Levina & Ross/The Vendor’s Value Proposition in IT Outsourcingpeople.stern.nyu.edu/nlevina/Papers/From the Vendors Perspective.pdf · value proposition. While the client’s sourcing

Levina & Ross/The Vendor’s Value Proposition in IT Outsourcing

MIS Quarterly Vol. 27 No. 3, pp. 331-364/September 2003 331

RESEARCH ARTICLE

FROM THE VENDOR’S PERSPECTIVE:EXPLORING THE VALUE PROPOSITIONIN INFORMATION TECHNOLOGYOUTSOURCING1, 2

By: Natalia LevinaInformation Systems Group/IOMSStern School of BusinessNew York University44 West Fourth Street, Suite 8-78 (KMC)New York, NY [email protected]

Jeanne W. RossCenter for Information Systems ResearchMassachusetts Institute of Technology3 Cambridge CenterCambridge, MA [email protected]

1V. Sambamurthy was the accepting senior editor for thispaper.

2An earlier version of this work was presented at the IFIPWorking Group 8.2 Research Workshop in Charlotte,NC, December 12, 1999. The presentation was basedon the working paper “Sources of Vendor ProductionCost Advantages in IT Outsourcing,” Center for Informa-tion Systems Research Working Paper #309, Massa-chusetts Institute of Technology, October 1999.

Abstract

To date, most research on information technology(IT) outsourcing concludes that firms decide tooutsource IT services because they believe thatoutside vendors possess production cost advan-tages. Yet it is not clear whether vendors can pro-vide production cost advantages, particularly tolarge firms who may be able to replicate vendors’production cost advantages in-house. Mixed out-sourcing success in the past decade calls for acloser examination of the IT outsourcing vendor’svalue proposition. While the client’s sourcingdecisions and the client-vendor relationship havebeen examined in IT outsourcing literature, thevendor’s perspective has hardly been explored. Inthis paper, we conduct a close examination ofvendor strategy and practices in one long-termsuccessful applications management outsourcingengagement. Our analysis indicates that thevendor’s efficiency was based on the economicbenefits derived from the ability to develop acomplementary set of core competencies. Thisability, in turn, was based on the centralization ofdecision rights from a variety and multitude of ITprojects controlled by the vendor. The vendorwas enticed to share the value with the clientthrough formal and informal relationship manage-ment structures. We use the economic concept of


332 MIS Quarterly Vol. 27 No. 3/September 2003

complementarity in organizational design, alongwith prior findings from studies of client-vendorrelationships, to explain the IT vendors’ valueproposition. We further explain how vendors canoffer benefits that cannot be readily replicatedinternally by client firms.

Keywords: Outsourcing of IS, case study, com-plementarity in organizational design, IS corecompetencies, management of computing and IS,systems maintenance, IS staffing issues, ISproject management

ISRL Categories: EL07, AI0102, FB08, BA,AF04, AF0403, USE, AF0412, EE

Introduction

Outsourcing is a phenomenon in which a userorganization (client) transfers property or decisionrights over information technology (IT) infrastruc-ture to an external (vendor) organization (Loh andVenkatraman 1992b). The brief history of IT out-sourcing includes episodes of both high hopesand bitter disappointment. Since Eastman Kodak’slandmark outsourcing of its IT services (Applegateand Montealegre 1991), the outsourcing industryhas been growing at a staggering rate of about 20percent a year (Caldwell and McGee 1997). Worldwide spending on IT outsourcing servicesreached almost $64 billion in 2001; in 2000, IToutsourcing represented about 30 percent of ITbudgets (Mason 2000). Despite these numbers,both vendors and their clients are struggling tounderstand the outsourcing value proposition:can vendors deliver economic and managementbenefits to their clients that outweigh contractingcosts and risks?

A number of studies indicate that the leadingreason behind outsourcing is the need to reduceand control IT operating costs (Ang and Cum-mings 1997; Ang and Straub 1998; Casale 2001;Loh and Venkatraman 1992a, 1992b; Slaughterand Ang 1996), followed by the need to improvemanagement focus and access technical talentnot available in-house (Casale 2001; Lacity and

Willcocks 1998). The intended benefits, however,often have not materialized (Hirschheim andLacity 2000; Scheier 1997) and risks are signi-ficant (Aubert et al. 1998, 1999; Earl 1996). Forexample, one study found that only 54 percent ofthe agreements realized expected cost savings(Lacity and Willcocks 1998). More recently, aGartner Dataquest Report claimed that about oneof every three outsourcing contracts targeting costreductions failed to match expectations (Caldwell2002a, 2002b). Moreover, there is evidence thatcompanies are willing to undergo the expense ofcanceling their contracts and rebuilding their in-house IS capabilities (Buxbaum 2002; McDougall2002). The mixed success of existing agreementshas not, however, led to disillusionment with theconcept of outsourcing.

Growth in the outsourcing market signals thatfirms of all sizes believe that IT vendors willultimately deliver value (Casale 2001). In fact,outsourcing results have been improving as thepractice of outsourcing has matured (Willcocksand Lacity 2000). Variations in outsourcing out-comes call for an investigation of factors thatshape the value delivered to clients through out-sourcing. The objective of the research presentedin this paper was to explore this question from thevendor’s perspective. In particular, we examinehow vendors create value in the case of appli-cation management3 outsourcing.

The case study presented here contributes to therecent stream of research that uses qualitativedata to examine the history of ongoing out-sourcing relationships (Kern 1997; Koh et al.1999; Lacity and Willcocks 1998; Lacity et al.1995; Sabherwal 1999; Saunders et al. 1997). Toexplain how vendors provide value, this paperdiverts from the more usual transaction costeconomics (TCE) (Williamson 1979), institutionaltheory (DiMaggio and Powell 1983), and neo-classical economics (reviewed in Williamson1985) accounts of IT outsourcing. Instead, it usesthe concept of complementarity in organizational

3Application management refers to ongoing main-tenance‚ support‚ and enhancement activities of all orpart of a firm’s application portfolio.


MIS Quarterly Vol. 27 No. 3/September 2003 333

design (Milgrom and Roberts 1995), the corecompetency argument on outsourcing (Hamel andPrahalad 1996; Quinn 1999), and findings fromthe literature on the vendor-client relationship(Elitzur and Wensley 1997; Kern and Willcocks2001; Willcocks and Lacity 2000). In this way, theanalysis extends neoclassical economics ac-counts of outsourcing to suggest why many largefirms may choose to outsource applicationsmanagement.

The paper is organized as follows. First, wereview existing relevant research on IT out-sourcing. We then explain our choice of method-ology for collecting and analyzing the data. In thefollowing section, we present an overview of thecase. Then we analyze and interpret data fromthe case using the theory of complementarity inorganizational design, and from the standpoint ofclient-vendor relationship factors, to build a frame-work for understanding the value proposition for IToutsourcers. Finally, we discuss the contributionsthis paper makes to research and practice.

Background: Perspectiveson Outsourcing

Historically, research on IT outsourcing hasfocused on the sourcing decision itself, trying tounderstand why organizations outsource. Drawingon economic literature, such as TCE (Williamson1979) and the theory of incomplete contracts (Hart1989), the sourcing decision is often seen as arational decision made by firms that haveconsidered transaction-related factors such asasset specificity, environmental uncertainty, andother types of transaction costs (Ang and Beath1993; Ang and Cummings 1997; Ang and Straub1998; Nam et al. 1996; Nelson et al. 1996;Richmond and Seidmann 1993; Richmond et al.1992; Walker and Weber 1984). An alternativetheory, neoclassical economics (reviewed inWilliamson 1985), posits that firms outsource IT toattain cost advantages from assumed economiesof scale and scope possessed by vendors (Angand Straub 1998; Loh and Venkatraman 1992a;

Slaughter and Ang 1996). This theory hasattained more empirical support in studies of out-sourcing decisions than TCE (Ang and Cummings1997; Ang and Straub 1998; Casale 2000, 2001;Loh and Venkatraman 1992a, 1992b; Slaughterand Ang 1996; Walker and Weber 1984).However, the economies of scale and scope argu-ment would predict that outsourcing has little tooffer larger firms, because they can generateeconomies of scale and scope internally by repro-ducing the production methods used by vendors.The data on outsourcing, however, indicates thatmany large firms continue to pursue outsourcingarrangements (Chabrow 2002; McDougall 2002).

Because of the focus on the sourcing decision offirms on average, this set of economic theories isgenerally not used for explaining outcomes of theoutsourcing decision. Alternative theories used tostudy IT sourcing decisions suggest that sourcingdecisions are motivated by political (Lacity andHirschheim 1993) or institutional (Ang and Cum-mings 1997; Hu et al. 1997; Loh and Venkatraman1992b) factors, but again these theories do nothelp in explaining variability among outsourcingoutcomes.

IS research on the value generation potential ofan outsourcing relationship has considered threefactors: client characteristics, vendor character-istics, and the vendor-client relationship (Goles2001). A key client characteristic is an under-standing of how to manage resources that a firmdoes not own (Elitzur and Wensley 1997;Henderson and Venkatraman 1990; Kern 1997;Lacity and Willcocks 1998; Lacity et al. 1995;Sabherwal 1999; Saunders et al. 1997; Useemand Harder 2000). This involves, for example,retaining in-house capabilities to ensure that ITresources are adequate and appropriately distri-buted to meet organizational requirements. It alsoencompasses vendor selection, relationship man-agement, managerial competence, architectureplanning, and monitoring emerging technologies(Currie 1998; Goles 2001; Lacity et al. 1995;McFarlan and Nolan 1995; Quinn 1999). Althoughthere are few in-depth investigations of how firmsdevelop these characteristics, their importance iswidely accepted.



The major thrust of the literature on IT outsourcingoutcomes investigates various aspects of thevendor-client relationship. From this literature, welearn that informal (interpersonal trust) and formal(contractual) aspects of the relationship areequally important (Poppo 2002; Sabherwal 1999)and need to be developed (Kern and Willcocks2001; Willcocks and Kern 1998; Willcocks andLacity 2000). Integrative work on this topic byWillcocks and Kern (2001) suggests that strategicintent as well as technical capability shape bothcontract structure and interpersonal relationshipdevelopment. For example, a relationship thataims to tap into the technical leadership capa-bilities of a vendor to achieve IT efficiency maygenerate higher value if it is run as a partnership,whereas one that aims to achieve IT efficiency bytapping into a vendor’s widely available resourcepool may be better managed as a technical supplypay-per-service relationship. A recent survey-based study of outsourcing proposed that highervendor-client alignment, teamwork, balance ofcontrol, and process agility in the relationship willlead to more successful outcomes (Goles 2001).From the standpoint of building effective relation-ships, we know that certain contractual methodsare more conducive to value sharing. Forexample, empirical studies show strong supportfor carrot and stick incentives, shorter term con-tracts, and engagement of multiple vendors(Currie 1998; Lacity and Willcocks 1998). Gametheoretic economics models advocate pilot pro-jects (Snir and Hitt 2002), penalties and multi-phased contracts (Whang 1992), proper choicebetween time and material versus fixed-pricecontracts (McDonnell and Lichtenstein 2002), andrisk sharing and reputation building contractingmechanisms (Elitzur and Wensley 1997).

The third factor shaping the outsourcing valueproposition is the vendor’s own capabilities (Goles2001; Saunders et al. 1997; Willcocks and Lacity2000). Despite growing interest in this factor,there has not been an in-depth examination ofthese capabilities and how they generate value inoutsourced relationships. Through theoreticalhypothesizing, Goles proposed that the vendormust possess such capabilities as technical com-petence, understanding of the customer’s busi-

ness, and relationship management. However,we are not aware of empirical investigations ofvendors’ competencies, which actually createvalue, and the ability or inability of client firms toinstead generate that value internally. This gap inthe literature limits understanding of outsourcingoutcomes and how value is generated and trans-ferred from vendor to client. By analyzing what avendor does on a successful outsourcing contract,we can start to explain when and why firms findvalue in outsourcing.

Methodology

We chose a case study methodology for ourinvestigation of the research question. The casestudy method is preferred “when ‘how’ or ‘why’questions are being posed, when the investigatorhas little control over events, and when the focusis on a contemporary phenomenon within somereal-life context” (Yin 1984, p. 16). The investiga-tion of how vendors deliver value in outsourcingsatisfied all of these criteria. Specifically, we con-ducted an explanatory case study (Yin 1984, p.16) with the goal of posing competing explana-tions and developing new ones. The case studymethod is well established in IS research, espe-cially when it is used for “sticky, practice-basedproblems” such as the value delivered by ISservices vendors (Benbasat et al. 1987).

