Changing Offshore Relationships – Barney, Moe, Low and Aurum

Indian Intimacy Ends as the Chinese Connection Commences: Changing Offshore Relationships

Hamish T. Barney1, Nils Brede Moe2, Graham C. Low1 and Aybüke Aurum1

1School of Information Systems Technology and Management
University of New South Wales
Sydney Australia

2SINTEF Information and Communication Technology
NO-7465 Trondheim, Norway

Corresponding author: Hamish T. Barney

Address: Information Systems, Technology and Management
UNSW Sydney
NSW 2052
Australia

Email Address: )

Indian Intimacy Ends as the Chinese Connection Commences: Changing Offshore Relationships

Keywords: Outsourcing, Backsourcing, Global Software Development, Virtual Teams, single-case study

Abstract

A large Norwegian company with a medium-sized internal software development department decided to outsource part of their software development function to an Indian outsourcer. After a year and a half, and what was perceived to be an unsatisfactory performance, the decision was made to wind-down and terminate this outsourcing deal. It was decided that all software development should be backsourced (i.e. returned in-house), partly sent off-shore to the company's fast growing offices in China, which the company decided to expand to include a software development department. The remainder being returned to the responsibility of the office in Norway. The contribution of this paper is to provide one of the first case studies focusing on sourcing decisions made after the an existing outsourcing relationship has been terminated or not extended beyond the length of the original contract term and the development of an improved framework for understanding sourcing decisions of this kind.

Introduction

What are the factors in a firm’s decisions about what to do after an outsourcing relationship ends, either terminated prematurely or not extended beyond the term of the original contract? Should the company find a new outsourcing partner or should it backsource and fulfill that information systems function internally? What factors are important in determining how this new information systems function should be governed or organized? These questions have not been addressed very thoroughly in the existing literature on information systems sourcing despite the fact that such situations are important for industry, being made with increasing frequency and substantially different from initial outsourcing decisions on a number of dimensions [1].

Information systems outsourcing has matured considerably since its inception and has become an integral part of the landscape in information technology. This maturation has led to a change in the kind of sourcing decisions that companies are making. Now rather than deciding whether to outsource parts of their information systems function many companies are facing decisions about what to do after ending an existing outsourcing relationship. Should the information systems function be backsourced (i.e. returned in-house) or should the company look for a new outsourcing partner? How should this new information systems function be organized? These post-outsourcing sourcing decisions are quite different decisions to green-fields outsourcing decisions that have traditionally been the object of study in information systems research [1-3].

One example of such a decision is the case presented in this paper. A large Norwegian company, NORWARE, (the company name has been changed for reasons of confidentiality), that outsourced part of its software development effort to India. It became apparent to NORWARE that this particular outsourcing relationship was not working as anticipated. As a result, NORWARE made the decision to backsource the previously outsourced parts of their software development, partly sent off-shore to the company's fast growing offices in China. Their operations in China were expanded to also encompass software development with apparent success. This case is interesting because it is a mix of outsourcing, offshoring and global software development (GSD).

Motivation

Several information systems authors have argued that sourcing decisions made after an existing outsourcing relationship ends are not only becoming increasingly important they are different from initial or green-fields sourcing decisions [1-3]. The importance of these decisions is demonstrated by the frequency with which these decisions are made, and because they differ in a number of ways from initial or green-fields outsourcing decisions.

An international survey conducted by Lacity and Wilcocks [4] found that a third of the 79 firms surveyed had cancelled their outsourcing contracts. Similarly, a recent industry survey on information systems outsourcing found that of companies who engage in outsourcing, 49% have terminated outsourcing contracts prematurely (i.e. before the end of the contract period) [5]. Furthermore, in 57% of those cases the function was passed to a new outsourcer while in the remainder the function was backsourced [5]. These figures are lower bounds, they do not include companies that decided not to renew expiring outsourcing contracts. The proportion of companies that have faced and will face a post-outsourcing sourcing decision is considerably higher.

Not only do decisions about what to do after an existing outsourcing relationship comes to an end arise relatively frequent they are often of high value. [2] have compiled a list of high-profile backsourcing decisions. One case they highlight is that of JP Morgan Chase who backsourced their information systems function after prematurely terminating a $5 billion (USD) outsourcing agreement with IBM.

