METHODS

Figure  2-­‐1  Analytical  model  

rapid growth at 22 percent per annum from 2010 until 2015 (Nasscom, 2011). Currently, India attracts the bulk of global KPO, over 70 percent, and this success model is being replicated in Latin America, Eastern Europe and South East Asia (FinancialExpress, 2014). The rapid growth in this industry and the potential for learnings that are relevant on a global scale make India the obvious selection for our study.

The unit of analysis in this paper is the offshored service-production process in the service provider firm. We focus on subsequent capability developments within the provider firm that are influenced by client interactions and by the type of task interdependence. We study five cases of offshore outsourced service production, and in all cases the service provider performs advanced services, i.e. high-skill services involving the most creative and skill intensive type of service work (UNCTAD, 2004). In view of the Indian context of the study, our experiences from previous research project and our ongoing dialogue with client and service provider firms engaging in offshore outsourcing we may consider the five case firms as typical cases of advanced services that are outsourced offshore. In his discussion on strategies for selection of cases in qualitative research, Flyvbjerg (2006) describes such cases as “paradigmatic cases”.

This implies that findings from the case in question can: “develop a metaphor or establish a school for the domain that the case concerns” (Flyvbjerg, 2006: 230). In this study the relevant

“domain” concerns offshore outsourcing of advanced services to service provider firms in emerging markets. However, as Flyvbjerg (2006) notes it is difficult to know beforehand whether a case truly is “paradigmatic”, and whether and to what extent the lessons learned from studying the case can be generalized beyond the boundaries of the case. In view of this limitation we do not claim that our findings will be readily generalizable on the basis of our five cases. Nevertheless, an important quality of a multiple case study design is that it shows when a

finding occurs in more than one case. This, in turn, can indicate that this finding is not merely an idiosyncratic observation that does not hold any value, except for the case itself, and it therefore could point at something that potentially is a more widespread phenomenon. It is in this context that we later in the paper develop a set of propositions which we present as statements about possible causal relations that can form the basis for further investigation in studies with a larger number of observations than ours.

Despite the common characteristic of being advanced services, the studied services reflect various levels of task interdependence, and can be categorized as either sequential or reciprocal tasks. The five services under study are: measurement sciences (Case A), client services (Case B), market research (Case C), competitive intelligence (Case D), and intellectual property and R&D (Case E) (see Table 2-2 below). Cases A, B, and C involve an Indian multinational that offers business-process and knowledge-process services (BPO and KPO, respectively). We focus on the firm’s KPO department, which has representative offices and production sites in India and around the world; we refer to this firm as ‘ServiceNow.’ Cases D and E involve a service provider that focuses on KPO services. The firm has sales representatives around the globe who travel to client locations. It also owns production sites in India, Chile, and Romania. We call this firm ‘COVALU.’

Table  2-­‐2  Case  descriptions  

Charac- teristic

Case A Case B Case C Case D Case E

Service Measurement science

Client services Market research

Competitive intelligence

Intellectual property and R&D research Service

description

Statistical analysis and global trend estimations;

client provides data

Analysis of and insights into business issues; client provides data

Analysis of and insights into markets;

data needs to be generated

Analysis of and insights into

competition and business environment;

data needs to be generated

Analysis of and insights into global intellectual property and R&D

activities; data needs to be generated Activity type Sequential Sequential Reciprocal Reciprocal Reciprocal

Firm name ServiceNow ServiceNow ServiceNow COVALU COVALU

Client industry

Multimedia Media consulting

Business consulting

Chemicals Chemicals Client

location US/Europe US US Switzerland Switzerland

Year of offshoring

2009 2010 2009 2006 2008

# of

interviews 14 8 8 12 13

Interviewee

positions - Business analyst - Manager (Client, Delivery, Division, HR, Partnership, Regional, Service, and Transition) - Trainer

- Business analyst - Manager (Client, Delivery, Division, HR, Regional) - Trainer

- Business analyst - Manager (Client, Delivery, Division, HR, Team,

Transition) - Trainer

- AVP - Business analyst - Division, HR, Team) - On-side representative - Trainer

- AVP - Manager (Division, HR, Team,

Transition) - On-site representative - Research associate - Trainer Required

educational background

Statisticians, researcher,

Commerce graduates, media experts, statisticians

Business analysts, economists

Chemical engineers, business analysts

Chemical engineers, lawyers

In order to ensure consistency with our advanced services construct and the distinction between reciprocal and sequential task interdependence, the authors carefully studied the service activities, the skills and roles of employees executing the services, the type of knowledge utilized, the extent of required knowledge transfers, and the type of data and inputs required for the production of the services. Two cases (A and B) fall within the category of sequential task interdependence, while three cases (C, D, and E) are characterized by reciprocal task interdependence. The cases that fall under the first category reflected a production process that could be distinguished into stages that were executed sequentially. After the task was executed, activities were handed over to the next employee in line who then continued with the service production. Employees were predominantly young, newly educated, and newly hired without much work experience and the client provided data for the service production to them. While the focus was on the final service delivery in sequential services, reciprocal services placed also a strong emphasis on production process activities. Reciprocal services required more independent judgment by the employees and data was neither provided nor easily accessible. The employees producing these services had experience in collecting and generating data, which was required in the service production.

All services contribute to core operations and/or strategic decision-making in the client firms. The services in cases A, B, and C contribute to a service the client sells to an end-customer. The services in cases D and E are not distributed by the client, but used by them for operational and strategic decision-making purposes.

Data  generation  

Data were predominantly generated through primary data collection from semi-structured interviews with personnel playing key roles in the production of services, including involved employees, team managers, employee trainers, and knowledge managers. The interviews lasted an average of 45 minutes, with interview lengths ranging from 30 minutes to 1.5 hours. In total, we conducted 55 interviews between October and December 2011. All interviews were recorded, transcribed, and analyzed using NVivo 10. In addition to the interviews, secondary data in the form of internal documents and publicly available information were used to triangulate information (Yin, 2003).

Validity (construct/credibility, internal/integrity, and external/transferability) was ensured in several ways. The data-collection process (e.g., interviews with top-level managers, trainers, knowledge-platform managers, and involved employees) and the analysis of various data sources allowed us to gain vast insights into various organizational levels. These activities also enabled us to gain an understanding of various organizational topics related to knowledge platforms, management, training, and service production, such as production processes, challenges, employee activities, and the offshoring process. For all cases, the data were purposefully generated to represent common reciprocal and sequential service-production processes within each type of service production. Although only one researcher conducted the interviews, the transcribed interviews were shared, collectively coded, discussed, and analyzed by all authors.