Following Eisenhardt (1989), we used the casestudy to build theory in a grounded and inductivefashion. We drew on a grounded theory approach(Glaser and Strauss 1967) similar to the way it wasused by Orlikowski (1993) to develop theory fromqualitative data. Grounded theory is a way ofiteratively collecting and analyzing data in order tobuild first a substantive theory of a particular phe-nomenon and than a formal theory on its basis (Dey1999; Glaser and Strauss 1967; Myers 1997). Akey difference between our use of the groundedtheory and that of Orlikowski was that we found aformal, positivistic theory (i.e., concepts of corecompetence and complementarity in organizationaldesign) that explained some of our findings,whereas Orlikowski’s account was interpretive.



Eisenhardt (1989) outlines the steps necessary forusing a case study to build theory grounded indata. After identifying our research question, wereviewed the literature on IS outsourcing anddetermined that no current theory answered theresearch question. Thus, we started our work withno theory under consideration and no hypothesisto test—an ideal for this kind of research (Eisen-hardt 1989, p. 536). We then proceeded to selectour case based on the concept of theoreticalsampling so that we could best answer the ques-tion posed (Glaser and Strauss 1967).

Site Selection

Since our goal was to understand how vendorsdeliver value to clients, we needed a case where(1) the vendor provided extensive access toindividuals at multiple levels who could describemanagement practices and how they delivervalue, (2) the client acknowledged receiving valuefrom outsourcing, (3) the client was willing toshare perceptions as to how the vendor deliversvalue, and (4) the contract had been active longenough to demonstrate long-term outcomes. Thecase we studied satisfied all of these criteria. Oursite presented a rare opportunity for broad accessto a successful outsourcing engagement. Thiscase was revelatory (Yin 1984, p. 48) or exemplarin the sense that we had an opportunity to studysomething previously not researched, but notunique. The research method did not requiremultiple sites. However, replicating our study tocontrast or compare it with studies of less suc-cessful outsourcing cases would provide furtherinsights.

Data Collection and Analysis

The study employed qualitative methods to under-stand the socially rich nature of the vendor’smanagement practices and of the vendor’s value,as they were perceived by the client. Data collec-tion took place in the spring of 1998. It involved avariety of techniques including unstructured andsemi-structured interviews, documentation, archi-

val records, direct observations, publishedsources, physical artifacts such as manuals,forms, and project archives, and follow-up e-mailand telephone interviews (Yin 1984). Semi-structured interviews lasted from 40 minutes to 2hours. Table 1 shows the types of intervieweesand number of interviews.4 One author alsoobserved several project meetings and trainingsessions on-site.

In addition to ongoing field notes, where theinvestigator tried to record what was going onwithout specific focus (Eisenhardt 1989), moretargeted interviews and document collectionfocused on six dimensions of inquiry described inTable 2.5 Each dimension was aimed at furtheringour understanding of the main unit of analysis: apractice by which the vendor provided inimitablevalue to the client. Our data collection was itera-tive: as data was collected, major themes wereidentified to guide further data collection, whichthen modified or built on prior themes andconcepts (Glaser and Strauss 1967)

As suggested by Pettigrew (1990) and Eisenhardt(1989), we broke down data analysis into over-lapping phases resulting in three different types ofcase write-ups. (Table 3 describes the analyticalprocesses associated with each level of output.)We started with a broad definition of the problem,which was sharpened through analysis of relevantliterature, on-site data collection, and discussionswith academic colleagues. This was followed byan open-ended and generative discovery of mainthemes, patterns, and propositions from interviewtranscripts and case notes (Glaser and Strauss1967) which led to additional data collection. Theinitial analysis resulted in a case write-up, or what

4Interview questions varied by type of participant.Interview protocols are available from the authors uponrequest.

5We identified the initial dimensions for the inquiry inanother research setting. This was based on an explora-tory study that took place between November, 1997, andFebruary, 1998, and involved a dozen interviews withproject managers from different firms. The subsequentfocus of the data collection effort was guided bycategories that emerged from data in the case.



Table 1. Interviewee Type and Number of Interviewsa

Interviewee Type Number of Interviews

Vendor OrganizationExecutive Leadership 3

Branch Management 4

Engagement Management 2

Project Management 4

Consultants 5

Client OrganizationIS Department Management 1

Engagement Management 1

Project Management 1

User Managers 4

User Staff 3

Total 28aMost of the data collection was conducted by the first author, with the second author conducting three of the inter-views with the executive leaders in the vendor organization.

Table 2. Dimensions Guiding the Inquiry

Dimension Description

Engagement History Reasons behind the outsourcing decision. Client’s prior experi-ences with outsourcing. Vendor selection process. Size, scope,structure, and length of contracts.

Client’s ManagementPhilosophy

Degree of control over operations. Approach to building the rela-tionship with the vendor.

Vendor’s ManagementPhilosophy and Processes

Degree of control over operations. Utilization of a software engi-neering methodology. Approach to building the relationship withthe client.

Knowledge Sharing Practices High-level goal setting. Everyday communication amonginvolved parties. Technology used to support sharing. Colla-borative culture. Incentives for sharing.

Staffing Decisions andChallenges

Hiring practices. Training. Assignment rules. Promotion rules. Employee satisfaction. Turnover.

Engagement Success Critical success factors. Perceived rating of success so far. Areas of concern.



Table 3. Data Analysis Phases

Case Write-Up Outputs Application to the Case Responsibilities

Level 1 (AnalyticalChronology)

Long detailed description of the history of theclient’s outsourcing decision, client’sorganizational context, various vendors’engagement practices and strategies, andthe engagement outcomes.

The write-up was largely createdby the first author. The secondauthor reviewed and edited thewrite-up on the basis of interviewsconducted by the second author(executive leadership of vendor).

Level 2 (DiagnosticCase)

• Focus on the theme of “How do vendorsdeliver value?”

• Broke the case down into key dimensions: (1) Environmental factors(2) Contractual structures(3) Vendor’s practices (staffing, methodo-

logy, ownership of process and product,collaborative culture)

(4) Engagement challenges• Two separate case write-ups were

prepared, one for vendor and one for client,to assure that the confidentiality of bothparties was not violated. Participants thenprovided feedback on the analysis.

Authors worked together onidentifying broad themes in thecase and the practical implica-tions of the case to be presentedto participants. The diagnosticcase went through four draftsbefore being sent to participants. Incorporating participant feedbackresulted in the final draft of thediagnostic case.

Level 3 (Interpretive/TheoreticalCase)

• In-depth inductive content analysis of thedata from multiple sources.

• Preparation of analytical displays, forexample, checklist matrices and causaldiagrams (Miles and Huberman 1984) todiscover relations among subcategories.

• Reducing number of emerging themes todevelop theoretical clarity.

• Link to the broader literature on IT out-sourcing, core competencies, resource-based view of the firm, economics oforganization, and applicationsmanagement.

The first author went through thecontent analysis exercise inde-pendently of the second author toproduce the first draft of the inter-pretive case. The second authorchallenged and debated firstauthor’s findings and wrote a dif-ferent version of an interpretivecase based on the chronologicalcase and transcripts. Authorsdebated interpretation of the casefrom different theoretical perspec-tives. The interpretive case wentthrough 40 iterations in 4 yearsuntil consensus was reached.



Pettigrew called analytical chronology (level 1output), incorporating multiple levels of analysis.In the next phase, we focused on current strategicconcerns of the organizations involved in thestudy in order to write a diagnostic case (level 2output per Pettigrew). The diagnostic case waspresented to the two participating organizationsfor feedback. We then incorporated the feedbackinto the case analysis. This iterative processallowed for further development of the analyticalframework.

The final phase was to create an interpretive/theoretical case (level 3 output per Pettigrew).Here we further interpreted the narrative devel-oped in prior phases and linked it to conceptualideas derived from the data and to wider theo-retical debates in the literature. In this phase, werelied on content analysis techniques to developthe analytical abstraction from multiple sources ofdata (Agar 1980). By this time, we had developedmajor conceptual themes. We then used causaldiagrams and checklist matrices (Miles andHuberman 1984) to help us discover relationshipsamong concepts and to do axial coding (Corbinand Strauss 1990). Appendix A shows the resultof one such analysis. We read and rereadinterview transcripts, archival records, and fieldnotes to link empirical evidence to recurringthemes and develop new themes. Thereexamination of the data led to clarification ofconcepts and their properties with the goal ofincorporating as much data as possible into theinductive reasoning process. In this process, wechallenged each other’s interpretations of data.Finally, sets of concepts were linked together in aframework that represented client needs, vendormanagement practices, and market strategies.

In the inductive generation of theory from data, werelied heavily on triangulation of different sourcesof evidence, which is a major strength of the casestudy research methodology (Yin 1984). This ledto stronger concept development in the inductivetheory building (Eisenhardt 1989; Glaser andStrauss 1967). Finally, we compared ourgrounded framework to various theories fromeconomics of organization and strategic literature,treating them as another data set and doing com-

parative analysis as suggested by Glaser andStrauss (1967).

The analysis involved consideration of variouseconomic perspectives on firm boundaries,theories on incentives, promotions, distribution ofdecision rights, and network organizations. Weconcluded that the concept of complementarity inorganizational design (Milgrom and Roberts1995),6 along with the core competency concept(Hamel and Prahalad 1996), provided the greatestconceptual insights in analyzing how a vendorgenerates value, while existing work on client-vendor relationships provided insights into why avendor shares this value with the client. Table 4provides an example of how the analytical theme,vendor’s control over a variety and multitude ofprojects fits vendor’s competencies and viceversa, was developed from data.

Data Sources

The Vendor

The vendor, ABC,7 is a medium-sized, decades-old firm, providing large organizations with avariety of IT outsourcing services through a NorthAmerican network of branch offices. In thedecade preceding 1998, the company had demon-strated financial success through an averagecompound total return over 50 percent a year—much higher than the software industry average of26.9 percent. Revenues for 1997 were up 40percent from 1996, and net income for 1997 wasup 81 percent from a year earlier.

At the time of the study, ABC’s top managementconsidered its Metropolis branch, the focus of this

6A paper that provides an in-depth analysis of explana-tory power of various economics of organization theorieswith respect to the case data is available from authorsupon request.

7The vendor’s and the client’s company names andlocations are disguised.



Table 4. The Development of the Analytical Theme on Vendor ControlSteps Data

Analyze interviewtranscripts and level2 and 3 case write-ups to identifyprominent themes

Example of a quote that motivated the investigation of a new theme:

One of the benefits of an outsourcing contract, particularly a big one, isthat we can control the staffing, so that we can bring people in androtate people out, and provide good career advancement and goodtraining for our employees, while still maintaining a level of service. Thelarger the engagement, in general, the more flexibility we have in doingthat, so we as an organization can be more successful that way andgive more opportunities to our employees. (Vendor’s Branch Manager)

Triangulate acrosssources

Looked for other data supporting and extending the theme:• Vendor’s executives reporting on growth achieved through the methodology,

personnel career development, and customer relationship management andabout using these practices to reduce delivery costs.

• Vendor’s engagement level management reporting on flexibility afforded bylarge engagements.

• Employees reporting on horizontal and vertical rotations enabling multiplecareer choices in and across engagements.

Create data displayin the form of acausal map

See Appendix A, which indicated that various vendors’ management practiceswere dependent on the vendor’s access to many projects and vice versa.

Use all the evidencecollected to confirmor reject the theme.

Coded interview transcripts to check the postulated relationship.

Consolidate themesinto distinctconcepts

Concepts that appeared in the causal map were consolidated: vendor neededscale and scope to reduce the costs of developing competencies, and com-petencies were used to reduce the cost of delivery on each project.

Generate challenges For example asked questions such as:• Is an entire practice bundle dependent on managing many projects or just a

subset of practices?• Are there practices that are weakening the strategy?

Compare inductiveconcepts to existingtheories, analyzeand challengelinkages

Compared inductive themes to management theories: 1. TCE (Williamson 1979) and property rights theory (Hart and Moore 1990)2. Relational contracts theory (Baker et al. 2001)3. Knowledge-based view of the firm (Conner and Prahalad 1996; Grant 1996)4. Resource-based view (Barney 1999; Wernerfelt 1984)5. Core competency concept (Hamel and Prahalad 1996)6. Complementarity in organizational design (Milgrom and Roberts 1995)7. Labor market economics view of outsourcing (Slaughter and Ang 1996)8. Neoclassical economics

Neoclassicaleconomics and corecompetency conceptprovide mostinsights

The theory was used to reinterpret the theme as saying that the scale andscope in IS services was used to develop an experience-based set of practicesthat increase vendor’s productivity and profitability. This doesn’t answer themain research question completely, as it is not clear that a large client couldn’tdevelop these practices in-house.