Researchers have documented cases where companies have learned over the course of a single outsourcing contract [6, 7], it seems likely that firms would also use their outsourcing experience in subsequent sourcing decisions. An acrimonious end to the old outsourcing relationship could bias the firm against the further outsourcing of information systems functions despite its potential advantages as a result of potential psychological biases like the availability heuristic [8, 9]. Having outsourced the particular information systems function, the firm may no longer have the capability to fulfill the function internally in the short to medium term at reasonable cost. These short-term constrains may lead to the company making a sub-optimal sourcing decision. The results of this research will be of interest, not only directly to firms who are terminating an outsourcing contract, but also to firms considering outsourcing. It is hoped that this research will help them to consider ways of structuring their outsourcing agreements in light of potential contingencies that may occur should they decide to terminate their outsourcing contract.

Similarly the results of this research may be of interest to outsourcers. It may help them identify outsourcing deals that are at high risk of being backsourced in the future. The results may also aid in the identification of customers that are at risk of ending their outsourcing contract to either backsource or find an alternative vendor.

The related literature on backsourcing, outsourcing and vertical integration will be reviewed. Based on the literature a preliminary model of the factors that appear to be relevant to the decision about what to do after terminating an outsourcing contract will be discussed. Finally, a validation of the model will be presented based on a recent sourcing decision.

Post-outsourcing decisions

Whitten and Leidner [3] analysed the perceptions of information technology executives in a recent decision about outsourcing contracts. A combination of transaction cost economics theory [10, 11] and social exchange theory [12] was used as the theoretical underpinning of their research. Their results showed a significant difference between the three groups summarized in Table 1.

McLaughlin and Peppard [28] reviewed publicly available records on high-profile backsourcing deals and their causes. In their analysis they encountered six main reasons given for backsourcing decisions: problems arose during the contract and/or specific objectives were not achieved, information technology had come to be seen as strategic, changes in the business environment or business change and evolution, changes in the available technologies and changes in the management. This research gives some interesting preliminary clues about the difference between an initial decision to outsource and the subsequent decision about what to do after a contract with an outsourcer has been terminated.

Hirschheim and Lacity [13] describe a qualitative analysis of fourteen different firms' insourcing and backsourcing deals. In this paper they refer to both backsourcing and insourcing (as defined above) as `insourcing'. For their study they interviewed a total of 41 participants in fourteen firms. Twelve of the firms studied had insourced while two had backsourced their information systems functions. In both cases of backsourcing, the cost savings expected of the previous outsourcing deal did not eventuate. As a result, the firms both backsourced their information systems function and apparently achieved cost reductions over the previous outsourcing relationship. It is, however noted that neither firm formally reviewed the comparative performance of the two arrangements.

Based on a review of the literature on outsourcing and sourcing decisions made after an existing outsourcing relationship has been terminated, Barney, Low et al. [1] have produced a conceptual model that attempts to explain how individual sourcing options are evaluated after an outsourcing contract has been terminated (see fig. 1). Each sourcing option, whether it be backsourcing or re-outsourcing would be evaluated on these dimensions proposed in their model and then the most appropriate sourcing option chosen. The current outsourcing arrangement provides a base line against which the other outsourcing alternatives can be compared.

The factors in the model are as follows:

  • Transaction cost factors: The costs of using the market versus internal organization[13], The higher the relative cost of using markets (decentralization) to internal organization, the more likely that a particular transaction will be conducted internally to the firm (or hierarchically). Here, Williamson’s [10, 14] instantiation of this theory was adopted, as the most widely used in outsourcing research [15] and rigorously empirically verified[16].
  • Degree of complimentarity between objectives: Makadok and Coff [17] assert that the degree to which the objectives of current production, investment in future production, private activities and leverless objectives (those objectives that are difficult to motivate extrinsically) will be an important determinant of governance.
  • Governance: This is the way the particular option is managed and organized. At its most basic the governance could be in-house or outsourced. The dimensions of asset ownership, perfomance incentives and authority are the most important dimensions along which governance can be understood [17-20]. The degree to which the optimal form of governance is feasible will be a factor in the evaluation of a particular sourcing option.
  • Outsourcing experience: Experience with the previous outsourcer [3] and the sourcing option under consideration will be important determinants of how positively that particular sourcing option will be perceived.
  • Resourcing factors: Short to medium-term costs or constraints associated with a particular sourcing option (such as switching costs [3])may render that option unacceptable. As such, resourcing factors may play a role in the evaluation of a particular sourcing option.

