Kt

Sa

b

c

a

KKKSIO

1

ofitfi2caaesttts

0d

International Journal of Information Management 29 (2009) 342–352

Contents lists available at ScienceDirect

International Journal of Information Management

journa l homepage: www.e lsev ier .com/ locate / i j in fomgt

nowledge transfer processes in IT outsourcing relationships andheir impact on shared knowledge and outsourcing performance

tefan Blumenberga,∗, Heinz-Theo Wagnerb, Daniel Beimbornc

E-Finance Lab/Institute for Information Systems, Goethe University Frankfurt, Grüneburgplatz 1, 60323 Frankfurt, Germanyheilbronn business school, Bahnhofstrasse 1, 74072 Heilbronn, GermanyChair of Information Systems and Services, University of Bamberg, Feldkirchenstr. 21, 96049 Bamberg, Germany

r t i c l e i n f o

eywords:nowledge-based theorynowledge transferhared knowledgeT outsourcing relationshiputsourcing performance

a b s t r a c t

What is the impact of specific knowledge-transfer processes on the level of shared knowledge and, inturn, on outsourcing performance in outsourcing relationships? Drawing on a series of case studies cov-ering IT providers and banks, we investigate several applied knowledge-transfer processes dedicated tothe transfer of explicit or tacit knowledge between outsourcing banks and their providers. We exam-ine the differential influence of various types of knowledge transfer on shared knowledge between theparties and on the resulting outsourcing performance. Results depict the differential impact of variousknowledge-transfer processes dedicated to the transfer of explicit, or tacit knowledge, respectively, on thedevelopment of shared knowledge. Interestingly, the combination of both knowledge-transfer processesdedicated to the transfer of explicit knowledge and those dedicated to the transfer of tacit knowledgeproves to be most effective. Furthermore the results indicate that high levels of shared knowledge posi-

tively influence outsourcing performance. In addition to previous literature, we found transfer processesfor explicit knowledge in an outsourcing context to consist of two dimensions: The content dimension, pri-marily focused on in literature, and the sender–receiver dimension of transfer processes which are rarelyaddressed in outsourcing literature. The content dimension embraces mechanisms such as trainings, SLAsand standards that define how content has to be interpreted, whereas the sender–receiver dimension oftransfer processes of explicit knowledge defines explicit, documented interaction structures between parties.

. Introduction

During the last decades of IS research, a vast amount of workn knowledge has accumulated and proved to be important forrms. Knowledge allows firms to add value, and it is argued thathe ability to generate knowledge is at the core of the theory of therm (Amit & Schoemaker, 1993; Hitt, Bierman, Shimizu, & Kochhar,001; Peppard & Ward, 2004) and that knowledge is the most criti-al asset of the firm (Grant, 1996). This importance is highlighted byrecent study concerning supply chain flexibility. Gosain, Malhotra,nd El Sawy (2004) found that deep coordination-related knowl-dge was the single most important factor positively influencingupply chain flexibility. This knowledge was far more important

han data connectivity and modularity issues, or the standardiza-ion of process and content interfaces. In their conclusion regardinghe importance of knowledge, they stated: “This shows that thetate of an enterprise’s knowledge relevant to its sensing and adap-

∗ Corresponding author. Tel.: +49 69 798 34001; fax: +49 69 798 33910.E-mail address: blumenberg@wiwi.uni-frankfurt.de (S. Blumenberg).

268-4012/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.oi:10.1016/j.ijinfomgt.2008.11.004

© 2008 Elsevier Ltd. All rights reserved.

tation capabilities for coordination should be assessed by referenceto underlying mechanisms by which knowledge is acquired, con-textualized, integrated, maintained, retrieved, and used” (Gosainet al., 2004, p. 32).

Knowledge, defined as justified true belief (Nonaka, 1994), isthe core of the knowledge-based theory (KBT) (Grant, 1996), orknowledge-based perspective, that builds upon the resource-basedview (RBV) (Alavi & Leidner, 2001). The knowledge-based theoryviews “the firm as a dynamic, evolving, quasi-autonomous systemof knowledge production and application” (Spender, 1996, p. 59).This perspective contends that knowledge is the principal resourceof firms and that production requires the integration of a broadrange of knowledge (Grant, 1996). To develop this principal resourceNickerson and Zenger (2004) suggest that the knowledge stock canbe expanded by acquiring or absorbing knowledge from outside thefirm or by generating new knowledge by, first, the identification of

a problem and, second, the discovery of a valuable solution.

In particular, knowledge is a crucial factor in IT outsourcing deci-sions. IT outsourcing is defined as the “the handing over to a thirdparty management of IT/IS assets, resources, and/or activities forrequired results” (Willcocks & Kern, 1998). Organizations outsource

of Info

I&e(katptpksTatpipp

rl

•

•

kEgoi

sbpe

slwcpc

2

pvweKw

ffTwrit

S. Blumenberg et al. / International Journal

T activities that are not regarded as their core competencies (FeenyWillcocks, 1998). Therefore, they are acquiring external knowl-

dge that has to be integrated into their routines and processesDibbern, Goles, Hirschheim, & Jayatilaka, 2004). This transfer ofnowledge between outsourcers and their providers is two-sideds, on the one hand, knowledge is transferred from the providero the outsourcer (technology-specific knowledge regarding, e.g.rovided services) and, on the other hand, from the outsourcer tohe provider (business-specific knowledge regarding processes androcedures) (Quinn, 1999). In both cases the goal is to increase thenowledge of the other’s knowledge domain, that is, to increasehared knowledge through these knowledge-transfer processes.he importance of knowledge transfer in IT outsourcing becomespparent when examining the outsourcing lifecycle. Knowledgeransfer between both parties is crucial in the pre-outsourcinghase, during which vendors are selected and contracts are crafted,

n the transition phase in which services are transferred to therovider, and has to be sustained over the years of the deliveryhase (Dibbern et al., 2004).

Combining research on designing an effective IT outsourcingelationship and on the knowledge-based perspective, we formu-ate the following research questions:

First, how do different types of knowledge-transfer processesinfluence the level of shared knowledge between the partiesinvolved in an IT outsourcing relationship?Second, how does shared knowledge affect the performance ofthe IT outsourcing relationship?

In this context, we examine, in particular, the process in whichnowledge is transferred and integrated, and follow the caveat ofisenhardt and Santos (2002, p. 160), who state: “More focus shouldo to knowledge integration processes, in which the developmentf meaning and the creation of new knowledge occurs throughndividual interactions and is affected by social contexts.”

To answer the research questions, we have conducted a casetudy series that surveys the IT outsourcing relationships of 12anks and their IT providers. Different applied knowledge-transferrocesses are analyzed regarding their influence on shared knowl-dge and outsourcing performance.

The remainder of this paper is structured as follows: the nextection provides an overview of the theoretical foundation, fol-owed by our research model and the hypotheses that guide our

ork in the subsequent section. The third section describes thease study setting and methodology. Afterwards, the results areresented. In the final section, limitations are discussed and con-lusions are drawn.

. Theoretical foundation and research model

Knowledge, in the interpretation of Alavi and Leidner (2001), isersonalized information related to facts, judgments, ideas, obser-ations, etc. For personalized information, it is not importanthether the information is accurate, new, unique, or useful. Knowl-

dge, in this view, results from the cognitive processing of stimuli.nowledge is also defined as justified true belief by Nonaka (1994),ho distinguishes between explicit and tacit knowledge.