Link to othertheoretical andempirical themes

Link this theme to other themes developed from data or literature such as theanalyses of market factors, of each practice, of complementarity among prac-tices, and others.



case, to be one of its most successful regionaloperations. In addition to consistently high finan-cial results in both profit margins and revenuegrowth, this branch pioneered a number of opera-tional innovations including software engineeringmethodology improvements, broader employeerecognition programs, and new structures toenable further control over projects.

The Client

The client organization is the Human ResourcesInformation Technology (HRIT) department ofTelecom, one of the world’s premier voice anddata communications companies. Telecom servesmillions of customers including consumers, busi-nesses, and the government, and provides tele-communication outsourcing, consulting, and net-working integration services to large businesses.In recent years, unprecedented changes in thetelecommunications industry, such as the Tele-communications Act of 1996, the emergence ofthe Internet, and the opening of global markets forworldwide competition, forced Telecom to re-evaluate its cost base and streamline its opera-tions. Subsequently, in 1996 and 1997, the com-pany underwent a major restructuring effort,which, among other things, resulted in a reducedIT budget and greater centralization of its IToperations.

Engagement HistoryOverview

At the time of the study, HRIT consisted of about250 people who were responsible for IT support ofthe firm’s corporate human resource function.Throughout the 1990s, Telecom struggled withdecisions on governance of IS resources. Thecontrol over budget related decisions was periodi-cally shifted between HRIT and its corporateclient, the Human Resources (HR) department.One of the users, an HR department manager,explained the history of the relationship:

In 1990, all the budget for system devel-opment was with HRIT. After a couple of

years of tremendous dissatisfaction, thebudget was transferred to the userorganization and we were then called“the customers.” We had the option ofwhere we wanted to spend our dollars,theoreticallybe that with HRIT orexternally. At that point in time HRITwas at its low point in terms of esteem inthe customers’ eyes. Most of us, had wereally been able to realistically, wouldhave gone out of house to have ourneeds fulfilled.…but HRIT really ownedthe Human Resource databases so wereally needed to work with them and wewould’ve been driving a stick to our heart[if we’d gone outside]. Then I think twoor three years ago [in 1995/1996], themoney went back to HRIT.

The struggle that the user described was the well-known struggle between the user’s immediateneed for support and legacy enhancements andthe need to invest in corporate infrastructure (e.g.,PeopleSoft package and client/server computing).Several users referred to the atmosphere as beingpolitical and full of discontent.

In 1993, HRIT was charged with maintaining allsystems according to corporate infrastructurestandards, which required regular upgrades tosoftware and hardware platforms. At the sametime, HRIT was trying to (1) reduce IT main-tenance costs, (2) focus on strategic systems(migration to client/server applications and pack-ages), and (3) sustain or improve customer (user)satisfaction.

IT maintenance costs were high because HRIT’sorganizational processes and incentives weresuch that individual programmers were engagedin addressing each maintenance request ratherthan attempting to reengineer the systems andmake them run more efficiently. One user com-mented that the existing IT environment had lowinteroperability and high levels of redundancyamong various systems, making it difficult forHRIT to fulfill its mandate:

It was an absolute mess. It had to becontrolled and organized.



HRIT had morale issues as well. Individualsresponsible for maintaining legacy systems hadfew opportunities to do new development becausemuch of the new development work was handledby outside IT consultants with expertise on newertechnologies. Assigning the more interestingdevelopment tasks to consultants not only ignitedskill envy, it meant that some HRIT managerswere over-extended due to the administrativerequirements of hiring, training, and compensatingever-changing contracted resources.

A PeopleSoft implementation led to customercomplaints about corporate technology platformupgrades and the struggle for enhancements oncurrent (legacy) systems. By 1994, HRIT wasconcerned about its ability to sustain customersatisfaction. About this time, HRIT’s managementdecided to outsource part of its legacy applicationsupport to reduce maintenance costs while pro-viding HR users with a satisfactory response tochanging business and technological conditions.

In launching the outsourcing solution, HRIT choseABC from a set of about 10 existing contractorsfor a pilot project, which outsourced maintenancefor just a couple of applications. ABC was con-tractually charged with not only maintaining thesystems, but with improving processing efficiencyso as to reduce data center bills. The pilot projectmet all of HRIT’s cost and quality requirementsand led to outsourcing of 25 applications in early1995. Despite its satisfaction with the pilot, HRITused a negotiated competitive bidding approach toselect its service provider.8 According to HRIT’smanagement, the biggest factor behind choosingABC was its demonstrated methodologies and itsfit with the clients’ needs

The contracts were fixed-price for $13 million fortwo years with an agreement on the level ofservice to be provided for the price (called level of

service agreements or LOSAs).9 The agreementsspecified the number of production calls, correc-tive maintenance requests, and enhancements tobe performed for a given system. The agree-ments did not include ABC hiring HRIT personnel.Some HRIT employees took early retirementpackages offered by Telecom, while othersrelocated to other Telecom IT departments. Therunning and performance of the technology wasthe responsibility of ABC (ABC “had the beeper”),but the financial asset ownership remained withTelecom. Telecom also designated a full-timecoordinator to oversee the contract.

At the end of the first full year of the outsourcingcontract, HRIT realized estimated cost savings ofover $1 million dollars in data center processingcosts:

They did a wonderful job.…After sixmonths we were showing annual savingsof somewhere between $250,000 andhalf a million dollars10.…They did a verygood job and they saved us a lot ofmoney. And they kept our customersatisfaction levels very high. (HRIT’sDistrict Manager, the head of HRITorganization)

HR managers (the users) were surprised by thevendor’s effectiveness:

Generally they really do a very thoroughjob. I can’t fault their technical expertise.…If it is approved, by and large theydeliver. [Laughs.] I am laughingbecause we haven’t gotten much by wayof enhancements that we wanted.…They are as effective as they are allowedto be given the budget constraint. Theydo their job well. (HR manager)

8Negotiated competitive bidding allows bidders to submita proposal along with follow-up questions. After theclient answers the questions, vendors submit new bidsbased on the new information.

9The size of the contract appears to be typical. Even in2001, 81 percent of contracts were under $25 millionwith 22 percent between $6 million and $25 million(Casale 2001).

10Other estimates placed savings as high as $2 million.



All interviewed users were satisfied with ABC, andsome rated the success of their relationship as 9.5out of 10. The head of HR simply stated, “We arebetter off.”

Subsequent years’ results also exceeded HRIT’scost savings and quality of service objectives. Inparticular, ABC drove down the number of produc-tion calls and corrective maintenance requests,which permitted more enhancements to existingsystems than anticipated. In 1997, the contractwith ABC was extended for a year and, in 1998, anew contract was granted for two years withoutsoliciting competitive bids. At the same time,HRIT handed over seven additional systems forABC to maintain. ABC was also increasinglytaking on separate assignments for systemcloning associated with HRIT restructuring, legacysystem retirement, and Year 2000 compliance.HRIT’s district manager noted that they were verypleased with the results of the outsourcingagreements:

ABC came in and they were as moti-vated as we were.…They were reallyexceptional in terms of delivering morewith less.

ABC management was also pleased with theoutsourcing outcomes. Like any client relation-ship, the Telecom contract was important forgenerating revenues. ABC’s branch manageremphasized that her focus was on revenue andprofit generation and that the Telecom contractwas hitting the mark on both. Perhaps even moreimportant, the Telecom contract provided ABCwith access to, and control over, a large engage-ment. Across its growing list of contracts, this kindof access and control enabled ABC to apply andhone its competencies:

One of the benefits of an outsourcingcontract, particularly a big one, is that wecan control the staffing, so that we canbring people in and rotate people out,and provide good career advancementand good training for our employees,while still maintaining a level of service.The larger the engagement, in general,

the more flexibility we have in doing that,so we as an organization can be moresuccessful that way and give moreopportunities to our employees.…Wewould be able to successfully proposeand engage in these kinds of assign-ments using our methodology….Themore that we can control and manage,the better it is for us, and if we can do itsuccessfully, the better it is for every-body. (Metropolis Branch Manager)

Case Analysis

ABC’s Market Conditionsand Practices

Figure 1 illustrates the types of challenges thatmany IT organizations, including HRIT, werefacing in the early 1990s. They were under pres-sure to simultaneously cut IT costs and respond torapid business and technology changes.

These needs had to be satisfied in a labor marketcharacterized by high turnover, rising IT workers’salaries, and a scarcity of advanced technicalskills (Slaughter and Ang 1996).11 In addition,legacy application management involved findingways of motivating workers interested in updatingtheir skills by providing them with opportunities fortechnical challenge (Lee et al. 1997; McMurtrey etal. 2002). Driven by these internal needs andlabor market constraints, prospective clients con-tracted IT projects out to specialized vendors whohad demonstrated that they had developed prac-tices to achieve client satisfaction. In this section,we discuss how ABC (1) developed a set of com-petencies that addressed market needs and con-straints, (2) increased the value of each com-petency through patterns of mutual reinforcement,and (3) capitalized on its control over relevantdecision rights on a growing number and variety of

11In 1998, turnover rates of IT staff were approaching 22percent with predicted raises of as much as 25 percentwith every job change (Garner 1998).



Vendor'spractices

Number andvariety ofprojects

contracted out

Clientsatisfaction

Applications ManagementMarket Characteristics

IT Labor Constraints

a. High level of turnover

b. Rise of IT workers' salaries

c. Scarcity of advancedtechnical skills

d. Rewards for skills updating

Client Needs

a. Rapid response to businessand technological changes

b. Cost effectiveness

define

driv

es

leads to higher

increases

Vendor'spractices

Number andvariety ofprojects

contracted out

Clientsatisfaction

Applications ManagementMarket Characteristics

IT Labor Constraints

a. High level of turnover

b. Rise of IT workers' salaries

c. Scarcity of advancedtechnical skills

d. Rewards for skills updating

Client Needs

a. Rapid response to businessand technological changes

b. Cost effectiveness

define

driv

es

leads to higher

increases

Figure 1. Vendor’s Operating Environment

projects to develop competencies and to reducethe marginal costs of service delivery. In the nextsection, we will show why ABC was compelled toshare the value generated by its competencieswith HRIT.

ABC’s Competencies AddressedMarket Needs and Constraints

ABC developed a set of three competencies torespond to client needs and market demands:personnel development, methodology devel-opment and dissemination, and customer relation-ship management.12 These are summarized inTable 5 and discussed below.

IT Personnel Development addressed existing ITlabor market constraints in ways that HRIT hadnot. ABC replaced experienced, high-cost HRITstaff with mostly lower-cost, junior programmers

and then developed their skills through training,mentoring, and team-based project work. Juniorstaff valued the professional growth while theirmentors often relished opportunities to “watchsomebody take off.” As a professional servicesfirm, ABC viewed maintenance work as a first stepin a career development path, which involvedrotating professionals within engagements,assigning personnel development managers, andcreating both technical and management hier-archies. By creating regular opportunities to learnnew skills, interact with team members, andlaunch more exciting careers, ABC was able tomake maintenance work more fulfilling:

Sometimes the work is not overlyexciting. We’ll have 20 applications, andwe can still move people around withinthose applications, and you’re stilllearning something new. It may not bethe technology, but you can still learn…you never get done learning. We haveone guy, for example, who has beenhere for a year and a half who has been

12Researchers’ interpretation of the strategy describedby case study participants.



Table 5. Summary of Competencies

CoreCompetence Personnel Development

MethodologyDevelopment and

DisseminationCustomer Relationship

Management

ConstituentSkills

• Staffing office• Promotion from within• Staff rotation• Junior staff use• Redundant skill

creation• Mentorship• Skill and project

management training• Team-based environ-

ment• Collaboration across

teams

• Engagement levelsoftware engineeringprocess group

• Project office• Corporate practice are

heads• Identification of best

practices• Standardization of

processes• Process documenta-

tion• Methodology training• Work tasks documen-

tation

• LOSA based contracts• Sharing efficiency bene-

fits (surplus) with clients • Management of client’s

goals and priorities• Communicating priorities

and work status• Sharing expertise with

client’s staff• Communicating priorities

and work status

on four different applications, all thesame technology, and he’s loved it. Heabsolutely loves it, because he’s gettingsomething new every six months or so…we can keep the salaries in line andmake it more of an enjoyable environ-ment. (ABC engagement manager)

Users did not favor frequent changes in their ITsupport personnel and cited frequent personnelchanges as their key concern with ABC’s qualityof service. At the same time, users stated thatABC had high quality people and processes tocompensate for turnover effects.

