2.5 Empirical programme of research

To help answer the question of ˜how do GSA relationships evolve over time? we have discussed six themes that provide the lenses for micro-level analysis of GSA relationships.

Before going into detail of how the cases that we have researched relate to particular themes, we provide an overview of the empirical programme that included these cases:

• We investigated three relationships:

• A Canadian company, GlobTel, in GSA relationships with four Indian companies, MCI, Witech, SRS and Softsys.

• A UK firm, Gowing, in a GSA relationship with an Indian firm, Eron.

• A UK firm, Sierra, in a GSA with its subsidiary in India.

• In Japan and Korea we did not investigate any particular relationship but interviewed a number of managers from Korean and Japanese companies who were at a relatively early stage of trying to establish relationships with Indian firms.

The analysis of these three GSAs is presented in seven chapters:

• Chapter 3 : an overview of the GlobTel GSA programme, extending over their four relationships in India.

• Chapter 4 : case study of GlobTel “Witech, analysed under the theme of ˜standardization .

• Chapter 5 : case study of GlobTel “ComSoft, analysed under the theme of ˜identity .

• Chapter 6 : case study of GlobTel “MCI, analysed under the theme of ˜space and place .

• Chapter 7 : case study of Sierra UK “Sierra India, analysed under the theme of ˜knowledge transfer .

• Chapter 8 : case study of Gowing “Eron, analysed under the theme of ˜power and control .

• Chapter 9 : an exploratory analysis of the relationships between Japanese and Indian firms, drawing upon the issue of cross-cultural communication challenges.

Syntheses of the theoretical and managerial implications that arise from these case studies are presented in chapters 10 and 11, respectively. These syntheses are based upon an inter-case comparison of various theoretical and managerial issues.

A note on methodology

The cases described in this book are part of an empirical research programme on GSA relationships in India. The names of all the companies and individuals mentioned in the description and analysis of the cases are pseudonyms. We have taken care to present the details in such a way that the identities of the entities are disguised as much as possible. The focus of the analysis constantly has been on issues rather than individuals.

The research started in 1996 with a five-person research team: three from a Canadian university and two from an Indian management institute. Over time, two of the Canadians and one Indian researcher left the project, and a UK-based researcher joined it. One member from the original team was present in all the interviews. We thereby ensured continuity in the interpretation of the data. The research programme started with the study of GlobTel, a large North American telecommunications firm, which had established relationships with four Indian companies in the early 1990s. Subsequently, the research program was extended to some UK, Korean and Japanese companies who were at various stages of outsourcing relationships in India.

Two aspects of our empirical program are salient. First, the underlying basis for the empirical programme has been our quest to obtain interpretations from managers and developers on both sides about the state of the relationship, the different challenges and how the respective organizations are dealing with them. The balance between the viewpoints obtained from either side, however, necessarily varied in the different cases owing to issues of access, cost and time constraints. Secondly, a longitudinal perspective was adopted in an attempt to try and understand the issues, the responses and future expectations, on a continuous basis. We tried to visit the sites at least once a year, to conduct repeat interviews at least with some of the key respondents and to get them to reflect on some of the issues that they had raised in earlier meetings.

As can be imagined, conducting such a large-scale programme of empirical research spread over not only multiple countries and organizations but also over time is an extremely complex endeavour. We tried to conduct the study through a network structure of researchers affiliated to different institutions spread over different countries and having close linkages with some of the organizations that were studied. Using this network structure, we conducted more than 200 interviews spread over ten organizations and seven countries in 1995 “2000, the details of which are summarized in table 2.2. The analysis of each relationship is anchored by the overall set of interviews, and thus the understanding of a particular case cannot be separated from the larger study of which it is a part. The analysis of any sub-set of interviews is inextricably based on our broader understanding of GSA relationships.

Table 2.2: Summary of interviews conducted

GSA relationship	Indian respondents	Alliance respondents	Total respondents
GlobTel		25	25
MCI	12		12
SRS	18		18
Witech	15		15
Gowing “Eron	22	19	41
Sierra	14	7	21
Japan	30	25	55
Korea	15	16	31
Total	126	92	218

Our research approach was aimed at developing a holistic and in-depth understanding of the process of evolution of a GSA relationship over time. We selected a longitudinal research design, except in the East Asian case. By interviewing people involved on both sides of the relationship in their respective contexts, we tried to understand the issues they were facing , the expectations they had and how these were changing over time. We did not have a fixed schedule of questions but rather a set of issues that we felt were important in shaping the relationship. We would start by asking the respondents to comment on these issues, identify new ones and describe how they were dealing with them. Over time, as we met people for a second and third time, there already existed a shared understanding between us and we could start the conversation from where we had left off at the last meeting. We provide below a brief overview of the different pieces of empirical work conducted in the four sets of case studies.