Explicit knowledge can be articulated, codified, and easily trans-erred. Thus explicit or codified knowledge “is transmittable inormal, systematic language” (Nonaka, 1994)(Nonaka, 1994, p. 16).

herefore, it can be defined as “articulated, generalized knowledge,”hich is, for example, the “knowledge of major customers in a

egion” (Alavi & Leidner, 2001, p. 113). In contrast, tacit knowledges linked to the individual and is very difficult, or even impossible,o articulate. Only through observation and doing something first-

rmation Management 29 (2009) 342–352 343

hand is it possible to learn this type of knowledge. As knowledgeis explored, put into action and socially justified, only some part ofit may be codified (made more explicit), by being converted intomessages that can then be processed as information and transmit-ted” (Eisenhardt & Santos, 2002, p. 140). Thus, “tacit knowledge isthe information that has been processed in the minds of individu-als through deliberation, learning, and judgment” (Pavlou, Housel,Rodgers, & Jansen, 2005, p. 208) and therefore this type of “knowl-edge is rooted in actions, experience, and involvement in specificcontext” (Alavi & Leidner, 2001, p. 113). An example of tacit knowl-edge is the “best means of dealing with specific customer” (Alavi& Leidner, 2001, p. 113). Tacit knowledge is partially embeddedin individuals and partially in collaborative relationships (Hitt etal., 2001) and encompasses a cognitive and a technical element(Nonaka, 1994). Technical refers to skills and know-how whereascognitive refers to mental models of an individual. Research oftenfocuses on tacit knowledge as a means of achieving a competitiveadvantage (see e.g. Hitt et al., 2001). However, this should not leadto the assumption that tacit knowledge is more valuable, becausetacit and explicit knowledge are mutually dependent and reinforc-ing, or as Alavi and Leidner explain: “Tacit knowledge forms thebackground necessary for assigning the structure to develop andinterpret explicit knowledge” (Alavi & Leidner, 2001, p. 112).

The generation of new knowledge takes place in individuals.Based on generic knowledge and skills that can be described ver-bally (King & Zeithaml, 2003), generic knowledge and skills areenriched by practice and lead to an increased recognition of unre-alized patterns, improved abilities to network knowledge domains,and lower levels of cognitive involvement. This so-called procedu-ral stage of knowledge generation refers to developing know-howand is inherently engaged in accumulating practical skills throughlearning (Kogut & Zander, 1992). This accumulation of practicalskills also encompasses the move from explicit forms of knowledgeto more tacit forms, which can be understood using the conceptof cognitive processing. Following the interpretation of Alavi andLeidner (2001), cognitive processing transforms information intoknowledge. If knowledge is articulated and thus made explicit itbecomes information (facts, axiomatic propositions, symbols) thatcan be exchanged with other individuals “once the syntactical rulesrequired for deciphering it are known” (Kogut & Zander, 1992, p.386). These individuals, in turn, process this information and trans-form it into knowledge interpreted by their mental models (Nonaka,1994).

Based on such knowledge integration, shared knowledge devel-ops. Sharing a common knowledge base is required for individualscoming to the same understanding of an issue. A shared knowledgebase is essential for coordination among agents who have differ-ent models of the world and do not know the models of others(Foss, 1999). The importance of a shared knowledge base is alsohighlighted in, e.g. several alignment studies and studies of IT per-formance (e.g. Nelson & Cooprider, 1996; Preston & Karahanna,2004; Ray, Muhanna, & Barney, 2005; Reich & Benbasat, 2000),dealing with the interaction of the business domain and the ITdomain. In this context, shared knowledge can be defined as “anunderstanding and appreciation among IS and line managers for thetechnologies and processes that affect their mutual performance”(1996, p. 411). Appreciation means sensitivity to the organizationalenvironment of the other group encompassing goals, constraints,interpretations, and behavior. Thus, shared knowledge forms thebasis for performance gains.

An initial step towards the formation of shared knowledge is

to create a common language. Human actors in the IT and thebusiness domain often speak different technical and procedurallanguages (Keen, 1991). Therefore, the requirements, goals andconstraints articulated by one domain can be perceived as unrea-sonable demands and as uncooperative by the other domain. Thus,

3 of Information Management 29 (2009) 342–352

aaacTib((to

2

atkiki

ftGte

d

ckpdptc

mtbcbckeeropmtf

titkd(fakvgf

Table 1Hypotheses.

Hypotheses Description

H1 Transfer processes of explicit knowledge have apositive impact on the level of shared knowledge.

H2 Transfer processes of tacit knowledge have a positive

44 S. Blumenberg et al. / International Journal

shared language can facilitate knowledge transfer as well as cre-te a positive social influence process (Galunic & Rodan, 1998). Insecond step, a shared knowledge base can be created. Based on a

ommon language and frequent interaction (Galunic & Rodan, 1998;iwana, Bharadwaj, & Sambamurthy, 2003), mutual understand-ng between the IT and the business domain can evolve, alleviatingarriers to cooperation and creating a common frame of referenceNelson & Cooprider, 1996) in which knowledge can be integratedGrant, 1996). Shared domain knowledge does not necessarily refero knowing how to do the other’s job, but requires an understandingf the other’s goals, constraints, and needs.

.1. Knowledge transfer

As discussed in the previous paragraphs, shared knowledge isn outcome of the process of knowledge transfer and the integra-ion of this transferred knowledge by learning and, in turn, sharednowledge influences the efficiency of this process by facilitat-ng the transfer of knowledge (Kogut & Zander, 1992). Thus, thesenowledge-transfer processes are the prerequisite for knowledgentegration that eventually lead to an increase of shared knowledge.

Knowledge-transfer processes are therefore important for theormation of shared knowledge. The term “process” is often usedo refer to explicit structures, but there are also implicit actions.alunic and Rodan (1998) call them tacitly- and explicitly-held rou-

ines to address the fact that routines can contain both tacit andxplicit knowledge.

Depending on the degree of explicitness, we will distinguishifferent types of knowledge-transfer processes.

First, there are knowledge-transfer processes primarily dedi-ated to the transfer of explicit knowledge. In order to transfernowledge it is necessary to be able to interpret the content (com-are Te’eni, 2001). This requires a common frame of reference byefining how content must be interpreted. Typical examples arerocess descriptions, standards, service level agreements, and con-racts designed to clarify terms, procedures, priorities and exchangeonditions.

Second, we can identify personal linkages appropriate for com-unicating more tacit knowledge. New knowledge is generated by

he influx of information into an individual’s mind and is then com-ined with the existing knowledge of this individual and, in turn,ommunicated and made explicit. Existing knowledge is exploitedy granting access to the knowledge of different actors which, whenombined, solves a problem. Nonaka (1994) sees the process ofnowledge creation as a permanent conversion between tacit andxplicit knowledge, requiring social interaction and influenced bynvironmental parameters. Eisenhardt and Santos (2002, p. 142),eferring to Brown and Duguid (1991) highlight “that codificationf work procedures can be quite different from actual workingractices, and sometimes it is even contradictory.” In their view,eaningful knowledge is significantly related to daily work, and

he acquisition of new knowledge (learning) is socially constructedrom working practices.

“Knowledge creation . . . is more dependent on the combina-ion and sharing of tacit knowledge . . . In knowledge creation,nformation exchange is frequently emergent, in that partners tohe exchange are often unable to articulate, a priori, the specificnowledge that they need. This kind of exchange requires veryirect interaction, as the parties grapple with research puzzles”McFadyen & Cannella, 2004, p. 737). This can be attributed to theact that explicit knowledge is “gradually translated, through inter-

ction and a process of trial-and-error, into different aspects of tacitnowledge.” (Nonaka, 1994, p. 20). Therefore, personnel linkage isery important, and was also found to be crucial during the strate-ic business and IS planning processes, to ensure that a commonrame of reference exists between business and IS planners (Teo &

impact on the level of shared knowledge.H3 Shared knowledge has a positive impact on

outsourcing performance.

Ang, 1999). Examples of mechanisms for transferring tacit knowl-edge are communities of practice or the transfer of personnel fromone organizational unit to another.

