innovative logistics firm, remakingterritory and redefining the meaning ofglobal production systems.


13 Third-party logistics providers that manageSLCs that are involved in the delivery of partsto Dell include such firms as BAX, MenloLogistics, Ryder, and Eagle Global Logistics.


Geographies of InnovationTime and distance conspired to pose a

formidable problem for Dell in its effort toimplement the two-hour material pulls ofdemand fulfillment, especially in the after-math of its expansion during the 1990s. In1990, the logistics of assembling PCs forDell, as a firm with one assembly location,consisted of controlling supply flows intoAustin, where the PC maker staged and thenpulled components into its factory. By theend of the decade, Dell was configuringproducts at six assembly sites on fourcontinents.14 In this expansionary configu-ration, the challenge for Dell as a just-in-time producer was how to overcome thegeographic separation between locations ofsupply and locations of assembly while main-taining the continuous flow of two-hourmaterial pulls at each assembly site withoutthe buildup of an inventory of compo-nents. Operationally, what Dell confrontedas a globally expanded firm was a set of logis-tical problems that were oriented funda-mentally around issues of time and space.

To overcome this challenge, Dell devel-oped two critical offsets for shrinking thegeographic separation between locations ofthe fabrication of components and locations

of final assembly. These two spatial fixesreveal the relationship of time to organiza-tion and territory. They also emphasize theimportance of space in the PC maker’soverall strategic vision of logistics as a busi-ness model.

On the one hand, Dell prevailed uponsome of its suppliers to establish componentfactories in the locations where it assemblesPCs. When Dell established its assemblyfacilities in Penang, in 1996, for example, itnegotiated with and convinced two of itsmanufacturers of motherboards, Jabil Circuitand SCI, to locate factories there (interviewwith Simon Wong, general manager, DellAsia Pacific, 2 October 2001). Althoughthese two suppliers undoubtedly benefitfrom this arrangement, the importance ofsuch proximity relationships is even moreparamount for Dell in implementing itsdemand-fulfillment system of two-hourmaterial pulls. Michael Dell was candid inacknowledging the importance of proximityto suppliers as a strategic aim for the PCmaker (Dell and Fredman 1999, 178–79).15Such proximity relationships are not onlyrecasting the linkages that form the basisof local economic geographies. In drivingthe locational decisions of other firms andcrafting proximity relationships with thesecompanies, Dell reveals itself as an agentin the creation of agglomeration economies,reinforcing existing place-based concentra-tions of high-technology activity where it haschosen to locate its own operations. In thisreshaping of the economic landscape, Delladmits to a special role in exercising choicesthat contribute to a broader story of spatialdevelopment focusing on forces of cumu-lative causation and agglomeration (seeFigure 2).


14 Its objective in this expansion was to ease thecapacity burden on the Austin facility as ordersbegan to escalate after 1994 and on supplymarkets throughout the world from designatedregional assembly sites located in each marketzone. Instead of a centralized assembly systemfocused on Austin, by 2001, Dell had created adecentralized assembly system that was spreadacross the globe but concentrated in specificregional locales. Dell’s locational preferencesreveal a pattern. First, the locations Dell chosefor expansion—Limerick (Ireland), Porto Alegre(Brazil), Xiamen (China), and Penang(Malaysia)—represent existing concentrations ofhigh technology, although, with the exceptionof Penang, they may accurately be termed theworld’s “second-tier” high-technology cities(Markusen, Lee, and DiGiovanna 1999). Second,Dell revealed a preference for places that offerit direct incentives. Finally, Dell chose some ofthe same locations that are preferred by itsmost critical supplier of materials, Intel.

15 “We came up with the phrase ‘proximity pays’as a result of translating the ROIC [return oninvested capital] metric down to each componentand each supplier.|.|.|. It was very clear thatsuppliers that located their factories close to ourshelped us to deliver a higher ROIC than thosewho were farther away.” Dell also acknowledgedhow its power of persuasion managed to prevailover these firms (Dell and Fredman 1999, 178).