Barney, Low et al.’s [1] conceptual model of sourcing decisions made after an existing outsouring relationship has been terminated prematurely or not renewed at the end of the contract will be the main model used to analyze the case study. This is one of the few models in the literature to specifically examine post-outsourcing sourcing decisions. Moreover, it encompasses the factors considered by the only other model of this decision [3].

Research Method

The goal of the research reported herein was to understand the sourcing decisions made after the an existing outsourcing relationship has been terminated. Therefore it is important to understand the software practitioners’ and the management’s perceptions of the terminating decision. Therefore, we collected data from a company that faced this decision recently.

We report on a single -case holistic study [21], in which we studied one phenomenon in one company. Before describing the results of this research, the context of the case and the research method are described

Study Context

NORWARE is a large Norwegian company. One large and important part of their operation is their software development arm and are in fact one of the largest software development organizations in Norway. They produce specialized software for the engineering domain. NORWARE sell mass-market software and they also write customer specific software on a contract-basis. Although it NORWARE is quite an old company, their software development division has grown quickly in recent times. China has, especially in the last couple of years, become an important market for their products.

NORWARE has very low staff turn-over, less than 10% per year in their software division. Most of the people working in their software development department have been trained as engineers rather than professional software developers (who represent about a third of their staff) but the proportion of software developers is increasing. NORWARE are used to getting the best people to work for them.

NORWARE has prior experience with offshoring and Global Software Development (GSD). A GSD team consists of distributed members who collaborate on a common software project while working across geographic, temporal, cultural, political, and organizational boundaries to accomplish an interdependent task [22, 23]. In addition to conducting software development work in the main office in Norway NORWARE also conducts software development in their offices in Eastern Europe and the UK. These are wholly owned subsidiaries of NORWARE. The case studied, was however their first experiment with outsourcing software development. Two Indian projects and one project involving the Eastern Europe office is described more in detail [24].

Analysis Method

For the analysis, we relied mainly on qualitative interviews, because these provide a rich picture of the reasons for the decisions taken. The semi-structured interviews were conducted based on a interview guide. The questions in this guide was identified from the literature, and tailored to this case by using experience from prior case-studies at this firm. The two first authors read the interviews and identified concepts related to the research problem. These concepts were then discussed among the authors. Author number two also discussed the findings with the company, to clarify and to verify our findings.

Interview subjects were selected based on their involvement with the original outsourcing effort in India, the decision to terminate this outsourcing agreement and move to China or the subsequent backsourced software development operation in China. While only six subjects were interviewed this sample represented the majority of the senior staff who were directly involved in one of these three ways with the sourcing decisions that were made. While the number of interviewees is not large it represents a fairly exhaustive sample for the purposes of investigating the way this sourcing decision was made at NORWARE. Each of the semi-structured interviews took approximately an hour.

Results

The outsourcing decisions took place in four main phases.First, the initial decision to outsource to India and the start of the project. Problems in the relationship with the Indian outsourcer started occurring and several different ways of improving the situation were tried unsuccessfully. There was a growing recognition at NORWARE that the outsourcing relationship was untenable, at the same the expansion of the Chinese office made backsourcing software development here the natural alternative to the outsourcing. The final phase is the situation as it currently exists, the Chinese office is handling software development work and the number of software developers working there is growing rapidly. Our findings will now be reported

“Something in the air”

The initial decision to outsource to India was made in a fairly ad-hoc or natural way [25]. Global software development (GSD) is becoming the norm [26], an there was a general feeling that lots of companies were outsourcing to India and that NORWARE should try it too. GSD claims to make it possible for organizations to gain time-zone effectiveness, leverage a large skill pool, and exploit low labour costs in certain part of the world [27]. NORWARE was tempted by this, since they had difficulty recruiting enough software developers in Norway and already had experience at GSD. Outsourcing software development seemed to offer access to a large pool of cheap software developers. Naïve comparisons of the costs were made but most of the initial justification to outsource software development seemed to be post-hoc. As one of the executives responsible for the decision said, “the Indian thing was part of… something in the air… I think we partly built up a rationale afterwards. I mean it was a sort of curve-fitting... Its not that very structured. ” (Executive).

It was decided that the outsourcer should work together with a one of NORWARE’s teams of developers on a project that was just about to start, extending one of NORWARE’s existing mass-market products. NORWARE was very happy with the employees of the outsourcer at this stage; “These are really really good guys! Really clever. I would employ them straight away if I could” (a project manager). After working closely with the Norwegian project team for a month, the Indian outsourcers returned to India to continue working on the project.