Methodology Development and Disseminationwas necessary for consistent delivery of best ofbreed solutions to client problems. WhereasHRIT’s staff focused on addressing users’immediate needs, ABC introduced methodologiesthat focused on overall operational improvementson projects. ABC had a long history of meth-odology development, which eventually led toadoption of the Software Engineering Institute’s

(SEI) Capability Maturity Model13 and enabledABC to move from level 1 to level 3 certification inless than a year.14 ABC formalized its meth-odology practices through its branch projectoffices, engagement level software engineeringprocess groups, and corporate practice areaheads. These structures ensured methodologydevelopment, improvement, and compliance. Themethodologies not only specified processes, theyalso standardized project documentation throughforms and templates such as change requestforms, lost time logs, and weekly status reportforms, to closely monitor project status. While HRclients complained about having to fill out long

13The Capability Maturity Model (CMM) is a five-levelmodel for judging the maturity of an organization’ssoftware processes (Paulk et al. 1993).

14According to SEI, between 1991 and 1997, companiestook an average of 27 months to move from level 1 tolevel 2 and 18 months to move from level 2 to level 3. InJune 1997 only 12.2 percent of commercial organi-zations were able to receive level 3 CMM compliancecertification or higher.



forms, they accepted this approach as “ABC’sway.” Both ABC and HRIT credited the meth-odology development and dissemination com-petency with introducing significant operationalimprovements and efficiencies in project delivery:

Our goal, and I think, if you look at themethodology, this is what’s supposed tohappen, is to go in and recommendsome process improvements. Find outexactly what’s causing our problems,what’s causing us all the time, and basic-ally that line [points to the LOSA chart]time wise for the production supportshould come way down, to be almostnothing. Now what the user can see isthat that time can be better spent onother things, such as the adaptive MR[maintenance requests], which ischanging the system the way they want,such as this Oracle release that’scoming down, year 2000 stuff, that typeof thing. That’s why methodology...allows us to do that stuff. (ABC Engage-ment Manager)

HR users also saw a significant improvement fromABC’s application of methodologies:

We had too many problems with thatpart of the system. ABC proposed todeal with that giant troubleshooting[problem] by making a minor adjustment[in the program code]. We got a bigimprovement. It was really beneficial.(HR Manager)

Customer Relationship Management was for-malized through LOSAs. Each LOSA set a fixedprice for agreed-upon services:

The major philosophy of outsourcing isthat ABC is taking a risk. ABC is respon-sible for whatever is defined in that clientinterface document as being our respon-sibility. (ABC Branch Manager)

While the LOSA might not lead to greater usersatisfaction with the level of IT services, it did

reduce uncertainty, thereby creating clearer ex-pectations and an acceptance of limits. As usersaccepted these limits, they recognized and appre-ciated services that exceeded contract require-ments. LOSAs clarified, for both users and cor-porate IT, the trade-offs between user needs andcorporate development priorities given budgetlimitations. The negotiation of LOSAs made Tele-com define which level of IT service would benecessary and acceptable over a two year timeperiod. The formal structure of the LOSAs wasenhanced by ongoing personal communicationbetween ABC and HRIT managers, through whichABC was able to clarify priorities, anticipateresource requirements, and report on issues andchanges in project status.

ABC’s Competencies WereMutually Reinforcing

Management practices targeted at one compe-tency tended to enhance the other competenciesas well. This reinforcing pattern was apparent inall three pairings of the competencies, as shownin Figure 3 and described below.

Personnel Development and MethodologyDevelopment and Distribution. The method-ology competency reinforced personnel develop-ment by helping junior staff learn quickly what theywere expected to do:

If I don’t know in the beginning what thegoal is, then I’m flying blind. I don’t wantto reinvent the wheel, so my incentive tofollow [the methodology] is that I wouldrather use something that’s there andhas been proven to work than have tocome up with it all on my own. (ABCJunior Consultant)

While methodologies were sometimes viewed asconstraining individual initiative, one junior con-sultant argued that the methodology empoweredher and others to challenge management direc-tives that might be inconsistent with documentedpractices. In addition, the standardization of prac-tices around methodology facilitated staff rotationsand scheduling.



In the same way, personnel development prac-tices, such as skill development, rotations, andpromotion policies, provided training, encourage-ment, and incentives that led to consistent useand improvement of methodologies across theorganization:

There is a fair amount of cross-pollination here in that when a branchthat has limited experience in this startsto do this kind of work, they usually gethelp from someone who has [experi-ence]. For example, this week I was inAtlanta for three days working with them,so they’re going to do pretty much whatwe would do. (ABC Practice Area Head)

Methodology Development and Distributionand Customer Relationship. When method-ology delivered operational improvements, ABCcould sometimes increase service levels with noadded cost to Telecom. In some cases, ABC hadbeen able to pull people off a project and hadelected to share the savings with Telecom. Thesevery visible improvements in IT service levelsreinforced the customer relationship:

As we improve the quality of the deliveryof services, we might gain more time.So we might actually increase the levelof service without additional cost. It isvery common for us because of ourmethodology to improve or increase thelevel of service during the contract.(ABC Project Manager)

Methodological approaches also improved custo-mer relationship management practices bydefining and standardizing best practices forcreating and managing LOSAs.

The customer relationship management compe-tence similarly reinforced the methodology compe-tence. ABC regularly communicated with HRIT todiscuss issues and expectations, and oneoutcome was to help HRIT managers understandthe methodologies so that they could facilitate,rather than hinder, ABC’s ability to meet expec-tations. Thus, HRIT managers shared their knowl-

edge of their systems with ABC and provided earlywarnings, where possible, when business orcorporate IT changes might have an impact onABC’s responsibilities:

They’ll call us today, for example, andsay, “Hey, I heard about this change thatwill be coming down.” We’ll take [thatinput] and take a high level look at howthat’s going to affect us. So when arequest comes down, we’ve already gotsome stuff ready to go. (ABC Engage-ment Manager)

Personnel Development and Customer Rela-tionship. Personnel development practices rein-forced customer relationships by ensuring thatstaff understood and accepted accountability formeeting contractual obligations. Personnel devel-opment practices also developed communicationskills to help staff establish customer expectationsand build trust. ABC’s practice of developingjunior staff through project teams and mentorsthen positioned ABC to deliver:

[HRIT] may not feel like they’re gettingtheir money’s worth if somebody wasnew and didn’t understand all thetechnologies. But out here, since it’sfixed price, ABC can sort of suck upsome of that initial loss of time due toramp-up because they decide how manypeople are out here and how much timethat they work. (Junior Developer)

At the same time, strong customer relationshipsled to better buy-in, on the customer’s part, topersonnel development policies that requiredrelease time or movement of personnel, such astraining programs, mentoring, and job rotations:

[HRIT managers] realize that rotation isa key. They realize that, and they’refans of it....If we’re rotating peoplearound and yet keeping them on theTelecom contract, that’s fantastic.... Butthe client feels better with the morepeople we can get with multiple appli-cation knowledge, that’s the advantage



of application outsourcing. If you lockedsomebody into that job full-time, thennobody is going to win long-term. All theknowledge base is going to be in onearea. If that person leaves, then theworld shuts down. That’s what we’retrying to prevent. (ABC EngagementManager)

ABC’s Control over Relevant Decision Rightson a Variety and Multitude of ProjectsEnabled Growth in Competencies

Consistent with neoclassical economics, owner-ship of a large number of projects provided ABCwith economies of scale in IT services provi-sioning. Its scale justified its upfront investment incompetencies. For example, developing a meth-odology takes considerable resources, such ascorporate development offices, compliance audits,and detailed documentation. However, the margi-nal cost of applying a methodology to an engage-ment is lower than regularly rediscovering whatworks (Banker and Slaughter 1997). Similarly,investing in personnel development managementand corporate training programs only pays offwhen the benefits are shared among many em-ployees, which eventually leads to lower marginalcost of personnel deployment on projects.

ABC’s access to, and control over, a large numberof diverse projects also bestowed economies ofscope that came with specialization in IT services.All three of ABC’s competencies were knowledge-intensive, so its broad scope allowed the firm tolearn from its experiences and apply its learning tonew, but related, situations (Hamel and Prahalad1996). For example, innovative methodologiesresulted from experimentation, documentation,and reuse across many different clients. Similarly,personnel development practices required theability to offer staff a wide range of experiences.ABC’s customer relationship management compe-tency depended on learning managerial skillsassociated with developing and managing LOSAs,such as communications, cost estimation, rela-tionship building, and metrics design, and fulfillingfixed-price contracts.

Most critically, ABC’s control over relevant ITservice management decisions on client projectsallowed it to simultaneously apply all of itscompetencies to engagements, thereby derivingthe benefits from mutual reinforcement of thecompetencies. As the branch manager stated, itwas the application of staffing and methodologypractices to the LOSA-based contracts thatallowed ABC to be more efficient and effective onprojects.

As shown in Figure 2, ABC’s access to, andcontrol over, a large number and variety of ITprojects provided economies of both scale andscope. The large number of projects gave ABCan ability to develop and improve competencies,as well as opportunities to reuse its competenciessimultaneously, so that it could increase clientsatisfaction. Client satisfaction helped ABCsecure additional engagements, which enabledeven greater efficiencies in the competencies. Inessence, ABC was getting the benefits ofspecialization in IT services as seen through thecore competency lens (Hamel and Prahalad 1996;Quinn 1999).

ABC’s three core competencies were valuable inthat they addressed the needs of clients andindustry constraints. The resource-based view ofthe firm (Barney 1991; Wernerfelt 1984) wouldsuggest that core competencies are uniqueresources of the firm, which are hard to imitate byother firms. Nonetheless, it is possible that ABC’spotential and existing clients could have devel-oped and applied one or two of these compe-tencies internally. Indeed, many firms outside theIT service industry have done so. The reason whyfirms purchased ABC’s services—or that ofanother vendor—is because they believed(despite mixed results) that the cost of purchasingthese services from a vendor, in both financial andother terms, was less than the combined cost ofdeveloping and applying them internally plus thecosts of contract monitoring. ABC gained a costadvantage in developing and applying its compe-tencies because it benefited not only from havingindividual competencies, but also from comple-mentarities among the competencies. The econo-mic concept of complementarities helps us under-



ABC's Core Competencies

Number and Variety ofProjects Controlled by ABC

ABC's ClientSatisfaction

drives andenables to build

and apply

provides higher levelsat lower costs

increases


Dissemination

IT PersonnelCareer

Development

ClientRelationshipManagement

ABC's Core Competencies

Number and Variety ofProjects Controlled by ABC

ABC's ClientSatisfaction

drives andenables to build

and apply

provides higher levelsat lower costs

increases


Dissemination

IT PersonnelCareer

Development

ClientRelationshipManagement

Figure 2. ABC’s Core Competencies: Development and Role

stand this phenomenon. The client-vendor rela-tionship literature will help us understand why theclients benefited as well.

Complementarity in OrganizationalDesign

The concept of complementarity in organizationaldesign posits that firms can improve productivityby engaging in complementary activities wherebenefits from doing more of one activity increaseif the firm is also doing more of the other activity(Milgrom and Roberts 1990). This concept drawslargely on a mathematical model developed byTopkis (1978),15 and allows us to conclude thecomplementarity of a whole system of activities

based on pair-wise complementarities. Its majorstrength is in having no strong assumptions aboutthe differentiability properties of the productionfunction.16

This concept of complementarity has been used instudies of manufacturing to show that modernmanufacturing approaches work as a system,rather than as a set of independent factors(Ichniowski et al. 1997; Milgrom and Roberts1995). IS researchers have applied the conceptto examine the complementarity of IT investmentswith other factors. For example, studies by Bryn-jolfsson and colleagues showed that IT invest-ments were complementary with the distribution ofdecision authority, human capital investments,and performance-based incentives (Brynjolfsson

15Ffor a more technical overview, see Barua et al. 1996.16For further clarification of what complementaritiestheory is—and is not—see Appendix B.



et al. 1997; Hitt and Brynjolfsson 1997). As aresult, firms that invested in IT and adjusted theirorganizational structure and processes performedbetter than firms that invested in IT withoutchanging complementary practices. Barua andcolleagues built a theoretical model that showedthat payoff from IT investments increases whenaccompanied by reengineering of businessprocesses and changes in decision authority andincentives (Barua et al. 1996) and extended themodel in relation to firms’ market-based objectives(Barua and Mukhopadhay 2000).