Knowledge transfer and the development of shared knowl-edge influence organizational performance. In a study regardingIS development capability, Tiwana et al. (2003, p. 251) found that“formal knowledge-transfer mechanisms such as change requestsor revised requirements specifications rarely allow rapid incor-poration of such information.” Santhanam, Seligman, and Kang(2007) found technology support mechanisms (e.g. help desks)to be insufficient for IT knowledge transfer. Rather, face-to-faceinteractions are crucial. Hansen (1999) investigated the sharing ofknowledge and found that, in particular, tacit knowledge can besuccessfully shared if there is a high frequency of interaction andcloseness. Through individual interactions, meaning is developedand knowledge created (Galunic & Rodan, 1998). This knowledgetransfer leads to an increase of shared knowledge that, in turn,may affect performance. Penrose (1959, p. 76) formulates that“. . . it is likely that increases in knowledge can always increasethe range or amount of services available from any resource.” Inseveral studies, shared knowledge was found to affect IT perfor-mance (Nelson & Cooprider, 1996), IT-line partnerships and IT use(Reich & Benbasat, 2000). In the context of IT and drawing on theknowledge-based perspective, delivering business value is essen-tially a set of knowledge-based activities. It involves integrating andcoordinating knowledge from many individuals of different disci-plines and backgrounds, with varied experiences and expectations,located in different parts of the organization (Peppard & Ward,2004). This demands a close partnership between IT units andbusiness units, both in formal processes and informal working rela-tionships whereas the informal structures and processes are moreimportant for the efficiency of partnerships (Chan, 2002). Althoughinteractions between the parties involved in IT outsourcing rela-tionships are of utmost importance to transfer knowledge overinter-organizational boundaries (Goles & Chin, 2005; Quinn, 1999),there is not much research on the influence of knowledge-sharingroutines on shared knowledge and outsourcing performance. Leeand Kim (1999) demonstrated the impact of information sharingon a generic construct of relationship quality in IT outsourcing rela-tionships, but could not show a correlation between informationsharing and business understanding (one of several dimensions ofrelationship quality). In a generic way, Lee (2001) illustrated theinfluence of knowledge sharing on outsourcing success, but did nottake specific transfer processes for knowledge sharing into account.

Based on the theoretical reasoning we derive the followinghypotheses that are described in Table 1.

Fig. 1 shows the research model with hypotheses that guide ourwork. Knowledge-transfer processes are expected to have a positiveimpact on shared knowledge, which, in turn, positively influencesthe outsourcing performance.

3. Approach

This section describes both the environment in which we gath-ered the series of case studies and the used case study methodology.

S. Blumenberg et al. / International Journal of Information Management 29 (2009) 342–352 345

searc

3

tbcwawctbtar

§Bati

3

saoka

•••

iFtWo“

by deriving hypotheses from the literature, discussing the relevancewith the case study partners, tape-recording and transcribing ques-tions and answers, and using the described case study database,which allows us to trace from conclusions back to questions.

Fig. 1. Re

.1. Case study environment

For testing our model, we conducted a series of case studies inhe German-speaking banking industry. Both the German and Swissanking systems consist of three different groups: private commer-ial banks, publicly owned saving banks and credit cooperatives,hich tend to be smaller, but much more frequent in the German

nd Swiss market. Savings banks and commercial banks operateithin clearly specified regional boundaries, thus, there is not much

ompetition within the groups of saving banks or credit coopera-ives. Commercial banks differ from credit cooperatives and savinganks, since they are free to choose their area of operation (mostlyhey operate on a nation-wide or international level) and are usu-lly in direct competition with the latter two, in their respectiveegions.

Overall, due to strict and detailed regulatory requirements (i.e.25a German Banking Act, guidelines from the Basel Committee onanking Supervision), the banks in the German financial industryre quite comparable. For instance, all of these organizations haveo follow strict requirements regarding control of their IT providers,mposed by Basel II regulations.

.2. Case study sample

Our case study approach included the IT outsourcing relation-hip of six German and Swiss banks with their IT providers, as wells the relationships of six IT providers with other banks (except inne case, which we were able to survey a “couple”) in the same mar-et. The comparison of the banks is supported by a very comparablend homogeneous setting:

All banks operate in the same market.All banks are subject to the same strict outsourcing regulations.All interviewees are in the position of provider manager (onbank side) or the key account manager (on provider side) and,therefore, responsible for the relationship that was investigated.Moreover, all of them are very experienced with outsourcing andwere involved throughout all phases of the outsourcing arrange-ment, including the pre-contract procedures.

Further information about the banks/the providers is presentedn Tables 2 and 3. The banks are denoted by F1 through F6, where

means “financial institution.” The providers are denoted by P1hrough P6, with P equal to provider. Based on Fontenot and

ilson’s (1997) prediction matrix of relationship activities, mostf the surveyed relationships are buyer–seller partnerships thatrepresent interdependence between firms in a particular area of

h model.

activity” (Fontenot & Wilson, 1997), except for three of the stud-ied relationships (F3, P3, P4), which can be classified as strategicalliances, where “each [involved organization] commits resourcestoward a strategic purpose that will further benefit each party”(Fontenot & Wilson, 1997). These strategic purposes were, e.g. jointproduct development (F3, P4) or joint marketing efforts. Usually, wespoke with one or two (when available) people at each organization.

3.3. Applied methodology

We tested our proposed research model, covering specificknowledge-transfer processes, shared knowledge and outsourcingperformance, with case studies “to explain the presumed causallinks in real-life interventions that are too complex for a survey orexploratory strategies” (Yin, 2002). To derive sufficient results froma case study approach, these have to be prepared, conducted andanalyzed thoroughly (Yin, 2002). Following Dubé and Paré (2003),we identified the research questions and made them explicit. Thequestionnaire was refined using a pretest and discussed withinthe research community. The interview guideline can be found inAppendix A.

Based on this semi-structured questionnaire, the interviewswere conducted with two researchers. They lasted about 2 h eachand were tape-recorded. The collected data was analyzed usingMAXQDA,1 an instrument for efficiently evaluating quantitativedata. To allow for a better comparison of the data, protocols wereentered in a self-developed case study database (Lee, 1989). Thistool allows cross sectional cuts through the data to, e.g. compareonly a single construct of all or selected cases. We derived theratings of the constructs (see Tables 3 and 4) by comparing the infor-mation regarding every single construct of a certain bank comparedto the overall sample in this paper. As a result, every construct israted as being high, medium or low.

To ensure construct validity (Yin, 2003), we used multiplesources of evidence by carrying-out interviews with key informantsand by using firm reports. Furthermore, the research model wasintroduced to the interviewees after the interview, the discussiontape-recorded and transcribed. A chain of evidence was maintained

1 For further information on MAXQDA see product website: http://www.maxqda.com/products/maxqda.

346 S. Blumenberg et al. / International Journal of Information Management 29 (2009) 342–352

Table 2Banks in the case study series.

F1 F2 F3 F4 F5 F6

Financial sector Commercial bank Commercial bank Credit cooperative Savings bank Savings bank Credit cooperativeEmployees <1000 >10,000 <1000 1000–5000 1000–5000 1000–5000Contract runtime 2005–2010 2003–2013 2000–2010 2004–2009 2004–2009 2004–2008Outsourced service IT infrastructure,

application developmentand maintenance

IT infrastructure,applicationhosting

Business process IT infrastructure IT infrastructure IT infrastructure

Table 3Providers in the case study series.

P1 P2 P3 P4 P5 P6

Employees 1000–5000 1000–5000 >50,000 1000–5000 >50,000 5000–50,000Contract runtime 2006–2009 2002–2009 1999–2013 2000–2010 2001–2012 1994–2010Provided service IT infrastructure Communication

infrastructureIT infrastructure,application

nt

Business process IT infrastructure IT infrastructure, mainframe

N

2wt

Cpia

rmg

1cow

4

bpifpe

managemeumber of contractperiodextensions

0 2 1

To ensure internal validity, we used pattern matching (Yin,003). As highlighted in the following sections, collected citationsere compared with theoretical propositions, to indicate whether

hese propositions can be supported.External validity is related to the generalizability of findings.

ase studies are similar to experiments and can be used to rejectropositions (Lee, 1989; Yin, 2003). While logical generalizabil-

ty is possible with case studies, statistical generalizability is notchievable (Lee & Baskerville, 2003).

For ensuring reliability, we used protocols to guide theesearcher by explicating research objective, propositions, researchodel and research questions, and a case study database as sug-

ested by Yin (2003).Based on the criterion of theoretical saturation (Eisenhardt,

989), no further cases were added, when the latest conductedases did not reveal significant new evidence, compared to the earlynes (Pandit, 1996). Hence, we decided to conclude the researchith the current number of cases.