Fig

ure

2. D

ell p

rodu

ctio

n ne

twor

k, 2

001.

CH

INA Xia

men

Aus

tinN

ashv

ille

TA

IWA

NG

ulf o

f Mex

ico

Lim

eric

k

IRE

LA

ND

Pen

ang

SIN

GA

PO

RE

TH

AIL

AN

D SUM

ATRAM

ALA

YS

IA

Eld

orad

o d

o S

ul

UR

UG

UA

Y

PA

RA

GU

AY

ARGE

NTIN

A

BR

AZ

IL

ME

XIC

O

UN

ITE

DS

TA

TE

S

UN

ITE

DS

TA

TE

S

Japa

n

Sea

ttle

Sili

con

Val

ley

Mex

ico

Kor

ea

Ass

em

bly

Site

Loca

lized

Sup

plie

r C

once

ntra

tion

Glo

bal S

uppl

ier

Con

cent

ratio

n

Su

pp

ly L

og

istic

s C

en

ter


While relocations of supplier factories arecritical in Dell’s geography of innovation, farmore widespread as a business practiceand spatial strategy is the intermediatestep inserted by Dell between the fabrica-tion of parts and the final assembly, craftedaround the SLCs. Pioneered by Dell andincreasingly imitated by its competitors, theSLCs reshape space by creating new rela-tions of proximity between supply andassembly. SLCs are always located nomore than a 20-minute drive from Dell’sassembly sites. Dell requires parts vendorswho do not have factories close to Dell’sassembly sites to maintain at least two weeks’supply of inventory in these SLCs so thatparts are always available to Dell to be pulledon an as-needed basis, regardless of short-term fluctuations in demand. Most signifi-cantly, in this arrangement, Dell forcessuppliers to bear the carrying costs of thisinventory.

In Dell’s time-driven business model, thestorage and staging of parts in SLCs is essen-tially a system for maintaining the rhythmsof just-in-time business practices bycollapsing the distance between the loca-tions of supplies of components, and thelocations of assembly. SLCs play the pivotalrole in bridging this distance. These facili-ties, the costs of which are imposed on thesupply base, create external economies forDell from proximity relationships betweenkey nodes in Dell’s network. They are whatenable Dell to manage the compressed timecycles for pulling material into productionon a just-in-time basis. Perhaps more impor-tant, however, is that the staging and storageof supplies in these SLCs is a response tothe greatest problem confronting Dell—therisk of securing access to supplies of compo-nents within specific parameters of time thatare consistent with its high-speed, build-to-order pull system. What Dell seeksthrough the staging and storing process iscontrol over the risks it encounters insecuring supplies in a just-in-time, envi-ronment.

In seeking to remedy this risk of timelyaccess to supplies of components, Dell reor-ganizes space as a substitute for inventory.

It implements this territorial reorganization,however, organizationally, creating proximitybetween itself and its suppliers as a sourceof speed and certainty. It is through orga-nization that Dell allocates the arrangementof activities in space and the linkagesbetween these activities. It enforces theseconditions of geography and proximity uponits suppliers owing to its power to controlthe behavior of these firms.

In effect, the economies of time compres-sion that Dell pioneered in coordinating thetwo-hour material pulls into each of itsassembly sites play the decisive role in influ-encing how the PC maker allocates anddistributes the linkages in its network acrossand within geographic space. Nevertheless,Dell does not simply remake territory insome mechanical adaptation to the efficiencyimperatives of time compression. Dellinstead relies on the business enterprise thatit created from the innovative process,specifically its power to influence and exertcontrol over other firms, as the organiza-tional precondition for this territorial recon-figuration.