While the term complementarity appears broadand can cover a wide range of market- andorganization-based phenomena that exhibitpositive feedback, we draw specifically on Milgromand Robert’s (1990) development of the conceptwhich pertains to complementarities in the factorsof production in the firm’s production function.This allows us to conclude from the rich theoryand studies of production factors complemen-tarities that those firms that invest simultaneouslyin several complementary activities perform betterthan those firms that increase the level of some ofthese activities, but not others. In fact, literatureon complementarity argues that firms thatincrease one factor without also increasing com-plementary factors may be worse off than firmsthat keep the factors at the same lower level (Hittand Brynjolfsson 1997; Ichniowski et al. 1997;Milgrom and Roberts 1990).

Milgrom and Roberts’ (1990) concept of comple-mentarities does not pertain to the fit betweenfirms’ competencies and labor market conditions(often referred to as strategic alignment). Forexample, labor market shortages, or the need tosupport frequent business changes of the client,are exogenous factors in the production function.The concept, however, pertains to complemen-tarities in endogenous factors that the firm cancontrol. Given certain technological and marketfactors, the economic theory postulates the effectsof relationships among production factors thatfirms do control on production outcomes.

This concept also does not describe the vendors’and clients’ activities as complementary even if

the vendors target their competencies to fit clientneeds. Indeed, vendor and client activities are notcomplementary because vendors and clientsoptimize their levels of investments in their corebusiness activities independently, rather than inunison. If vendor and client activities were com-plementary, it would make sense to integrate theminside firm boundaries so as to have an ability tooptimize levels of inputs simultaneously.

Finally, this concept does not refer to the fitbetween a vendor’s access to the variety andmultitude of projects and its ability to developcompetencies—the traditional core competencyargument. This is again because access to thevariety and multitude of projects is, for the mostpart, an exogenous factor to the vendor. At thesame time, the concept of complementarities doesenhance the core competencies-based argumentfor specialization in production (Milgrom andRoberts 1992, p. 554). While the core compe-tency theorists argue that experience-basedlearning lowers the costs of investing inknowledge-intensive competencies (Hamel andPrahalad 1996), the complementarities conceptadds that specialized firms would gain a naturalmomentum because applying complementarycompetencies together increases production andleads to greater market share and, hence, furtherexperience-based learning (Milgrom et al. 1991).

ABC enjoyed a natural momentum in the interplaybetween gaining control over additional projectsand building competencies to deliver results. Asalready noted,ABC’s three core competenciespaired to create three complementarities (seeFigure 3). Using the economic theory, we con-clude, based on pair-wise complementaritiesamong practices, that the whole system (i.e.,ABC’s production function) exhibited such com-plementarities. Hence, ABC was best off when itsimultaneously raised the levels of all of theseactivities.

Outsourcer’s Value Proposition

The concepts of complementarities and corecompetencies explain that ABC can increase



productivity and reduce costs on client projects byapplying a set of complementary applicationmanagement competencies. In this section, weexamine how ABC delivers value to clients as aresult of its ability to develop complementarycompetencies. First, we go beyond neoclassicaleconomics theories to explain why potentialclients are unlikely to develop these comple-mentary competencies internally. We thenexplore the mechanisms that ensure that thebenefits of ABC’s complementary competenciesare, in part, passed on to clients.

Why Clients Do Not Replicate andApply ABC’s Competencies

Telecom’s primary goal was to excel intelecommunications services, which had a dif-ferent set of market structures and resource con-straints than the IT services industry. Accordingly,Telecom had organized into divisions, createdorganizational structures, distributed decisionrights, developed business processes, and hiredpersonnel to address the market conditions andcustomer demands of the telecommunicationsindustry. Telecom could have attempted to buildIT application management competencies ratherthan outsource to ABC, but, unlike ABC, it mighthave found that optimizing the development andapplication of IT competencies would conflict withoptimizing core business activities. ABC, on theother hand, could shield itself from these conflictsthrough the structure provided by the contract,which specified deliverables rather than levels ofinvestment in competencies.

For example, to address labor market constraints,Telecom could increase the compensation oftechnical specialists, but non-IT workers mightperceive the inflated IT salaries as unfair.Similarly, Telecom was not as well positioned asABC to institute an IT personnel career devel-opment office or a practice of IT personnel rotationand promotion. HRIT had tried to institute appli-cation management methodologies similar tothose used by ABC, but had difficulty changingapplication development processes that accom-modated each user’s request at the expense of

system-wide improvements. Also, HRIT couldhave implemented service level agreements(SLAs) comparable to ABC’s LOSA, but researchshows that internal SLAs rarely satisfy users’desires for a sense of control over IT servicelevels (Ross et al. 1999).

Evidence from the case indicates that HRIT sawmany of the competencies developed by ABC asproper solutions to the business problems HRITwas facing. However, HRIT was unable to imple-ment these solutions internally due to the tensionthat the centralization of decision authority over ITdecisions caused with users. Outsourcing con-tracts helped manage the boundary with the usersclarifying which decision rights would stay withusers (e.g., asking which support requests toperform) and which ones would be given to thevendor and HRIT (e.g., deciding how many sup-port requests to perform in a given time period).As an outside firm, ABC experienced less tensionat the IT-user boundary because users werefocused on contractual obligations rather than onABC’s processes for selecting, assigning, andcompensating its staff and implementing itsmethodologies.

If it could overcome obstacles to centralization ofIT decision rights, HRIT might be able to developand apply one or more application managementcompetencies. Yet, theory supporting comple-mentarity in organizational design holds thatdeveloping one or two of ABC’s competencieswould leave Telecom at a disadvantage relative tothe services ABC could offer, because ABC couldgenerate a premium as a result of the comple-mentarities among its competencies. In addition,this premium is likely to grow because specializedfirms like ABC tend to grow by offering new kindsof IT services to a wider range of clients (Currie2000) thereby accelerating the reinforcing effectsof the competencies. A non-IT firm like Telecomwould put lower priority on increasing the pool ofIT projects to provide experience-based learning.Instead, Telecom’s IT governance structureswould—and should—prioritize projects and growIT budgets based on business needs, rather thanthe learning requirements of their IT departments(Brown 1999; Chan et al. 1997). Thus, ABC also



had the advantage of the momentum that resultsfrom experience-based learning over time, whichnon-IT firms—even if they developed similarcompetencies—could not replicate.

In short, consistent with core competency con-cepts (Quinn 1999), HRIT could develop compe-tencies similar to those offered by ABC, but itfaced more significant obstacles in developingand applying all of them than a specialist IT firm.Specialized firms can focus on optimizinginvestments in their core business (IT services)and use contractual mechanisms to manage theboundary with other business activities (coreactivities of their clients). The key insight offeredby the notion of complementarity in organizationaldesign is that optimizing business activitiesinvolves changing levels of different activitiessimultaneously, and unless firms have decisionrights over all related practices, they would beunable to do so. Application management ven-dors like ABC acquire such rights through servicelevel-based contracts. Many internal IT depart-ments lack this ability as critical decision rights aredistributed throughout user organizations andcorporate management. In this sense, the vendorcan gain productivity and cost advantages bycentralizing some IT decision rights through thestructure of a contract so as to develop and applya set of complementary core competencies in ITservice delivery.

The case presented here focuses on the specificset of competencies that ABC developed toaddress HRIT’s needs for low-cost, high-qualityapplication management and responsiveness toopportunities created by new technologies. Theseneeds, as well as the existing labor marketconstraints, were typical of many firms in the late1990s (Bresnahan et al. 2002; Moore and Burke2002; Slaughter and Ang 1996). Our case, aswell as existing literature on outsourcing (Hirsch-heim and Lacity 2000), indicates that outsourcingoften provides efficiency at the cost of flexibility.This was the trade-off that Telecom accepted inthe mid-1990s. Nonetheless, client needs candiffer from firm to firm and will certainly differ overtime. Changes, for example, in the IT labor mar-

ket, can render some competencies less valuable,while creating a market for new competencies.This study did not examine the implications ofchanging competencies, but core competencytheory suggests that the ability to accelerateexperience-based learning related to IT servicesshould provide a lasting value proposition forthose vendor firms which recognize and respondto changing client needs.

Why Vendors Share ProductivityGains with Clients

From the client perspective, the outsourcer’s valueproposition would not exist if the benefits of com-plementary competencies and strategy accruedsolely to the outsourcer. Fortunately, contract-based, interpersonal, and reputation-based mech-anisms encourage vendors to share advantageswith clients. Telecom effectively deployed somecontract-based mechanisms including pilot pro-jects (Snir and Hitt 2002), multiphased contractingwith penalties (Whang 1992), interpersonalrelationship building (Kern 1997; Lee and Kim1999; Sabherwal 1999), carrot and stick incen-tives and shorter-term contracts (Lacity et al.1995), and competent contract monitoring (Casale2001; Lacity et al. 1995; Sabherwal 1999). All ofthese mechanisms increased client control andmotivated ABC to demonstrate value to the client.Since the value of outsourcing to the client is veryhard to measure, most researchers have focusedon client satisfaction (Goles 2001; Lacity andWillcocks 1998; Saunders et al. 1997). In thiscase, HRIT managers, HR users, and vendors allpointed out that significant cost savings and clientsatisfaction were achieved.

ABC and HRIT emphasized the importance ofbuilding interpersonal relationships (consistentwith recommendations of Kern and Willcocks[2001], Poppo [2002], and Sabherwal [1999]).HRIT assigned a full-time engagement managerthat interfaced daily with ABC’s engagementmanager. To obtain support of HRIT’s IT person-nel, who still remained after outsourcing, ABCemphasized the role of consultants as helpers:



I remember this one guy....He basicallywas responsible for all the applications inthe running user interface between usand the actual computer operators.…Hehated ABC just because they were ABC.In about two or three months he was thebest guy in the world. He just called melast week and invited me to go fishingwith him. He just retired, and we’vegiven him a lot of information over theyears.…Telecom people will tell you thatright now. Whatever they need, they cancome to us and we’ll get them answers.(ABC’s Engagement Manager)

Reputation-based mechanisms (Elitzur andWensley 1997) provided ABC with another strongincentive to share productivity gains with Telecom.These mechanisms have not been studied indepth in the IS outsourcing literature, but are wellknown institutional incentive mechanisms ineconomics , sociology, marketing, and strategy(Baker et al. 2001; Fudenberg and Levine 1992;Hardin 2002; Herbig et al. 1994; Low 2002; Weisset al. 1999; Wilson 1985). A recent study byGoles (2001) also found that IT service vendorsfocus on reputation building in their relationshipswith clients. In addition to their current contractingstructure, vendors care about their long-termmarket position. Thus, the vendor is inclined toshare benefits with the client so that the infor-mation about the vendor’s contribution enables itto win future contracts. Developing a solidindustry reputation helped ABC win new, andextend existing, engagements, which led to theacquisition of, and control over, more projects.ABC’s track record of performance at HRIT helpedit win new contracts with HRIT, with other Telecomdivisions, and with other companies in tele-communications and other industries. ABCinvested heavily in proving its reputation forquality, for example, by investing in CMM com-pliance certification on many of its engagements.Industry research consulting groups such asGIGA, Gartner, IDC, Forrester, and Jupiter oftenevaluate the performance of IT service providers,thus influencing their reputations. ABC ranked

consistently high in those evaluations. To sum-marize, ABC was not only able to develop costadvantages, it also had strong incentives to sharethose advantages with its clients.

The outsourcer’s strategy and practices are de-picted in Figure 3. This model of the IT vendor’svalue proposition suggests that client needs, asshaped by market constraints, specify the require-ments for client satisfaction. Client satisfactionresults from services provided by vendors throughthe application of a complementary set of corecompetencies targeted at delivering higher serviceat a lower marginal cost. Client satisfaction isachieved when the application of core compe-tencies to projects is enabled by a healthy client-vendor relationship, which is in part influenced bythe vendor’s expertise in managing client relation-ships. Competencies, in turn, grow through thevendor’s firm-wide experience gained from con-trolling a large number and variety of projects,which, in turn, grow due to the reputation the ven-dor develops through its ability to satisfy custo-mers. The model represents a set of positivefeedback loops, which will result in negativeoutcomes if, for example, the competencies do notmatch client needs.