. Results

In this section, we analyze the knowledge-transfer processesuilt to allow knowledge flows between outsourcers and theirroviders, in the cases investigated. In our case studies, we could

dentify a broad range of mechanisms employed to ensure the trans-er of knowledge between outsourcer and provider. In the followingaragraphs, these processes are discussed and split into transfer ofxplicit or tacit knowledge, respectively.

Table 4Summary of knowledge transfer of banks.

Several times 1 Several times

4.1. Transfer of explicit knowledge

Based on the reasoning above, a common frame of referenceis necessary for the interpretation and integration of knowledge.Accordingly, one important variable investigated in the contingencytheory is organizational maturity, which is the degree to whichorganizational processes have been systematized and formalizedby rules, procedures, and management practices (Raymond, 1990).Organizational maturity, known as formalization, requires organi-zational processes to be well understood. Formalization is related tomore efficiency and involves the application of rules and standardprocedures to reduce ambiguity (Dewett & Jones, 2001), but mayalso lead to reduced flexibility, due to constraints imposed by theserules and standards (Galunic & Rodan, 1998). The prerequisite ofapplying those rules and procedures is to understand the relation-ships of the individual elements and activities of an organizationalprocess. One expression of well-understood processes could be thedegree of up-to-date documentation. What is well understood doesnot include causal ambiguities and can be made explicit (Nonaka,1994), implying that up-to-date documentation is correlated withhigher IS effectiveness and improved process performance (see alsoRaymond, 1990).

Referring to our cases, overall, provider P2 exhibits well-designed and documented processes. For instance, the require-ments process is regarded as one of the best-structured processesby the interviewee. The same applies to a continuous improvementprocess that has been established some years ago and is used on aregular basis within the relationship of P2. The following citationshows that well-documented processes are seen as beneficial to theoutsourcing relationship:

• “A well structured requirements process (e.g. for changes)reduces the amount of conflicts.”2 (P2)

Some of the banks and providers within the case study series

highlight the importance of detailed contracts and Service LevelAgreements (SLAs). For instance, F4 has a very formalistic relation-ship with its provider. A contract with a length of 1500 pages and4 contract renegotiations within 3 years illustrate the efforts that

2 Please note that the original interviews were conducted and transcribed in Ger-man. All citations presented in this paper have been translated by one of the authorsand have been validated by one other author.

of Info

FAnoip–tbtadda

•

•

•

katsmepntftsleattwgtss&

••

•

ctuctefite

d

S. Blumenberg et al. / International Journal

4 invests in improving the contract and in keeping it up-to-date.s a result, F4 regards its contract as “really good.” In turn, F5 isot very content with the outsourcing contract. From F5’s pointf view, some SLAs need major improvements. Nevertheless, thenterviewees pinpoint the importance of the SLAs in governing therovider and proposing that some of their SLAs are well defined andin case of failures – have sufficient explanatory power to illustrate

he importance of their systems to the provider. This is supportedy P5, who renewed and improved its contract recently. The con-ract improvement was required and solved many open issues, interlia the reduction of conflicts due to redesigned escalation proce-ures. The following citations from the cases show that SLAs andocumented escalation procedures are viewed as a means to reducembiguity and to increase understanding.

“SLAs foster mutual understanding. We define the criticality ofour services in SLAs and then the provider employees can under-stand how problematic failures of our systems are.” (F4)“The importance of our banking systems [with regard to recoveryin case of failures] is defined in SLAs.” (F5)“Good escalation procedures reduced the amount of escalationsheavily and keep our communication efficient.” (P5)

Other important facets of the interpretation and integration ofnowledge are standards providing a common frame of referencend definitions of key technical terms. Three providers refer tohe IT Infrastructure Library (ITIL) as an industry standard and aophisticated IT service process reference tool for enhancing com-unication and shared knowledge with their customers. This is

specially interesting in light of the provided services. P2 and P3rovide “only” IT and communication infrastructures, which areot core competencies of their customers. P3 also hosts applica-ions, but its main services are IT infrastructures. P2, a companyocused on communication infrastructures, points out the impor-ance of standards such as ITIL. They are “absolutely required toerve a customer in another industry.” P3 sees a significant corre-ation between communication standards and the communicationfficiency. P3 states that, since the introduction of ITIL in 1993, themount of conflicts decreased abruptly. P6 regards ITIL fundamentalo communicate with its customer. The standardized communica-ion allows – to a certain degree – to debate technical developmentsith the technically non-skilled customer. Thus, ITIL even “miti-

ates missing technical knowledge.” The following citations showhat the provision of a common language is promoted by using atandard, such as ITIL, that also leads to an increase of mutual under-tanding supporting the reasoning of Galunic and Rodan (GalunicRodan, 1998).

“Implementation of ITIL decreased the amount of conflicts.” (P3)“Standardized communication procedures, such as ITIL, enhancethe mutual understanding.” (P2)“We speak the same language because of ITIL based process stan-dardization.” (P6)

While tacit knowledge must be built up through learning pro-esses and is the result of experience, another possibility forransferring explicit knowledge, in addition to standards and doc-ments, is education (Hitt et al., 2001). Depending on the jobharacteristics, employees are required to have a certain educa-ional level, that is, articulable knowledge in their field. Basicducation, often prior to entering the firm, industry-, task-, and

rm-specific training before and during a job, are designed toransfer explicit knowledge in order to reach the desired level ofducation.

F6 experienced trainings to be crucial for improving the sharedomain knowledge of its provider. From the outset of the relation-

rmation Management 29 (2009) 342–352 347

ship, the provider employees “did not have any clue of the financialindustry.” F6 is very experienced in outsourcing and anticipated thislack of knowledge. The provider personnel have been trained in aseries of workshops in the beginning of the relationship to providebasic banking skills. To sustain and develop these skills, the provideremployees are requested to participate in a job shadowing program,and the bank gives regular speeches at the place of the provider. Asa result, F6 evaluates the banking skills as continuously increasingduring the relationship. The bank regards this knowledge as veryvaluable for the provider: “These [skills] can be used by the provideras sales argument in contract extensions or in other outsourcingdeals.” P2, the communication infrastructure provider, believes hisemployees must exhibit sufficient skills to provide good services.Hence, all new provider employees are shown the customer depart-ments, which receive communication services. The effect of thoseefforts is depicted in the following citations:

• “We train the provider employees which gives them an under-standing of our processes and fosters their business skills.” (F6)

• “We give presentations at the provider to enhance the businessunderstanding of its employees.” (F6)

• “Communication structures in terms of regular trainings keep thecommunication effective.” (F6)

• “Our employees have a good business understanding because weare used to show them the customer departments we care for.”(P2)

Besides the previously described transfer processes for explicitknowledge, which provide a common frame of reference, wealso found another mechanism contributing to the transfer ofthe explicit contents discussed in the previous paragraph, beingthe explicit definition of communication partners, between whichcontents have to be transferred. Nickerson and Zenger (2004)investigated hierarchical structures in an intra-organizational con-text and their effects on the creation and transfer of knowledge.Te’eni (2001) researched into organizational communication anddealt with the importance of so-called sender–receiver distance.Nevertheless, to our knowledge, the importance of making com-munication structures explicit for knowledge transfer within aninter-organizational (outsourcing) context has rarely been inves-tigated and empirically addressed.