Vendors of components are candid in theway this power relationship operates. “Dellhas a significant amount of power with itssuppliers based on the current and futurebusiness levels they offer,” argued one ofDell’s large parts producers. “They know itand they use it” (interview with Supplier 2,2 July 2002). Another supplier insisted thatthe requirement by Dell to maintain at leasttwo weeks’ supply of inventory in SLCs hasbeen and remains “the single biggest issuefacing Dell’s material suppliers.” Dell’ssuppliers, this vendor explained, have beentrying to negotiate with Dell to reduce thisrequirement to one week. “Dell has resistedfor the past three years,” this supplierexplained, “because they do not want therisk” (interview with Supplier 1, 28 May2004).16 Suppliers that do not want to


16 Dell, this supplier explained, has an explicitmonthly grading system for assessing vendors’performance. If a supplier allows inventory inSLCs to fall below the required two weeks, itreceives a warning from Dell. If the situation


conform to the inventory requirements ofSLCs in the PC maker’s demand fulfillmentsystem will not supply Dell.

In this way, while there is much that isinnovative about the logistics of demandfulfillment, Dell’s claim that it has createdan Internet-driven, just-in-time system ofmaterial balancing without procurementinventory is misleading. Indeed, despite theinnovative advance of demand fulfillment,there is inventory in the just-in-time systemof material balancing at Dell. This inventory,however, along with its costs, lies outside theformal boundaries of the PC maker. It iscomponent vendors that Dell forces to bearthese inventory costs as the price of inclu-sion in Dell’s interfirm business organiza-tion. Similarly, it is the same supplier firmsthat are forced to reallocate their economicactivities territorially in conforming to therequirements of Dell’s own innovativeroutines.17

Dell’s control over suppliers and use offorce to influence the locational behavior ofthese firms in remaking territory reveals acritical organizational feature of the PCmaker’s interfirm network. Dell interactswith the suppliers in its network not throughmarkets, but through highly controlled rela-tionships. Similarly, the relationship betweenDell and the entities managing the SLCs,although nominally between separate firms,is also far from what would qualify as amarket transaction between independent

agents. In Austin, for example, the SLCserving the Morton Topfer ManufacturingCenter, where Dell assembles finishedgoods, is literally on the Dell propertydirectly adjacent to the assembly activity.Moreover, this SLC facility was actuallyconstructed by Dell and then leased to EagleGlobal Logistics, a 3PL that is responsiblefor managing the inventory at this particularSLC. Formally, the relationships amongDell, Eagle, and the suppliers that maintaininventory in the facility are among separatefirms. Functionally, 3PLs, such as Eagle, thatmanage the SLCs and the suppliers thatposition inventory in these warehouses arecompletely integrated into the Dell opera-tion.

Both global supply planning and demandfulfillment, in effect, have compelled Dellto create a functionally integrated organiza-tion with the other firms in its network. Atthe same time, Dell is able to maintain clearorganizational boundaries between itself andthese other firms when such formal demar-cations are advantageous for it, as, forexample, in the postponement in the deliveryof components until the very moment whenDell wants them. Such arrangements repre-sent Dell’s strategy for simultaneously takingadvantage of external capabilities by passingcertain obligations and costs on to otherparties, while retaining necessary controlover critical adjacent steps in the PCcommodity chain. In assessing the time-sensitive nature of these linkages and theirinterfirm character, Dell has concluded thatcontracting with these other firms throughmarkets to execute the necessary steps, fromprocurement through assembly, posesgreater levels of risk than do relationshipsthat are consummated through force.

In this sense, Dell’s reliance on controlledrelationships to link the adjacent operationsin its network is strikingly similar to thedependence of vertically integrated firms inthe late nineteenth century on administra-tive controls to organize their procurementand production systems. In much the sameway that vertical integration represented aresponse to the risk of managing complexprocurement, production, and distribution


occurs more than twice, this vendor said, theoffending supplier will probably be dismissed.