GlobTel and the Indian firms

GlobTel, a large North American-based telecommunications MNC, had relationships with four Indian firms. In chapter 3, we first present the GlobTel case, and this is followed by analysis of three of the four relationships in chapters 4 “6. This includes interviews with all the GlobTel respondents in Canada, their UK lab and their Indian expatriate managers. This perspective is supplemented with some secondary data obtained from company websites and newsletters in order to develop some element of a contextual understanding. By first presenting the GlobTel perspective, we avoided repeating a number of the common elements across the three cases, for example the motivation of GlobTel to start relationships in India.

Atotal of seventy interviews were conducted (with GlobTel personnel as well as those from the four Indian firms), and except for one that took place over videoconference, all were conducted in face-to-face settings in the location of the interviewee s workplace. The videoconference meeting had four of the five researchers present and the hook-up took place between the University in Canada where the four researchers were present ^[1] and the respondent in his US office. The other face-to-face interviews had two of the researchers present (with one of the members being common across all interviews over time). Each interview lasted for about an hour except the videoconference, which lasted about 2 hours. All the interviews with the Indian respondents and also the seven GlobTel expatriates in India were recorded. Only four of the remaining GlobTel interviews conducted in North America were recorded owing to the reluctance of the interviewees. The reluctance of the others to be recorded could be because of the more formal requirements of non-disclosure arrangements in North America. We found the Indian companies in general to be more comfortable about recording. The recording of the videoconference was flawed, so we made use of our handwritten notes taken on this occasion.

The first set of interviews that took place in GlobTel s North American office in November 1996 helped to reconstruct the relationship historically, and to develop an understanding of the perceived future challenges. Our request to GlobTel for permission to conduct interviews with their partners in India was granted and we visited the Indian partners in February 1997. There were several limitations to this study. In 1998 we visited the North American office but could only get to meet the staff from the Research and Development office and not the line managers since they were either extremely busy or travelling. We acknowledge this to be a limitation of the research as the line managers views are crucial to a holistic understanding of the process of growth of the relationship. Another limitation of the empirical study was that after 1998 we did not get further access to do interviews in North America because of sensitive changes taking place in the relationship during that period. This limitation was, however, partly offset by the fact that we met with GlobTel s expatriate managers in India and the UK, and learned the nature of issues as from their perspective.

The Gowing “Eron relationship

Interviews in Gowing began in 1998 as a result of an introduction to the Indian company Eron. Forty-two interviews were conducted with the senior and middle management and the development staff, in both Britain and India. Early interviews concentrated on historical reconstruction of the relationship between Gowing and Eron, followed by interviews at different periods with actors at various levels in the organization. After the initial interviews, follow-up interviews would normally start with participants being asked to bring the ˜story up to date with key events in the evolving relationship. We would then explore these issues and discuss themes that were of relevance in previous interviews as well as future challenges. Information on the relevant context at various levels (company performance, industry trends, etc.) was also collected.

Two field trips to Chennai in 1998 and 1999 led to a total of three days of interviews with key participants, including the programming staff. The UK-based researcher undertook all interviews in the UK. All interviews were taped, transcribed and subsequently discussed with one of the Indian researchers. Subsequent trips were made to the UK offices of Eron and Gowing for formal interviewing and the attendance at meetings when requested by the Gowing management. The UK and Indian researchers, who still remain in close contact with Eron, made one subsequent field trip to Mumbai in 2000 to conduct further interviews. A lot of other secondary data including newsletters, press announcements, etc. were also collected to get a broader understanding of the relationship.

The Sierra subsidiary

The first interviews in Sierra, a UK-based software house, took place in 1998 during the setting up of their subsidiary in Bangalore. Subsequent meetings took place in the interviewees workplace in either the Bangalore office or in London. The UK-based researcher conducted the interviews in India with at least one or sometimes two of the other researchers present, but he conducted the interviews in the UK independently. Most of the interviews were recorded except for those with the India centre manager where handwritten notes were taken and transcribed in full afterwards. The transcripts were made available to all three researchers concerned for the process of analysis. The final set of interviews with the Indian staff planned for early 2000 were cancelled owing to the centre being closed down.

The analysis of the data was done at periods where the researchers met and discussed substantive themes and issues and related the discussion to theoretical concepts. The writing of reports was critical, and two such reports were written, one early in 1999 and the other in 2000 after the India centre was closed. The reports summarized interviews and identified substantive themes, together with our views on how the process could be improved. We felt it important to provide this input since Sierra was a small company with limited resources and they expected something in return for the time spent in interviews with us. Providing these reports helped to maintain access and served as a useful vehicle for data analysis involving discussion with interviewees.