Correspondingly, we found a set of quotes in our analysis, withregard to elements that have an influence on shared knowledge,that have not been considered in the knowledge-transfer literatureon inter-organizational relationships previously:

• “We train the provider employees which gives them an under-standing of our processes and fosters their business skills.” (F6)

• “Mutual understanding is influenced by a defined provider man-agement as only defined and skilled contact people from thecustomer communicate with the provider employees.” (P1)

• “Clearly defined responsibilities of the employees lead to efficientcommunication with little escalations.” (P4)

• “Finally, after we implemented a liaison unit [requirements man-agement], we [bank and provider] speak the same language.” (F3)

Although provider P4 provides a more business-centered(credit-processing services) service than P1 (IT infrastructure),both experienced that missing structures lead to problem-atic relationships. P1 has agreed upon a relationship with anoutsourcing-inexperienced customer. Main services are desktops

and help desk support, which are a standard product of P1’s port-folio. Despite the recommendations of P1, the customer did notprepare a dedicated provider management from the outset of therelationship, which resulted in a “higgledy-piggledy” communica-tion. Even though the provider has clearly defined contact persons,

3 of Info

tlbfpionctaidi

etisstdtsuigep

raoc

fkT(sb

k

4

eceac1

tlutcfikwtkts

48 S. Blumenberg et al. / International Journal

he customer is not able to interact with the stipulated managementevels. For instance, P1 complains that its account management isurdened by the client with a great number of problems comingrom the operational level. P4 is a bank spin-off and provides credit-rocessing services, a quite novel service in the German financial

ndustry. With regard to its market experience of 7 years, P4 is onef the most experienced providers in its market segment. The con-ection to its customer (F3) is very close. Both companies haveonnections to members of the executive board and have offices inhe same building. Nevertheless, their relationship required manydjustments and learning cycles, concerning the development ofnteraction structures. Finally, at the time of the interview, botheveloped a satisfactory set of structures which they believe excel

n some areas and require improvements in others.In contrast to the previously described transfer processes of

xplicit knowledge, which are required for the transfer of content,he above citations refer to a clear definition of sender and receivern an IT outsourcing relationship as being required. It should betated that the complete and explicit definition of communicationtructures, including assigned responsibilities for the strategic andhe operational level of an outsourcing service, cannot be always beefined completely ex ante. Of course, there are some communica-ion structures defined in almost every contract. However, if thosetructures are incomplete (especially at the operational level) ornclear (e.g. responsibilities not clarified), they will be built up in

nformal ways, with the effects mentioned in the previous para-raph. As these structures are explicit elements for guiding thexchange of knowledge between organizations, we regard them asart of transfer processes for explicit knowledge.

The above citations show that communication efficiency iselated to a defined interaction structure, as indicated by Nickersonnd Zenger (2004), discussing hierarchical structures. In particular,ur case study partners view structure as beneficial for communi-ation in an inter-organizational context.

Therefore, as an extension of the previously held view on trans-erring explicit knowledge, we pose transfer processes for explicitnowledge in an outsourcing context to consist of two dimensions:he content dimension defines how content has to be interpretede.g. with trainings, communication standards, SLAs), whereas theender–receiver dimension of transfer defines interaction structuresetween parties.

In the following, we will apply this described explicitnowledge-transfer process construct with its two dimensions.

.2. Transfer of tacit knowledge

In contrast to explicit knowledge, tacit knowledge is formed byxperience. While applying employees’ explicit knowledge in theontext of their jobs, considerable tacit knowledge is acquired. Forxample, this refers to firm-specific knowledge regarding the oper-tional tasks or the knowledge of the specificities of customers andolleagues and the social networks implied (Nahapiet & Ghoshal,998).

As discussed in the theory section, personnel linkages are impor-ant for the transfer of tacit knowledge. These patterns of personnelinkages are found in many of our cases. Many outsourcers aresed to transfer staff, in the beginning of the outsourcing rela-ionship, to their provider. These transferred employees work inlose cooperation with “old” provider employees to serve the clientrm. However, none of the banks transferred staff for the sake ofnowledge transfer. Their motives were mainly cost-driven, as they

anted to realize cost savings. Interestingly, these banks consis-

ently report the transfer of staff to have a positive impact on sharednowledge, even though the gains in skills were attained “acciden-ally.” The following citations from the case studies highlight theituation.

rmation Management 29 (2009) 342–352

• “The provider employees who previously worked for our ITdepartment and changed to the provider in the beginning of theoutsourcing relationship, exhibit much better banking skills thantheir colleagues” (F1)

• “Transfer of staff [from the bank to the provider] in the beginningof the relationship has a positive impact on mutual understand-ing” (P1)

• “We usually reach consensus quickly with former bank employeesthat changed to the provider in the beginning of the relationship.”(F1, F4)

• “Provider employees that changed from the bank to the providerhave good banking skills.” (F4)

• “Provider employees that changed from us to the provider inthe beginning of the outsourcing relationship, exhibit excellentbanking skills” (F2)

• “We transferred a lot of staff to the provider in the beginning ifthe relationship. These guys have good banking skills.” (F4)

As mentioned above, new knowledge can be regarded associally constructed. “This social construction of knowledge occurswithin informal communities-of-practice, where knowledge isfreely shared through collaborative mechanisms such as narrationand joint work” (Eisenhardt & Santos, 2002, p. 142). By exploit-ing what has been learned in previous jobs, employees transferknowledge from one organizational unit to another “by exchanginginformation and advice through informal communities of practice”(Haas & Hansen, 2005, p. 2). Joint work proves to be an efficientmeans for the transfer of tacit knowledge.

The relationship of P3 is the longest lasting in the surveyed sam-ple. P3 and its customer have a very trustful relationship, withstrong personal bonds. P3’s customer has, e.g. a last call optionand can always submit a second quote after other competitors. Thisstrong involvement of both parties in this relationship is demon-strated by their close collaboration on a daily basis. P3 and itscustomer always jointly develop business blueprints and businesscases together.

P4 is in a comparable situation to P3. As a spin-off from its cus-tomer, the integration between both is quite close. Both companiesstill have personal links on the board level. On the operational level,the collaboration becomes apparent in their mixed workgroups. P4provides employees, who work together with a superior from thecustomer within a joint team that has no functional segregation.

P2 has not only a less close relationship with its customer,but also points out the importance of business understandingfor a successful service. This may sound counter-intuitive, asP2 provides “only” communication infrastructure. However, theinterviewee highlights the importance of understanding, e.g. thecomplex regulatory requirements in German financial law, whichraises questions, like “how to deal with lost cellular phones oftraders according to German financial law” or “what is the impactof network breakdown during trading times.” Due to communi-cation problems during the relationship, the provider successfullyinsisted on involving its customer in the process of demand man-agement, which reduced the processing time of customer requests.The quotes from the case studies illustrate the impact of transfer oftacit knowledge on shared knowledge:

• “Joint development of business blueprints leads to high mutualunderstanding.” (P3)

• “Communication is more open and effective since we are jointly

performing demand management with our customer.” (P2)

• “We are performing the outsourced activity together with ourcustomer in mixed workgroups. These groups consist of our staffand one superior from our customer, who is totally integrated inthe process.” (P4)

S. Blumenberg et al. / International Journal of Info

Table 5Summary of knowledge transfer of providers.

Provider Knowledge-transfer process Shared knowledge

Explicit 1 Explicit 2 Tacit Professionalexperience

Own perception

P1 X X Medium HighP2 X X X Medium HighP3 X X Medium HighPPP

4p

c

ciaKawkkaPosTrcT

•

•

•

•

•

mttdfmAtvq

pcpo

t

ees are evaluated as having less banking knowledge. Therefore, F4

4 X X High High5 X @ Medium High6 X, @ Medium Medium

.3. Developing a chain of evidence of knowledge-transferrocesses on shared knowledge and on performance

What are the effects of the depicted knowledge-transfer pro-esses on shared knowledge and on outsourcing performance?

The indicated chain of arguments from knowledge-transfer pro-esses via shared knowledge to performance measures is laid downn the following tables. The quotes from the case studies describedbove are summarized in Table 4 (banks) and Table 5 (providers).nowledge-transfer processes are differentiated according thebove-presented classification in the transfer of explicit knowledge,ith respect to sender and receiver (Explicit 1), transfer of explicit

nowledge, with respect to content (Explicit 2), and transfer of tacitnowledge (Tacit). A cross marks the occurrence of quotes from thebove-described case studies on knowledge sharing processes (e.g.2 works together with its customer in joint workgroups (transferf tacit knowledge), which is represented by a cross on the inter-ection of row “P2” and column “Knowledge Transfer Processes,acit”). An “@” takes into account further information, like annualeports, or press reports handed out by the interviewee, whichontains relevant information about knowledge-transfer processes.ables 4 and 5 are complemented by this additional information.