Conclusion

Findings from the case study indicate that an ITapplication management vendor can deliver valueto its clients by developing a set of experience-based core competencies that (1) address clientneeds and market conditions, (2) exhibit com-plementarities that result in efficient servicedelivery, and (3) depend on the vendor’s controlover, and centralization of, decision rights on alarge number of projects from multiple clients.These core competencies result in a compellingvalue proposition when investing in the growth ofany of these competencies in-house conflicts withoptimization of the client’s core business activitiesand when contractual- and reputation-based incen-tives encourage vendors to share the efficiencygained from these competencies with their clients.



define

Reputation

Vendor's Core Competencies

help build

enables

ApplicationsManagement

MarketCharacteristics

ClientSatisfaction

Client-VendorRelationship

provides higher levels atlower marginal costs

improves

leads to

increased Number and Variety ofProjects Controlled by

the Vendor

drives andenables tobuild and

applydefine

Reputation

Vendor's Core Competencies

help build

enables

ApplicationsManagement

MarketCharacteristics

ClientSatisfaction

Client-VendorRelationship

provides higher levels atlower marginal costs

improves

leads to

increased Number and Variety ofProjects Controlled by

the Vendor

drives andenables tobuild and

apply

Figure 3. Vendor’s Value Proposition

Implications to Research

In this paper, we have added one of the missingpieces to the IT outsourcing puzzle. For the pastdecade, the majority of research on IT outsourcingfocused on client firms’ reasons for outsourcing.More recent literature has focused on the out-sourcing relationship itself, looking at how differentcontractual structures and noncontractual aspectsof the relationship influence outsourcing out-comes. Finally, some work looked at the client’sability to manage vendors. However, the vendor’sperspective has been largely left unexamined.

We have drawn on prior work on the vendor-clientrelationship and used core competency andcomplementarity in organizational design con-cepts to identify the vendor’s value proposition.We discovered that IT outsourcing vendors canoffer benefits to clients by developing a system ofcomplementary core competencies that depend

upon a vendor’s control over decision rights onmany projects. We have argued that this com-plementary system is hard to replicate inside non-IT firms because it is likely to conflict with theirmain business practices. Prior work on ISoutsourcing emphasized that the vendor selectionprocess was important and that vendors, at aminimum, had to provide technical expertise andbe able to manage the relationship (Goles 2001;Saunders et al. 1997). At the same time, priorwork pointed out that cost savings is the numberone reason for outsourcing IT services. Our workexplains how vendors are able to provide high-level technical capability and manage therelationship cost effectively. More studies of thevendor’s side of the outsourcing relationship areneeded to develop and validate our case studyfindings.

The application of the complementarities conceptto outsourcing extends neoclassical economics



(Ang and Straub 1998) and labor market-basedapproaches to outsourcing (Slaughter and Ang1996) by arguing that, in application management,vendors use economies of scale and scope tocombat specific labor market and business condi-tions by developing complementary practices. Italso adds strength to the popular core compe-tency approach to outsourcing (Barney 1999;Hamel and Prahalad 1996; Quinn 1999) byarguing that specialized firms gain momentum indeveloping world-class competencies due to thecomplementarities among their practices. Wehave shown, through the case data, how an ITvendor was able to develop competencies, whichwere costly to imitate by a client organization—one of the key questions considered in corecompetency and resource-based views of the firm(Barney 1986, 1991, 1996; Wernerfelt 1984).

A formally described theory of complementarityapplied to factors of production is only a decadeold and, to our knowledge, has not been used inpublished studies on outsourcing. While prior ISliterature that applied this theory used IT servicesas an input to production, from the vendor’sperspective IT services are outputs of a produc-tion function. In addition, our work could be usedas a starting point for understanding how thecomplementarities concept can be used to answerquestions on firm boundaries more generally. Forexample, a firm looking to expand into newmarketplaces (vertically or horizontally) may wantto examine whether the competencies needed toexcel in these markets are complementary,independent, or conflicting with the competenciesdeveloped for the current business. This kind ofexamination would also involve the question of therisk that the firm takes when outsourcing ITservices, traditionally addressed by TCE andrelated theories (Hart and Moore 1990; Richmondand Seidmann 1993; Richmond et al. 1992). Itwould be fruitful to develop a theoretical modelwhich ties together the trade-offs that firms facebetween, on the one hand, contracting out to avendor that has developed a complementary setof core competencies and facing contracting costsand risks and, on the other hand, forgoing thebenefits of specialization and complementarity inorganizational design. An empirical comparison of

different kinds of IS outsourcing, and between ISoutsourcing and outsourcing in other services andproducts, may provide interesting insights on this(see, for example, work in pharmaceutical industryby Azoulay 2001).

While we have not developed formal propositionsor mathematical models, future research canbenefit from such formalization as well as draw onrecent advancements in econometrics of comple-mentarities (Athey and Stern 1998). The greatestadvantage of this theory, from the standpoint ofeconometric studies, is that one does not need toassume a particular production function or even itsdifferentiability to test the theory. Interestingquestions to explore on a broad scale include

(1) Do IS vendor practices generally exhibitcomplementarities?

(2) Do successful internal IS practices exhibitcomplementarity?

(3) Are those outsourcing arrangements wherevendor’s complementary practices would notbe complementary to client’s own practices,if implemented internally, more successfulthan those where vendor’s practices arecomplementary among each other as well aswith client’s?

Controlling for client-vendor relationship factors,this latter question would be the most direct test ofour theoretical development. One can also testthis in different domains of outsourcing, controllingfor firm size (economies of scale) and experience-based learning factors.

A future challenge is to understand how firmsinterested in developing IT competencies canevolve their practices to address changing marketconditions. For example, as IT labor becomesless scarce, the set of competencies that clientsvalue is likely to change (McDougall 2002). Yet,when retention of employees becomes less of anissue, concerns about motivation in low-skill jobsremains (Lee 2000; McMurtrey et al. 2002; Mooreand Burke 2002). Moreover, there is almostcertainly variation in which core competencies are



valuable across IT services other than applica-tions management. For example, having stringentdevelopment methodologies might prove detri-mental to the success of more creative andinnovative application development projects. Acloser examination of the evolution of IT marketconditions, client needs, and valuable corecompetencies would seem a fruitful area for futureresearch. Generalizing from our study, we seethat centralization of decision rights over a varietyand multitude of projects, managing experience-based learning, and value sharing with clientswere key factors enabling the growth of compe-tencies. To build a broader theoretical modeldrawing on these factors, it would be fruitful tolook at the literature on dynamic capabilitydevelopment (Eisenhardt 1999, 2000, 2001;Makadok 2001), which examines the interplaybetween enabling market conditions and firmcapabilities.

Implications to Practice

The findings reported here suggest that thebenefits of IT application management out-sourcing can be significant to firms of all sizes.This is because IT vendors, by virtue of their focuson IT service delivery, have the opportunity todevelop a set of core competencies that generatesignificant benefits and are enhanced throughtheir complementarities. This does not mean thatfirms necessarily will generate benefits from out-sourcing. Actual benefits will depend on (1) theability of the firm making sourcing decisions todetermine the consistency between its own needsand the competencies available in the market-place, (2) the selection of the vendor and manage-ment of the relationship, and (3) the vendor’scompetencies and their complementarities.

First, clarifying the outsourcer’s value propositionshould help potential clients determine whetheroutsourcing is right for them. Firms can start byexamining the IT service market conditions thatthey face. They can then look at the compe-tencies developed by IT service vendors andexemplar internal IT departments to address

market conditions. They then need to judgewhether optimizing the level of investment in anduse of these competencies is consistent withoptimizing activities in their own core business.Where this is the case, they should be able togenerate complementarities similar to those ofvendors, and thus would likely realize little or novalue from outsourcing. Where firms’ core com-petencies are different from those of vendors, theyshould recognize the difficulty of first developingimportant competencies and then applying themtogether so as to benefit from complementarities.Meanwhile, the growing body of literature on ITgovernance structures can help firms learn aboutaligning IT competencies with non-IT activities(Brown 1999; Sambamurthy and Zmud 1999).

Second, vendor selection has been shown to bea critical determinant of outsourcing success(Casale 2001; Saunders et al. 1997). This studyindicates that potential clients should focus onvendor core competencies in making the choice.In addition to selecting the right vendor, effectiveoutsourcing appears to require that clients givetheir vendors some freedom to apply their compe-tencies. This means that they must carefullynegotiate the service level agreement, but thengive the vendor control over how the contract isfulfilled (Bendor-Samuel 2002). Even thoughclient firms must normally manage the relationshipand help with vendor methodology implemen-tation, they also must develop an understanding ofhow to manage resources that they do not own(Elitzur and Wensley 1997; Henderson andVenkatraman 1990; Kern 1997; Lacity and Will-cocks 1998; Lacity et al. 1995; Whang 1992;Sabherwal 1999; Saunders et al. 1997; Useemand Harder 2000). The outsourcing firm retainsresponsibility for applying a vendor’s compe-tencies to practice. For example, clients mayneed to provide dedicated personnel for relation-ship management. They may need to work withinthe host organization to help with vendor method-ology implementation, produce documentation,and understand the need for vendor staff rotation.Most importantly, clients must specify their expec-tations and priorities.

Finally, in order for clients to benefit from out-sourcing, vendors must recognize the compe-



tencies that constitute a compelling value propo-sition. Market needs will change, thus changingthe particular competencies that firms will findvaluable. For example, by 2002, the market for ITskills had significantly turned around. The needfor a personnel development competency con-cerned with retention was reduced. Other compe-tencies may still be important and complemen-tarities among the competencies will be a criticalsource of value, but vendors should recognize theneed to reassess market conditions and clientneeds and regularly re-create a complementaryset of competencies that addresses them.

Acknowledgements

This work was sponsored by MIT’s Center forInformation Systems Research (CISR). We aredeeply grateful to Jack Rockart for his role insetting up this project and for supporting us, bothfinancially and intellectually, in completing thiswork. We would also like to thank CISR’s supportstaff for helping us throughout our writing andresearch process. Insightful comments on earlierdrafts of this paper were provided by WandaOrlikowski, JoAnne Yates, and Lotte Bailyn as wellas the BPS Fall 1998 doctoral research seminarparticipants at the Sloan School. We thank themfor helping us in our numerous rewrites. We arealso very grateful to the companies participating inour studies. They were most generous with theirtime and energy and we appreciate their open-ness in sharing their experiences with us. Thispaper benefited greatly from the tireless efforts ofthe senior editor, V. Sambamurthy, SandraSlaughter, the associate editor, and three anony-mous reviewers.

References

Agar, M. The Professional Stranger: An InformalIntroduction to Ethnography, Academic Press,New York, 1980.

Ang, S., and Beath, C. M. “Hierarchical Elementsin Software Contracts,” Journal of Organi-zational Computing (3:3), 1993, pp. 329-361.

Ang, S., and Cummings, L. L. “Strategic Re-sponse to Institutional Influences on Informa-tion Systems Outsourcing,” OrganizationScience (8:3), May-June 1997, pp. 235-256.

Ang, S., and Straub, D. “Production and Trans-action Economies and IS Outsourcing: A Studyof the U.S. Banking Industry,” MIS Quarterly(22:4), December 1998, pp. 535-552.

Applegate, L. M., and Montealegre, R. “EastmanKodak Co.: Managing Information SystemsThrough Strategic Alliances,” Harvard BusinessSchool Case 192-030, Boston, MA, 1991.

Athey, S., and Stern, S. “An Empirical Frameworkfor Testing Theories About Complementarity inOrganizational Design,” Working Paper #6600,National Bureau of Economic Research, 1998.

Aubert, B. A., Dussault, S., Patry, M., and Rivard,M. “Managing the Risk of IT Outsourcing,” inProceedings of the Thirty-Second Hawaii Inter-national Conference on System Sciences, IEEEComputer Society Press, Los Alamitos, CA,1999, pp. 7068-7079.

Aubert, B. A., Patry, M., and Rivard, M.“Assessing the Risk of IT Outsourcing,” inProceedings of the Thirty-First Hawaii Inter-national Conference on System Sciences, IEEEComputer Society Press, Los Alamitos, CA,1998, pp. 685-693.

Azoulay, P. “Acquiring Knowledge Across andWithin Firm Boundaries: Evidence from Clini-cal Development,” in Three Essays on theDiffusion and Development of PharmaceuticalInnovations, Unpublished Ph.D. Dissertation,Massachusetts Institute of Technology, 2001.

Baker, G., Gibbons, R., and Murphy, K. J. “Rela-tional Contracts and the Theory of the Firm,”The Quarterly Journal of Econmics (117:1),2002, pp. 39-84.