17 Although there is inventory in the Dellmodel, the costs of which are born by suppliers,there are nevertheless aggregate efficiency gainsin the system. In overseeing the supply chainsof its suppliers, Dell forces the suppliers tobecome more proficient in managing their owninventory and supply chains (Kapuscinski et al.2004). As another of Dell’s suppliers conceded,“Dell dragged us over the coals to make certainthat we knew about the risks in our own supplychain” (interview with Supplier 3, 24 July2002). These suppliers, in conforming to theinventory requirements of Dell in the SLCs, haveenormous incentives to cut their own expensesrelated to inventory.


systems without disruption, so, too, doesDell’s virtually integrated enterprise confrontsimilar types of risk by deploying similarcontrol mechanisms. The difference is thatwhereas manufacturers in the early mass-production era tended to exert such controlthrough mechanisms of administrative plan-ning in concert with their ownership ofassets, Dell exerts control through mecha-nisms of administrative planning in combi-nation with assets owned by different firms.Although the asset structures of the twotypes of organization differ—vertically inte-grated firms own the assets in their networkwhile Dell does not—the rejection ofmarket-based interactions to accomplishoperational objectives is fundamentallysimilar.

Thus, the idea that market linkages amongnominally independent firms are emergingas the mechanism of governance in inter-firm production networks is far removedfrom the experience of logistics-orientedproduction networks that are driven by thegoal of time compression and dominated byfirms like Dell. While the proliferation ofthe interfirm network as an organizationalphenomenon is undeniable in the currentperiod, the power relations of Dell’s inter-firm enterprise tell a far more revealing storyabout interfirm networking than does thefocus on Dell’s structural characteristics perse. In the high-risk, high-speed, time-drivenbusiness systems being pioneered by Dell,the story of interfirm cooperation is notbased on markets. Instead, it is a story ofhow power is exercised within networks offirms and how the exercise of such powermobilizes resources within the network forinnovation and profit. Geographic proximityis one pivotal resource that Dell mobilizesthrough organizational power to accommo-date a set of time-sensitive innovativeroutines.

What Dell has established in response tothe problem of time and distance that isposed by the innovative routines of demandfulfillment is a decentralized system ofsupply-and-demand balancing for the allo-cation and transfer of parts at two-hour inter-vals into each different assembly site. At the

same time, the company has standardizedthis system so that the routines it has createdfor the allocation and transfer of componentsare identical in each location. Commentingon this practice, Dell’s vice president andgeneral manager for Latin Americaexplained: “We execute this same businessmodel everywhere. It’s like McDonald’s”(interview with Daryl Robertson, 12 March2002). Although there is undoubtedlysome variation in the way this businessmodel is implemented at the different Delllocations, the geographic impacts of thiseffort at standardization are profound. Increating a standardized set of innovativeroutines for the transfer of parts betweensources of supply and assembly, the PCmaker has contributed to homogenizing aset of territorial practices across space. Thus,by organizing relationships of proximitybetween sources of supply and assembly inits chosen regional locales, Dell has assumedthe role of agent in crafting territorialfeatures of the contemporary regionaleconomic world.

ConclusionWhat makes Dell a compelling and

paradigmatic case is that it is part of abroader trend among firms to capitalize onnew communications technology and usecorporate power to rationalize supply-chainsystems and forge innovative businessmodels on the basis of distribution and thelogistics of time compression. In this regard,Dell is a close counterpart to Wal-Mart, theone firm that is changing the economic envi-ronment worldwide most decisively. LikeDell, Wal-Mart is essentially a logisticsfirm that operates an enormously efficientsupply-chain and inventory-managementsystem through a combination of highlyadvanced communications and a system ofthrottling its supply base. As a dominantpurchaser, Wal-Mart, like Dell, is able toimpose technological imperatives on itssuppliers—in 2004, it forced its 120 largestsuppl

innovation, time, and territory: space and the business … · 2020. 12. 11. · space and the...

Documents