This additional information is summarized in the following:

F2 provided internal documents that prove the application of ITILin the daily communication together with the provider.A press report proves the transfer of staff from F3 to their providerin the beginning of the relationship.The SLA’s in the outsourcing contract of F6 are well defined onthe process level and instruct the provider on the relevance of itscustomer’s systems.P5 applies job rotation to let employees experience the customer’sprocesses first-hand.P6 occasionally trains the customer employees in technical skills.

Furthermore, both tables contain the level of shared knowledge,easured in terms of the general banking skills of the employees,

he ability of the provider to understand business requirements andhe level of application of a shared language (see Appendix A foretails). The last column of Table 4 presents the outsourcing per-ormance, measured in terms of the service quality assessed by the

anager in charge of the outsourcing deal, at the respective bank.ll efforts of a bank in an outsourcing relationship are geared toward

he achievement of superior IT service quality in effectively pro-iding banking services to customers. Hence, the achieved serviceuality is used as a measure for outsourcing performance.

We derived the ratings of shared knowledge and outsourcingerformance by comparing the information regarding every singleonstruct of a certain bank compared to the overall sample in this

aper. As a result, every construct is rated as being high, mediumr low.

Following our reasoning above, we would expect that a tacitype of knowledge transfer, combined with explicit forms, serves as

rmation Management 29 (2009) 342–352 349

the strongest contributor regarding the level of shared knowledge,and this ultimately contributes to outsourcing performance. Table 4summarizes the knowledge-transfer processes, shared knowledgeand performance. While F2, F3 and F6 exhibit a rating as expected,F1, F4 and F5 (shaded rows) behave contradictory.

As expected, F2, F3 and F6 reflect our reasoning that, in thepresence of transfer processes of explicit AND tacit knowledge,shared knowledge and, in turn, outsourcing performance are higherthan in other cases. For example, F6 has only created the trans-fer of explicit knowledge mechanisms by defining the relevanceof its systems in SLAs, without considering further exchange rou-tines with its provider, such as trainings or transfer of staff in thebeginning of the relationship. As a result, its provider is assessed ashaving mediocre banking skills and a mediocre outsourcing perfor-mance.

F2 and F3 exhibit both high levels of knowledge transfer, whichis reached through the transfer of staff in the beginning of the rela-tionship to the provider. The interviewee of F2 explicitly highlightsthe excellent banking skills of its former colleagues, who are nowworking for the provider. Even his counterpart on the providerside is a former colleague. Likewise, F3, a spin-off from its cus-tomer, absorbed staff in the beginning of the relationship. Bothare working in close cooperation, as they are jointly performingtheir activities. In addition to this strong cooperation on the oper-ational level, F3 and its customer have robust connections on theboard level. Both F2 and F3 rate the outsourcing performance asvery good, which is explained by the significant banking skills ofboth providers.

F6 argues along the same lines. It has defined the criticality ofits systems in SLAs, and trains the provider employees regularly. Asa result, he assesses the provider employees as moderately skilledand evaluates the outsourcing performance to be moderate.

Three cases require a closer examination, as they producedresults contradictory to the theoretic reasoning. F1, F4 and F5(shaded rows in Table 4) do not fit to the causality between thehypothesized knowledge-transfer process types, shared knowledgeand performance, described above. Both exhibit “tacit” knowledge-transfer processes, but only report about a medium level of sharedknowledge and a medium or low level of outsourcing perfor-mance.

Although F1 transferred personnel to the provider in the begin-ning of the relationship, the shared knowledge is assessed as beingonly moderate. The banking skills of the former employees, whichchanged to the provider, are regarded as adequate, but the skills ofthe other provider employees also engaged in the provision of ser-vices for F1 are evaluated as poor. Therefore, overall banking skills ofthe provider employees are evaluated as moderate. Additionally, F1suffers from an inappropriate contract. F1 is a local unit of an inter-nationally operating bank. Their outsourcing deal was developedby the corporate international headquarters, without the nationalunit having the possibility to adapt the contract to its local require-ments. The result of this centrally-agreed upon contract for the localunit is disastrous. The cost and processing times after outsourcingare estimated to be 30% higher than before outsourcing. In partic-ular, time-to-market is evaluated as unacceptably slow (“Since theoutsourcing of our IT, we are paralyzed”).

F4 exhibits explicit and tacit knowledge-transfer processes, butonly a moderate level of shared knowledge. The banking skills of theprovider are regarded as very diverse, dependent on the employees.Again, as described by F1, transferred employees are regarded ashaving good banking skills. By contrast, the other provider employ-

rates the overall level of shared knowledge as moderate. F5 imple-mented SLAs as only knowledge-transfer mechanisms. However,F5 evaluates these SLAs to be insufficient. Both the quantity andquality of SLAs are inadequate. Many elements of the service are

3 of Info

ncAicu

ifb(ditwc

arctwat

pHkthniefgsemsbk

50 S. Blumenberg et al. / International Journal

ot covered by SLAs. Existing SLAs often only address very techni-al values without reaching the level of service that F6 expected.s a result, its provider shows only limited banking skills result-

ng in a poor outsourcing performance with a lot of inefficientommunication between both, due to a non-existing shared vocab-lary.

Table 5 shows the results of the provider side and is structuredn the same way as Table 3, with two exceptions: (1) the per-ormance of the outsourcing relationship could not be captured,ecause providers cannot objectively rate their own performancebasically, the self-assessment exhibits only “high” ratings); (2) weistinguished shared knowledge (i.e. banking skills) of the provider

n an objectively measurable value (professional experience ofhe interviewee) and the subjective perception of the intervie-ees (self-assessment of banking skills), illustrated by the last two

olumns in Table 5.All of the interviewees have long-term experience in the man-

gement of relationships with banks, some within the surveyedelationships, and others with other customers from the finan-ial service industry. Besides the experience as account manager,he interviewee of P4 started to work for his current customer. Heorked there for 5 years and then changed to his current position

s account manager on the provider side. Therefore, – in contrasto the other ratings – his professional experience is rated as high.

Overall, the types of knowledge-transfer processes of theroviders fit with the proposed levels of shared knowledge.owever, there are minor deviations between the types ofnowledge-transfer processes and the professional experience ofhe account managers for P1, P2, P3 and P5. These providers exhibitigh levels of knowledge transfer, either by transferred person-el or by the involvement of the customer in joint processes,

n combination with a medium level of professional experi-nce (i.e. long-lasting experience in the provision of servicesor banks without working for a bank). This divergence is miti-ated by the perception of the account managers own bankingkills: P1, P2, P3 and P5 rate these high. The practical experi-

nces of provider employees as bankers could be very helpful inanaging outsourcing relationships with banks. Anyhow, maybe

ufficient relationship management experience with banks coulde adequate for getting a good understanding of banking specificnowledge and for successfully governing the relationship.

Fig. 2. Research model after

rmation Management 29 (2009) 342–352

5. Conclusion

Before concluding, several limitations of our study are discussed.First, the use of perceptual measures to assess the outsourcingperformance might create a limitation by creating a bias in thedependent variables. Second, we investigated one point in time andare, therefore, not able to analyze the dynamics of the knowledge-transfer processes and the development of shared knowledge overtime. Third, the provider cases cannot be used to investigateour second research question regarding the connection betweenshared knowledge and performance, because we doubt that theseproviders are able to realistically, sufficiently, and objectively assesstheir own performance. Fourth, in most cases we used a sin-gle respondent to assess outsourcing performance, as well as thelevel of shared knowledge, which might imply a bias. This limi-tation was diminished by interviewing the expert in charge andby balancing the view of the manager, using additional informa-tion, such as annual reports and press clippings.Referring to thesummary table for banks in the previous section, our case studyindicates a connection between shared knowledge and performance.There is an indication that high levels of shared knowledge posi-tively influence outsourcing performance (H3) (see Fig. 2). Surveyedbanks that have high levels of shared knowledge, e.g. by speakingthe same language as the provider, regarding content, or havingbusiness-trained employees, overall, are very content with theoutsourcing performance. Thus, H3 is supported. As a result, weaddressed a gap in the literature, regarding this link, and providedan indication of the influence of shared knowledge on outsourcingperformance.