Banker, R., and Slaughter, S. “A Field Study ofScale Economies in Software Maintenance,”Management Science (43:12), 1997, pp. 1709-1725.

Barney, J. B. “Firm Resources and SustainedCompetitive Advantage,” Journal of Manage-ment (17:1), 1991, pp. 99-120.

Barney, J. B. “How a Firm’s Capabilities AffectBoundary Decisions,” Sloan ManagementReview (40:3), Spring 1999, pp. 137-145.

Barney, J. B. “The Resource-Based Theory of theFirm,” Organization Science (7:5), 1996, p. 469.



Barney, J. B. “Strategic Factor Markets: Expec-tations, Luck, and Business Strategy,” Manage-ment Science (32:10), 1986, pp. 1231-1241.

Barua, A., Lee, S. C. H., and Whinston, A. B.“The Calculus of Reengineering,” InformationSystems Research (7:4), December 1996, pp.409-428.

Barua, A., and Mukhopadhay, T. “InformationTechnology and Business Performance: Past,Present, and Future,” in Framing the Domainsof IT Management, R. Zmud (ed.), PinnaflexInc., Cincinnati, OH, 2000, pp. 65-84.

Benbasat, I., Goldstein, D. K., and Mead, M. “TheCase Research Strategy in Studies of Infor-mation Systems,” MIS Quarterly (11:3), Sep-tember 1987, pp. 369-386.

Bendor-Samuel, P. “CIO Radio Interview,” 2002(http://www2.cio.com/radio/cio/cat.pl?interview=peter_bendor2_01, June).

Bresnahan, T., Brynjolfsson, E., and Hitt, L. M.“Information Technology, Workplace Organi-zation, and the Demand for Skilled Labor:Firm-Level Evidence,” The Quarterly Journal ofEconomics (117:1), February 2002, pp. 339-376.

Brown, C. V. “Horizontal Mechanisms UnderDiffering IS Organization Contexts,” MIS Quar-terly (23:3), 1999, pp. 421-454.

Brynjolfsson, E., Renshaw, A. A., and VanAlstyne, M. “The Matrix of Change,” SloanManagement Review (38:2), Winter 1997, pp.37-54.

Buxbaum, P. “Bringing IT Back Home,” Com-puterworld (36:31, July 29, 2002, p. 38.

Caldwell, B. “2001 Trends in IT OutsourcingDelivery, Solution Development, Marketing,Sales and Alliances,” Gartner DataquestReport, Stamford, CT, January 22, 2002a.

Caldwell, B. “Outsourcing Cost ReductionCreates Paradox: How to Still Make a Profit,”Gartner Dataquest Report, Stamford, CT, April12, 2002b.

Caldwell, B., and McGee, M. K. “OutsourcingBacklash,” Information Week (650), September29, 1997.

Casale, F. “Outsourcing Index 2000,” The Out-sourcing Institute /Dunn and Bradstreet Report,November 15, 2000 (available online at http://www.outsourcing.com/).

Casale, F. “IT Index 2001,” The OutsourcingInstitute/Sunguard eSourcing Report, April 15,2001 (available online at http://www.outsourcing.com/).

Chabrow, E. “The IBM-Amex Deal: A RadicalChange In IT Outsourcing,” InformationWeek,March 4, 2002.

Chan, Y. E., Huff, S. L., Copeland, D. G., andBarclay, D. W. “Business Strategic Orientation,Information Systems Strategic Orientation andStrategic Alignment,” Information SystemsResearch (8:2), 1997, pp. 125-150.

Conner, K. R., and Prahalad, C. K. “A Resource-Based Theory of the Firm: Knowledge VersusOpportunism,” Organization Science (7:5),September/October 1996, pp. 477-501.

Corbin, J. M., and Strauss, A. L. “GroundedTheory Research: Procedures, Canons, andEvaluative Criteria,” Qualitative Sociology(13:1), 1990, pp. 3-21.

Currie, W. “The Supply-Side of IT Outsourcing:the Trend Towards Mergers, Acquisitions andJoint Ventures,” International Journal of Physi-cal Distribution & Logistics Management(30:3/4), 2000, pp. 238-254.

Currie, W. “Using Multiple Suppliers to Mitigatethe Risks of IT Outsourcing in Two UKCompanies: ICI and Wessex Water,” Journalof Information Technology (13:3), 1998, pp.169-180.

Dey, I. Grounding Grounded Theory: Guidelinesfor Qualitative Inquiry, Academic Press, SanDiego, CA, 1999.

DiMaggio, P. J., and Powell, W. W. “The IronCage Revisited: Institutional Isomorphism andCollective Rationality in Organizational Fields,”American Sociological Review (48:2), April1983, pp. 147-160.

Earl, M. J. “Limits of IT Outsourcing,” Sloan Man-agement Review (37:3), Spring 1996, pp. 26-32.

Edgeworth, F. Y. Mathematical Physics, KeganPaul, London, 1881.

Eisenhardt, K. M. “Building Theories from CaseStudy Research,” Academy of ManagementReview (14:4), October 1989, pp. 532-550.

Eisenhardt, K. M. “Dynamic Capabilities: Whatare They?,” Strategic Management Journal(21:10/11), October/November 2000, pp. 1105-1121.



Eisenhardt, K. M. “Strategy as Simple Rules,”Harvard Business Review (79:1), January2001, p. 107-116.

Eisenhardt, K. M. “Strategy as Strategic DecisionMaking,” Sloan Management Review (40:3),Spring 1999, pp. 65-72.

Elitzur, R., and Wensley, A. “Game Theory as aTool for Understanding Information ServicesOutsourcing,” Journal of Information Tech-nology (12:1), 1997, pp. 45-60.

Fudenberg, D., and Levine, D. “Maintaing aReputation When Strategies are ImperfectlyObserved,” Review of Economic Studies (59:3),July 1992, pp. 561-579.

Garner, R. “R-E-S-P-E-C-T,” Computerworld,February 2, 1998.

Glaser, B. G., and Strauss, A. L. The Discovery ofGrounded Theory: Strategies for QualitativeResearch, Aldine Publishing Company,Chicago, 1967.

Goles, T. The Impact of Client-Vendor Rela-tionship on Outsourcing Success, UnpublishedPh.D. Dissertation, University of Houston,Houston, TX, 2001.

Grant, R. M. “Toward a Knowledge-Based Theoryof the Firm,” Strategic Management Journal(17), Special Issue on Knowledge and the Firm,Winter 1996, pp. 109-122.

Hamel, G., and Prahalad, C. K. Competing for theFuture, Harvard Business School Press,Boston, 1996.

Hardin, R. Trust and Trustworthiness, RussellSage Foundation, New York, 2002.

Hart, O. D. “An Economist’s Perspective on theTheory of the Firm,” Columbia Law Review(89:7), 1989, pp. 1757-1774.

Hart, O. D., and Moore, J. “Property Rights andthe Nature of the Firm,” Journal of PoliticalEconomy (98:6), December 1990, pp. 1119-1158.

Henderson, J. C., and Venkatraman, N. “StrategicAlignment: A Model for Organizational Trans-formation via Information Technology,” WorkingPaper 3223-90, Sloan School of Management,Center for Information Systems Research,Massachusetts Institute of Technology, 1990.

Herbig, P., Milewicz, J., and Golden, J. “A Modelof Reputation Building and Destruction,”Journal of Business Research (31:1),September 1994, pp. 23-31.

Hirschheim, R. A., and Lacity, M. C. “The Mythsand Realities of Information TechnologyInsourcing,” Communications of the ACM(43:2), February 2000, pp. 99-107.

Hitt, L. M., and Brynjolfsson, E. “InformationTechnology and Internal Firm Organization: AnExploratory Analysis,” Journal of ManagementInformation Systems (14:2), Fall 1997, pp. 81-101.

Hu, Q., Saunders, C., and Gebelt, M. “ResearchReport: Diffusion of Informaiton Systems Out-sourcing: A Reevaluation of InfluenceSources,” Information Systems Research (8:3),1997, pp. 288-301.

Ichniowski, C., Shaw, K., and Prennushi, G. “TheEffects of Human Resource ManagementPractices on Productivity,” American EconomicReview (87:3), 1997, pp. 291-313.

Kern, T. “The Gestalt of an Information Tech-nology Outsourcing Relationship: An Explora-tory Analysis,” in Proceedings of the EighteenthInternational Conference on InformationSystems, K. Kumar and J. I. DeGross (eds.),Atlanta, GA, 1997, p. 37-58.

Kern, T., and Willcocks, L. The Relationship Ad-vantage: Information Technologies, Sourcing,and Management, Oxford University Press,Oxford, UK, 2001.

Koh, C., Tay, C., and Ang, S. “Managing Vendor-Client Expectations in IT Outsourcing: APsychological Contract Perspective,” in Pro-ceedings of the Twentieth International Con-ference on Information Systems, P. De and J.I. DeGross (eds.), Charlotte, NC, 1999, pp.512-518.

Lacity, M. C., and Hirschheim, R. A. InformationSystems Outsourcing: Myths, Metaphors, andRealities, J. Wiley, Chichester, UK, 1993.

Lacity, M. C., and Willcocks, L. P. “An EmpiricalInvestigation of Information TechnologySourcing Practices: Lessons From Experi-ence,” MIS Quarterly (22:3), September 1998,pp. 363-408.

Lacity, M. C., Willcocks, L. P., and Feeny, D. F.“IT Outsourcing: Maximize Flexibility and Con-trol,” Harvard Business Review (73:3), May-June 1995, pp. 84-93.

Lee, J. N., and Kim, Y. G. “The Impact of Knowl-edge Sharing, Organizational Capability and



Partnership Quality on IS Outsourcing Suc-cess,” Journal of Management InformationSystems (15:4), 1999, pp. 29-61.

Lee, P. C. B. “Turnover of Information Techno-logy Professionals: A Contextual Model,”Accounting, Management and InformationTechnologies (10:2), April 2000, pp. 101-124.

Lee, P. C. B., Ang, S., and Slaughter, S. “TurningOver and Away of Information Systems Profes-sionals,” in Proceedings of Eighteenth Inter-national Conference on Information Systems,K. Kumar and J. I. DeGross (eds.), Atlanta, GA,1997, pp. 470-471.

Loh, L., and Venkatraman, N. “Determinants ofInformation Technology Outsourcing: A Cross-Sectional Analysis,” Journal of ManagementInformation Systems (9:1), Summer 1992a, pp.7-24.

Loh, L., and Venkatraman, N. “Diffusion of Infor-mation Technology Outsourcing: InfluenceSources and the Kodak Effect,” InformationSystems Research (3:4), December 1992b, pp.334-378.

Low, J. Invisible Advantage: From Innovation toReputation—How Intangibles Are Driving Busi-ness Performance, Perseus Publications,Cambridge, MA, 2002.

Makadok, R. “Toward a Synthesis of theResource-Based and Dynamic-CapabilityViews of Rent Creation,” Strategic ManagementJournal (22:5), May 2001, p. 387-402.

Mason, P. “Report Shows ‘Two Nations’ of itSpend Health Services Lag Behind GeneralGrowth,” Computer Weekly, March 2, 2000, p.19.

McDonnell, A., and Lichtenstein, Y. “Pricing Soft-ware Development Services,” UnpublishedPaper, University College Dublin, 2002 (avail-able online at http://mis.ucd.ie/staff/Ylichtenstein/alan.doc; accessed March 10,2003).

McDougall, P. “New Deals May Be Start of Some-thing Big,” Information Week, November 18,2002 (available online at http://www.informationweek.com/story/IWK20021115S0027).

McFarlan, F. W., and Nolan, R. L. “How to Man-age an IT Outsourcing Alliance,” Sloan Man-agement Review (36:2), Winter 1995, pp. 9-23.

McMurtrey, M. E., Grover, V., Teng, J. T. C., andLightner, N. J. “Job Satisfaction of InformationTechnology Workers: The Impact of CareerOrientation and Task Automation in a CASEEnvironment,” Journal of Management Infor-mation Systems (19:2), Fall 2002, pp. 273-302.

Miles, M. B., and Huberman, A. M. QualitativeData Analysis: A Sourcebook of New Methods,Sage Publications, Beverly Hills, CA, 1984.

Milgrom, P. R., and Roberts, J. “Complementa-rities and Fit: Strategy, Structure, and Organi-zational Change in Manufacturing,” Journal ofAccounting & Economics (19:2/3), March-May1995, pp. 179-208.