Interviews in Korea and Japan

Unlike the earlier cases, the interviews in Korea and Japan were not focused on any one particular relationship. Rather we spoke to a cross-section of managers and developers from Japan and Korea who were either at an early stage of outsourcing work to India, or were thinking about doing so. So, in general we have an understanding of the perceived challenges and opportunities of doing work in India. Interviews in Korea only took place in the first round in 1998 and the subsequent meetings focused entirely on Japan. The reason for this was primarily logistical . The interviews in Korea took place in two large Korean conglomerates, in the divisions that were coordinating with the software development activities in India. We did a total of thirty individual and group interviews in Japan and half that number in Korea.

Interviews in Japan were conducted over four visits during which various companies were seen. We also visited the Singapore subsidiary of one of the Japanese companies because software development was being outsourced to India through this ˜hub . To address the problem of language, we conducted interviews together with a Japanese speaker. On the Indian side, we conducted twenty-five interviews spread over four Indian firms that we had studied in the GlobTel case. These firms also had Japanese projects and since the researchers had developed a good rapport with them as a result of the prior relationship, access to their Japanese projects was facilitated. Another reason was that we were interested in developing a comparative analysis with the GlobTel case, and by taking the same Indian firms; we could study the difference in approach to the two sets of clients . In the Korean case, we conducted fifteen interviews in two of the Korean software development centres in India.

Process of data analysis

All the interviews were transcribed (if tape-recorded) or written up as detailed field descriptions (if not recorded). The different researchers present in the interview would make up their own notes and impressions and share them with others electronically . Through discussions we tried to develop cross-perspectives on the interpretations. These different notes and discussions provided the basis for a ˜first-cut on the analysis in the form of a report that was submitted to both the partners in the GSA. We tried to prepare and circulate this report within about 2 “3 months of having completed a set of interviews, and submitted three reports over the course of the GlobTel study and at least one or two reports in each of the other cases. The reports served a number of purposes. First, they summarized and consolidated our impressions of the relationship and as such provided a basis for subsequent theoretical analysis. Secondly, they became points of communication between the researchers and the respondents and useful in ˜continuing the conversation over time . When starting a new set of interviews, we would refer back to the previous reports and would urge the respondents to comment on their contents and how they felt things had changed with respect to some of the issues mentioned. These discussions on the reports helped us to ˜verify and cross-interrogate our interpretations. Furthermore, even though these reports were not part of an explicit and predetermined ˜action research strategy, they provided useful feedback and helped to continue our research access. They also improved our credibility at the research site and led to management openly discussing strategic issues with us.

Although data collection was done mainly through semi-structured interviewing, additional data was obtained through company websites, newspaper reports and company newsletters. The broad scope of people interviewed provided us with a wide range of interpretations over time. Our research approach reflects the tradition of studies that can be broadly classified as ˜interpretive case studies (Walsham 1993, 1995). There is an increasing body of work in the information systems (IS) literature based on this approach (for example Suchman 1987; Boland and Day 1989; Orlikowski 1992; Walsham and Sahay 1999).

Although the approach to data collection was empirically grounded and ˜bottom-up , the data analysis was interpretive and conducted in a holistic and ˜top-down manner. The interpretive analysis, as mentioned earlier, necessarily had the frame of reference of the different interviews we were conducting and was not restricted to the particular case studies reported . We did not attempt to adopt the use of computerized tools such as ˜Nudist as we felt they would be reductionist as they would not be able to take into account these various other interviews and data sources we were engaged in reading and discussing over time, countries and organizations. We decided to rely on our individual and collective abilities to integrate and synthesize the more than 1,000 pages of data, notes and reports. Key to this integration process were periods of intensive discussion between the researchers who came from different academic and cultural backgrounds and countries and introduced multiple perspectives into the analysis. For example, one of the researchers who lived in India had a deep understanding of the Indian software industry and provided various ˜place-based insights that otherwise might have been missed. A senior Canadian researcher had studied JVs for a long time and could provide the group with insights on the working of such arrangements, especially as viewed from the perspective of a large North American MNC.

The researchers met at least twice a year for periods extending to a few weeks in India, Canada, the UK or Norway for detailed discussions on individual interpretations of the transcripts. During these meetings we discussed various issues and related them to relevant theoretical concepts. This helped to clarify and refine the interpretations, to suggest alternative interpretations and to identify further areas for theoretical investigation. The approach thus involved a continuing dialogue between data collected, our interpretations and feedback from the case participants, discussions with colleagues and our continued reading of related literature. The rapid changes taking place in the industry forced us to constantly review of the interpretations and explore theories to develop insights into the processes of change.

^[1] Three of the four researchers were based in this particular university in Canada, while the fourth had travelled from India for a scheduled research meeting of the four researchers. The videoconference was scheduled during this period to take advantage of his presence.