As an enhancement to the transfer of explicit knowledge,we found transfer processes of explicit knowledge to consist oftwo parts (illustrated in Fig. 2 by two elements representingtransfer of explicit knowledge). Transfer of explicit knowledgein IT outsourcing relationships consists of elements to transfercontent (Galunic & Rodan, 1998; Tiwana et al., 2003), such ascommunication standards, service level agreements and train-

ings, and structures that define the sender and receiver in an IToutsourcing relationship. Organizations that have clearly definedstructures, which guide the communication between both par-ties, report a high level of shared knowledge. Therefore, in orderto emphasize the importance of such knowledge-sharing struc-

empirical evaluation.

of Info

tpatpttnte

pdthpptataTohtsittfr

aebutottamoddftaomFwtHwampwb

afg

q

S. Blumenberg et al. / International Journal

ures in IT outsourcing relationships, we divide the transferrocesses of explicit knowledge in two parts: the content part,s described by previous literature, focusing on elements forransferring explicitly definable content, and the sender/receiverart, illustrating which explicit structures are required to posi-ively influence shared knowledge in an IT outsourcing context. Ashese sender/receiver structures are clearly defined (i.e. commu-ication partners with documented responsibilities), we evaluatehem as an extension of the transfer processes of explicit knowl-dge.

We could also show that the combination of knowledge-transferrocesses, dedicated to the transfer of explicit knowledge, with those,edicated to the transfer of tacit knowledge, is even more effec-ive. Organizations with both kinds of transfer processes exhibitigher levels of shared knowledge than organizations with transferrocesses either for explicit or tacit knowledge. This result sup-orts Eisenhardt and Santos reasoning (2002, p. 140), who statehat explicit knowledge cannot entirely replace tacit knowledge,nd partially supports the argument that “tacit knowledge formshe background necessary for assigning the structure to developnd interpret explicit knowledge” (Alavi & Leidner, 2001, p. 112).here seems to be evidence in our sample that transfer processesf explicit knowledge and transfer processes of tacit knowledgeave an impact on shared knowledge. Banks with only one of thewo kinds of transfer processes have at least a medium level ofhared knowledge. However, as we do not have any organizationn our sample that uses neither transfer processes for explicit noracit knowledge, we do not have a valid comparison between noransfer processes in use and the use of only one kind of trans-er processes. Therefore, H1 and H2 can neither be supported norejected.

A comparison of knowledge-transfer processes applied by banksnd providers is interesting. Whereas the transfer processes ofxplicit knowledge within this sample are quite comparableetween banks and providers (both use elements such as liaisonnits, SLAs, trainings and communication standards), the tacit-ransfer processes differ. Banks achieved transfer of tacit knowledgenly by means of transfer of staff. In none of the cases was thisransfer of staff planned as a strategic instrument for knowledgeransfer, but rather as a method for turning fixed cost into vari-ble cost and cost reduction. Besides a transfer of staff (whichight be imposed by their customers), providers instead applied

ther methods for the transfer of tacit knowledge, such as jointevelopment of business blueprints and cases, jointly performingemand management or employing mixed workgroups withoutunctional segregation between members. Therefore, transfer ofacit knowledge, from the point of view of banks, seems to bechieved only “accidentally,” without strategic intention. On thether hand, providers seem to apply the above-described instru-ents strategically, in order to achieve a transfer of tacit knowledge.

urther research is required to get a better understanding ofhether there really is a difference between the strategic applica-

ion of transfer processes of tacit knowledge of banks and providers.owever, providers seem to be more interesting for researchers,hen analyzing processes for transfer of tacit knowledge, as they

pply a multitude of transfer processes of tacit knowledge. Thisay be due to outsourcing being the core competence of the

roviders, which is reflected in their, overall, higher experienceith the governance of outsourcing relationships compared to

anks.We explicitly showed the successful integration of knowledge

cross organizational boundaries and provide specific examplesrom a real-world context for the hypothesized possibility to inte-rate knowledge externally (Grant, 1996; Kogut & Zander, 1992).

Summarized we found the following answers to our researchuestions:

rmation Management 29 (2009) 342–352 351

Research question Finding

How do different types ofknowledge-transfer processes influencethe level of shared knowledge

We found transfer processes forexplicit knowledge in anoutsourcing context to consist oftwo dimensions: the contentdimension, which defines howcontent has to be interpreted, andthe sender–receiver dimension oftransfer, which defines interactionstructures between parties. Thecombination of processes designedto transfer explicit and tacitknowledge has the most influenceon the level of shared knowledge.

How are shared knowledge andperformance connected?

The level of shared knowledge ispositively related withperformance, because theintegration of different knowledgedomains provides insights for theexploitation and exploration ofresources.

In conclusion, we found evidence for a linkage between cer-tain types of knowledge-transfer processes, the level of sharedknowledge and outsourcing performance. This provides insightsfor managers who seek to influence outsourcing performance inprovider relationships by employing an orchestration of specificknowledge-transfer mechanisms.

Future research should incorporate additional factors that mightinfluence the creation of shared knowledge. Moreover, additionalperformance measures should be employed.

Appendix A. Interview guideline

Demographics

• Experience in banks/with management of banks as customers.

Knowledge-transfer processes

• Which processes have been applied to transfer knowledge fromthe bank to the provider (i.e. transfer of staff, job rotation, jobshadowing, training, workshops, conferences, liaison unit, SLAs)?

• Are these processes used periodically or once?• Are mixed workgroups without functional segregation applied?

Shared knowledge

Bank:

• The employees of our provider have sufficient banking skills.• Our provider understands our requirements.• We use a common language together with our provider.

Provider:

• Perception of the own banking skills.• We use a common language together with our customer.

Performance (measured in terms of service quality)

• Rating of the overall service quality.• Reliability of the service.• Responsiveness of the provider.• Proactivity of the provider.

3 of Info

R

A

A

B

C

D

D

D

E

E

F

F

F

G

G

G

G

H

H

H

K

K

K

L

L

L

L

M

N

N

N

52 S. Blumenberg et al. / International Journal

eferences

lavi, M., & Leidner, D. E. (2001). Review: Knowledge management and knowl-edge management systems: Conceptual foundations and research issues. MISQuarterly, 25(1), 107–136.

mit, R., & Schoemaker, P. J. H. (1993). Strategic assets and organizational rent.Strategic Management Journal, 14(1), 33–46.

rown, J., & Duguid, P. (1991). Organizational learning and communities of prac-tice: Towards a unified view of working, learning, and innovation. OrganizationScience, 2(1), 50–57.

han, Y. E. (2002). Why haven’t we mastered alignment? The importance of theinformal organization structure. MIS Quarterly Executive, 1(2), 97–112.

ewett, T., & Jones, G. R. (2001). The role of information technology in the organiza-tion: A review, model, and assessment. Journal of Management, 27, 313–346.

ibbern, J., Goles, T., Hirschheim, R., & Jayatilaka, B. (2004). Information systemsoutsourcing: A survey and analysis of the literature. The DATA BASE for Advancesin Information Systems, 35, 6–102.

ubé, L., & Paré, G. (2003). Rigor in information systems positivist case research:Current practices, trends, and recommendations. MIS Quarterly, 27(4), 597–635.

isenhardt, K. M. (1989). Building theories from case study research. Academy ofManagement Review, 14(4), 532–550.

isenhardt, K. M., & Santos, F. M. (2002). Knowledge-based view: A new theoryof strategy? In A. M. Pettigrew, T. Howard, & R. Whittington (Eds.), Handbookof strategy and management (pp. 139–164). London/Thousand Oaks/New Delhi:Sage Publications Ltd.

eeny, D. F., & Willcocks, L. (1998). Redesigning the IS function around core capabil-ities. Long Range Planning, 31(3), 354–367.

ontenot, R. J., & Wilson, E. J. (1997). Relational exchange: A review of selected modelsfor a prediction matrix of relationship activities. Journal of Business Research,39(1), 5–12.