Milgrom, P. R., and Roberts, J. “The Economicsof Modern Manufacturing: Technology, Stra-tegy, and Organization,” The American Econo-mic Review (80:3), June 1990, pp. 511-528.

Milgrom, P. R., and Roberts, J. Economics,Organization, and Management, Prentice-Hall,Englewood Cliffs, NJ, 1992.

Milgrom, P. R., Qian, Y., and Roberts, J.“Complementarities, Momentum, and theEvolution of Modern Manufacturing,” TheAmerican Economic Review (81:2), May 1991,pp. 84-88.

Moore, J. E., and Burke, L. A. “How to TurnAround ‘Turnover Culture’ in IT,” Communi-cations of the ACM (45:2), February 2002, pp.73-78.

Myers, M. “Qualitative Research in InformationSystems,” MIS Quarterly Discovery, 1997(available online at http://www.misq.org/misqd961/isworld).

Nam, K., Rajagopalan, S., Rao, H. R., andChaudhury, A. “A Two-Level Investigation ofIn formation Systems Outsourc ing, ”Communications of the ACM (39:7), July 1996,pp. 36-44.

Nelson, P., Richmond, W., and Seidmann, A.“Two Dimensions of Software Acquisition,”Communications of the ACM (39:7), July 1996,pp. 29-35.

Orlikowski, W. J. “CASE Tools as OrganizationalChange: Investigating Incremental and RadicalChanges in Systems Development,” MIS Quar-terly (17:3), September 1993, pp. 309-340.

Paulk, M., Curtis, B., Chrissis, M., and Weber, C.“Capability Maturity Model, Version 1.1,” IEEESoftware (10:4), July 1993, pp. 18-27.



Pettigrew, A. M. “Longitudinal Field Research onChange: Theory and Practice,” OrganizationScience (1:3), August 1990, pp. 267-292.

Poppo, L. “Do Formal Contracts and RelationalGovernance Function as Substitutes or Com-plements?,” Strategic Management Journal(23:8), August 2002, pp. 707-725.

Quinn, J. B. “Strategic Outsourcing: LeveragingKnowledge Capabilities,” Sloan ManagementReview (40:4), 1999, pp. 9-21.

Richmond, W. B., and Seidmann, A. “SoftwareDevelopment Outsourcing Contract: Structureand Business Value,” Journal of ManagementInformation Systems (10:1), Summer 1993, pp.57-72.

Richmond, W. B., Seidmann, A., and Whinston, A.B. “Incomplete Contracting Issues in Informa-tion-Systems Development Outsourcing,”Decision Support Systems (8:5), September1992, pp. 459-477.

Ross, J. W., Vitale, M., and Beath, C. M. “TheUntapped Potential of IT Chargeback,” MISQuarterly (23:2), June 1999, 215-237.

Sabherwal, R. “The Role of Trust in OutsourcedIS Development Projects,” Communications ofthe ACM (42:2), February 1999, pp. 80-86.

Sambamurthy, V., and Zmud, R. W. “Arrange-ments for Information Technology Governance:A Theory of Multiple Contingencies,” MISQuarterly (23:2), June 1999, pp. 261-290.

Samuelson, P. A. “Complementarity,” Journal ofEconomic Literature (12:4), December 1974,pp. 1255-1289.

Saunders, C., Gebelt, M., and Hu, Q. “AchievingSuccess in Information Systems Outsourcing,”California Management Review (39:2), Winter1997, pp. 63-79.

Scheier, R. L. “Businesses Outsourcing More, ButLess Thrilled with Results,” Computerworld,July 21, 1997.

Slaughter, S., and Ang, S. “Employment Out-sourcing in Information Systems,” Communica-tions of the ACM (39:7), July 1996, pp. 47-54.

Snir, E., and Hitt, L. M. “Vendor Screening inInformation Technology Contracting with a Pilot

Project,” Unpublished Manuscript, The WhartonSchool, University of Pennsylvania, 2002 (avail-able online at http://tecom.cox.smu.edu/esnir/Screen.pdf; accessed March 27, 2003).

Topkis, D. M. “Minimizing a Submodular Functionon a Lattice,” Operations Research (26:2),March/April 1978, pp. 305-321.

Useem, M., and Harder, J. “Leading Laterally inCompany Outsourcing,” Sloan ManagementReview (41:2), 2000, pp. 25-36.

Walker, G., and Weber, D. “A Transaction CostApproach to Make-or-Buy Decisions,” Administ-rative Science Quarterly (29:3), September1984, pp. 373-391.

Weiss, A. M., Anderson, E., and Maclnnis, D. J.“Reputation Management as a Motivation forSales Structure Decisions,” Journal of Mar-keting (63:4), October 1999, pp. 74-89.

Wernerfelt, B. “A Resource-Based View of theFirm,” Strategic Management Journal (5:2),1984, pp. 171-180.

Whang, S. “Contracting for Software Develop-ment,” Management Science (38:3), 1992, pp.307-324.

Willcocks, L., and Kern, T. “IT Outsourcing asStrategic Partnership: The Case of the UKInland Revenue,” European Journal of Infor-mation Systems (7:1), 1998, pp. 29-45.

Willcocks, L., and Lacity, M. C. “Relationships inIT Outsourcing: A Stakholder Perspective,” inFraming the Domains of IT Management, R.Zmud (ed.), Pinnaflex Inc., Cincinnati, OH,2000, pp. 355-384.

Williamson, O. E. The Economic Institutions ofCapitalism, The Free Press, New York, 1985.

Williamson, O. E. “Transaction Cost Economics:The Governance of Contractual Relations,”Journal of Law and Economics (22:2), 1979,pp. 232-262.

Wilson, R. “Reputations in Games and Markets,”in Game-Theoretic Models of Bargaining, A.Roth (ed.), Camridge University Press, NewYork, 1985, pp. 27-62.

Yin, R. K. Case Study Research: Design andMethods, Sage Publications, Beverly Hills, CA,1984.



About the Authors

Natalia Levina is an assistant professor in theInformation, Operations, and Management Sci-ences Department at the New York UniversityStern School of Business. Her current researchinterests include understanding collaborationpractices on multiparty systems developmentprojects, knowledge management in IT servicedelivery organizations, IT outsourcing strategy andimplementation, and technology design principlesin heterogeneous environments. Her work hasappeared in Applied Intelligence and Reflections:The SoL Journal. Natalia has a Ph.D. in Informa-tion Technology from the Massachusetts Instituteof Technology and a Master’s in Mathematics fromBoston University.

Jeanne W. Ross is a principal research scientistat the Massachusetts Institute of Technology’sCenter for Information Systems Research whereshe conducts research, lectures, and directsexecutive education courses on IT managementpractices. Her research focuses on the manage-ment and use of information technology inorganizations. Recent work has examined the lifecycle of IT architectures, the impacts of enterprisesystems on organizational processes, and thedevelopment of IT infrastructures for e-business.Her work has appeared in Sloan ManagementReview, MIS Quarterly, Journal of ManagementInformation Systems, and Harvard BusinessReview. Jeanne has an MBA from The WhartonSchool at the University of Pennsylvania and aPh.D. in Management Information Systems fromthe University of Wisconsin-Milwaukee.

Control of Varietyand Mutlitude of

Projects

Methodology

Junior Staff

StaffRotation

Promotion fromWithin

Training

Team-basedEnvironment

EmployeeSatisfaction

ReducedTurnover

Productivity

Sense ofOwnership

Employee CareerDevelopment

CustomerRelationshipManagement

PersonnelDevelopment

Office

enables

prepares

provides shielding for

results in

increases

enables

increases through best practice dissemination

enablessupports

enables

increases

increases

increases through flexibility and friendliness

creates

helps evaluate for

increases

allows for skilldiversification in

Profitabilityleads to

WorkDocumentation

ensures

increases

builds reputation that leads to

increases

empowers

managers

enriches

Control of Varietyand Mutlitude of

Projects

Methodology

Junior Staff

StaffRotation

Promotion fromWithin

Training

Team-basedEnvironment

EmployeeSatisfaction

ReducedTurnover

Productivity

Sense ofOwnership

Employee CareerDevelopment

CustomerRelationshipManagement

PersonnelDevelopment

Office

enables

prepares

provides shielding for

results in

increases

enables

increases through best practice dissemination

enablessupports

enables

increases

increases

increases through flexibility and friendliness

creates

helps evaluate for

increases

allows for skilldiversification in

Profitabilityleads to

WorkDocumentation

ensures

increases

builds reputation that leads to

increases

empowers

managers

enriches

Appendix AVendor’s Using Control Over Projects to Develop and Use Competencies



Appendix B

A Discussion of Complementarity inOrganizational Design Concept

While the economic concept of complementarity among goods in the consumption function dates back toEdgeworth (1881) and Samuelson (1974), the concept of complementarity in organizational design isrelatively new. This concept provides a powerful mechanism for defining the relationships between inputsand outputs when, as is usually the case, a production function is unknown. It focuses on mutuallyreinforcing pairs or bundles of factors. The impact of one factor is complementary with another factor ifits impact on outcomes is amplified by the application of one or more other factors. Economic notion ofcomplementarity in organizational design holds that when all inputs demonstrate pair-wise comple-mentarities (i.e., they are mutually reinforcing), the whole system is complementary. This theory is usefulin identifying pairs or sets of inputs that work in tandem to produce outcomes.

Although similar, this theory is distinct from the core competency argument. While both concepts focuson a positive feedback among inputs and outputs of the production function, the core competencyargument does not focus directly on the pairs or sets of competencies. In their book on the topic Hameland Prahalad (1996) talk about “bundles of complementary skills,” but they do not formalize the notion ofcomplementarity. While complementarities theory derives from economic arguments, the core competencyargument is based on learning theories. It says that firms build their skills from experiential learning, (i.e.,doing something repeatedly) as in economies of scale and scope: firms build capabilities through exper-ience. For example, if a soccer team plays many games with many different opponents, the team can builda unique style and competence: a core competency in playing soccer. In short, core competency theoryrefers to the ability to generate desired outcomes by enhancing an input—a specific core competence.

The complementarities argument is also distinct from the concept of fit, which considers exogenous factors,such as market demands. Strategic alignment is the process of fitting a firm’s IT capabilities to thedemands of the marketplace. This alignment is critical to business success. Marketplace demands, how-ever, are not controllable by firms. Since the theory of complementarity in organizational design isconcerned only with endogenous (i.e., controllable) factors, it is not concerned with fit. Certainly, long-termsuccess demands that a firm’s complementarities do, indeed, fit with customer needs, but the fit betweenorganizational capabilities and marketplace demands is not germane to the theory.

Soccer can also provide an example for the discussion of complementarities. A soccer player whopractices his skills for many hours a day should become a stronger player (this is a basic core competencyargument). If he then practices with his teammates and learns his role within the team, his skills willgenerate value for the team. However, if the player fails to play position properly, the impact of hisimproving skills in a game situation will likely be quite limited. In fact, a player with improving skills but poorteamwork may disrupt his team (by, for example, stealing the ball from his teammates) and thus, as hepersonally improves, the team’s performance worsens. Thus, the individual player’s skill and his ability toplay position are complementary input into the team’s performance. The theory of complementaritiesargument is that it makes more sense for the player to split his practice time between improving his positionplaying skills and his individual skills rather than on one or the other alone. Note that this player’s perfor-mance, and that of his team, might be affected by weather, field conditions, the skills of the opponents, andthe competence of the referees, but those are exogenous factors and thus are not complementary with the



player’s skills. They affect both the player’s ability to leverage his skills and, conceivably, the performanceof the team, but they are not complementary with either.

Hitt and Brynjolfsson (1997) provide an example of the complementarities concept in an IS setting. Theyobserved that increasing IT investments generate higher returns when accompanied by changes in threeother factors: (1) distribution of decision authority, (2) human capital investments, and (3) performance-based incentives. Because these four factors are complementary, firms that invested in IT and adjustedtheir organizational structure and processes performed better than firms that invested in IT withoutchanging complementary practices. In contrast, a core competency argument would be that gaining ITdeployment competence could improve firm performance. However, core competency theory would failto fully explain the phenomenon of what is involved in gaining such competence. It is noteworthy thatBrynolfsson and Hitt do not discuss the fit between IT strategy and firm strategy; this would be analignment argument.

Documents

Levina & Ross/The Vendor’s Value Proposition in IT Outsourcingpeople.stern.nyu.edu/nlevina/Papers/From the Vendors Perspective.pdf · value proposition. While the client’s sourcing