oss, N. J. (1999). Edith T. Penrose and the Penrosians—Or, why there is still so muchto learn from The Theory of the Growth of the Firm. Economies et Societes, 29,143–164.

alunic, D. C., & Rodan, S. (1998). Resource recombinations in the firm: Knowledgestructures and the potential for Schumpeterian innovation. Strategic Manage-ment Journal, 19(12), 1193–1201.

oles, T., & Chin, W. W. (2005). Information systems outsourcing relationship factors:Detailed conceptualization and initial evidence. The DATA BASE for Advances inInformation Systems, 36(4), 47–67.

osain, S., Malhotra, A., & El Sawy, O. A. (2004). Coordinating for flexibility in e-business supply chains. Journal of Management Information Systems, 21(3), 7–45.

rant, R. M. (1996). Prospering in dynamically-competitive environments: Organiza-tional capability as knowledge integration. Organization Science, 7(4), 375–387.

aas, M. R., & Hansen, M. T. (2005). When using knowledge can hurt performance:The value of organizational capabilities in a management consulting company.Strategic Management Journal, 26(1), 1–24.

ansen, M. T. (1999). The search-transfer problem: The role of weak ties in sharingknowledge across organization subunits. Administrative Science Quarterly, 44(1),82–111.

itt, M. A., Bierman, L., Shimizu, K., & Kochhar, R. (2001). Direct and moderatingeffects of human capital on strategy and performance in professional servicefirms: A resource-based perspective. Academy of Management Journal, 44(1),13–28.

een, P. G. W. (1991). Shaping the future: Business design through IT. Boston: HarvardBusiness School Press.

ing, A. W., & Zeithaml, C. P. (2003). Measuring organizational knowledge: A con-ceptual and methodological framework. Strategic Management Journal, 24(8),763–772.

ogut, B., & Zander, U. (1992). Knowledge of the firm, combinative capabilities, andthe replication of technology. Organization Science, 3(3), 383–397.

ee, A. S. (1989). A scientific methodology for MIS case studies. MIS Quarterly, 13(1),33–52.

ee, J.-N. (2001). The impact of knowledge sharing, organizational capability andpartnership quality on IS outsourcing success. Information and Management, 38,323–335.

ee, A. S., & Baskerville, R. (2003). Generalizing generalizability in information sys-tems research. Information Systems Research, 14(3), 221–243.

ee, J.-N., & Kim, Y.-G. (1999). Effect of partnership quality on IS outsourcing suc-cess: Conceptual framework and empirical validation. Journal of ManagementInformation Systems, 15(4), 29–61.

cFadyen, M. A., & Cannella, A. A. (2004). Social capital and knowledge creation:Diminishing returns of the number and strength of exchange relationships.Academy of Management Journal, 47(5), 735–746.

ahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital, and the organi-zational advantage. Academy of Management Review, 23(2), 242–266.

elson, K. M., & Cooprider, J. G. (1996). The contribution of shared knowledge to ISgroup performance. MIS Quarterly, 20(4), 409–432.

ickerson, J. A., & Zenger, T. R. (2004). A knowledge-based theory of the firm—Theproblem-solving perspective. Organization Science, 15(6), 617–632.

rmation Management 29 (2009) 342–352

Nonaka, I. (1994). A dynamic theory of organizational knowledge creation. Organi-zation Science, 5(1), 14–37.

Pandit, N. R. (1996). The creation of theory: A recent application of the groundedtheory method. The Qualitative Report, 2(4)

Pavlou, P. A., Housel, T. J., Rodgers, W., & Jansen, E. (2005). Measuring the return oninformation technology: A knowledge-based approach for revenue allocation atthe process and firm level. Journal of the Association for Information Systems, 6(7),199–226.

Penrose, E. T. (1959). The theory of the growth of the firm. Oxford: Basil Blackwell.Peppard, J., & Ward, J. (2004). Beyond strategic information systems: Towards an IS

capability. Journal of Strategic Information Systems, 13(2), 167–194.Preston, D., & Karahanna, E. (2004). Mechanism for the development of shared men-

tal models between the CIO and the top management team. In Proceedings of the25th international conference on information systems (ICIS).

Quinn, J. B. (1999). Strategic outsourcing: Leveraging knowledge capabilities. SloanManagement Review, 40(4), 9–21.

Ray, G., Muhanna, W. A., & Barney, J. B. (2005). Information technology and the perfor-mance of the customer service process: A resource-based analysis. MIS Quarterly,29(4), 625–652.

Raymond, L. (1990). Organizational context and information systems success: Acontingency approach. Journal of Management Information Systems, 6(4), 5–20.

Reich, B. H., & Benbasat, I. (2000). Factors that influence the social dimension of align-ment between business and information technology objectives. MIS Quarterly,24(1), 81–113.

Santhanam, R., Seligman, L., & Kang, D. (2007). Postimplementation knowledgetransfers to users and information technology professionals. Journal of Manage-ment Information Systems, 24(1), 171–199.

Spender, J. C. (1996). Making knowledge the basis of a dynamic theory of the firm.Strategic Management Journal, 17(Special Issue: Knowledge and the Firm), 45–62.

Te’eni, D. (2001). Review: A cognitive-affective model of organizational communi-cation for designing IT. MIS Quarterly, 25(2), 251–312.

Teo, T. S. H., & Ang, J. S. K. (1999). Critical success factors in the alignment of IS planswith business plans. International Journal of Information Management, 19(2),173–185.

Tiwana, A., Bharadwaj, A., & Sambamurthy, V. (2003). The antecedents of informationsystems development capability in firms: A knowledge integration perspective.In A. Massey, S. T. March, & J. I. DeGross (Eds.), Proceedings of the twenty-fourthinternational conference in information systems Seattle, Washington, USA, (pp.246–258).

Willcocks, L., & Kern, H. J. (1998). IT outsourcing as strategic partnering: The case ofthe UK inland revenue. European Journal of Information Systems, 7(1), 29–45.

Yin. (2002). Case study research. Design and methods. Thousand Oaks: Sage Publica-tions Ltd.

Yin, R. K. (2003). Case study research: Design and methods. ThousandOaks/London/New Delhi: Sage Publications Inc.

Stefan Blumenberg is IT portfolio and relationship manager at the Department ofthe Interior of the German state of Hesse and he is researcher at the E-Finance Lab inFrankfurt. His Ph.D. thesis “A relational view on IT outsourcing” addresses the ques-tions how relationship quality in IT outsourcing relationships between banks andtheir IT service providers can be measured and influenced. Besides IT outsourcingrelationship management, his main research interests are IT business alignment andIT portfolio management. After his studies of Information Systems at Frankfurt Uni-versity, Germany, he worked for 4 years as business manager for the private-publicresearch partnership E-Finance Lab whose research addresses questions in the areaof the industrialization of the financial industry.

Heinz-Theo Wagner is Professor of Management and eBusiness at the heilbronnbusiness school. Previously, he has been working as Operations Director at anaerospace company and was responsible for process redesign and process man-agement including the IS support. Before it, he has been working for more than 15years as a senior consultant and principal on business and IT projects for variouscompanies with focus on the financial services industry. He received his Ph.D. forhis research on IT business alignment with the E-Finance Lab at Johann WolfgangGoethe University, Frankfurt, Germany.

Daniel Beimborn is Assistant Professor at the Department for Information Sys-tems and Services at University of Bamberg (Germany) and he is researcher at theE-Finance Lab in Frankfurt. Daniel graduated in Economics and Business Adminis-tration at Goethe University, Frankfurt, where he also received his Ph.D.. He has beenmember of the Ph.D. program 492 “Enabling Technologies for Electronic Commerce”

of the German National Science Foundation (DFG) and he is author of more than 30reviewed articles and four books on outsourcing of financial processes, standardiza-tion, IT management, and IT business alignment. His current research activities coverbusiness process outsourcing in financial industries, management of outsourcingrelationships, controlling and standardization of IS infrastructures, and the businessvalue of service-oriented architectures.