Data collection

scarcity of customer companies that had adopted NAV 2009 RTC left little choice in this regard. Additionally, since the third research question addressed the use of the NAV 2009 RTC and not the relationship between the client organizations and the partner companies, the risk of getting biased cases was deemed acceptable.

A total of five customer organizations that had adopted the NAV 2009 RTC agreed to participate as cases in the research. Fortunately, two of the customer companies had migrated from a different ES solution, while three had upgraded from a previous version of the NAV product line. This combination gave the possibility of comparing differences across cases to assess whether findings differed between companies that had migrated and companies that had upgraded. The main characteristics of the five customer cases are shown in Table 5.

Table 5. Selected client companies.

Client organization alias

No. of employees

Industry ES prior to implementing NAV 2009 RTC

Customer 1 800 Aviation (airport) Other ES Customer 2 50 (150

worldwide)

Fashion design Previous version of NAV

Customer 3 75 Furniture

manufacturing

Previous version of NAV

Customer 4 90 Geographical

services

Previous version of NAV

Customer 5 10 Packaging Other ES

2.6.1 Interviews

Interviews were conducted as the primary technique for providing the data for both the case study research and the GTM. Selection of respondents for the interviews was based on their knowledge of, or involvement in, the topic of role-oriented ESs. The simultaneous collection of data for both the case study research and the GTM entailed that while a single interview was often enough to provide data for the case study research of the partner companies in the ecosystem, the opportunity for more interviews in the same company could help saturate the emerging concepts of the GTM. More than one interview was thus carried out in some of the partner companies.

All interviews were semi-structured (Kvale and Brinkmann 2008, p. 130) and an interview guide with a list of topics and suggested questions was prepared prior to the interviews and offered to the respondents before the interview was conducted. While some questions were static to allow cross-case comparison, other questions and topics evolved as the research progressed, to allow evolution of the grounded theories and to explore topics that seemed relevant. A total of 24 interviews were carried out as part of the research project divided between: 6 interviews with representatives from Microsoft and SAP, lasting between 51 and 108 minutes with an average of approx. 60 minutes;

15 interviews with representatives from the partner companies in the Microsoft ecosystem, lasting between 19 and 108 minutes with an average of approx. 60 minutes;

and 5 interviews with the respondents in the client organizations, lasting between 21 and 102 minutes with an average of approx. 45 minutes. All the 23 interviews were audio recorded and fully transcribed resulting in over 280 pages of transcription to support a detailed data analysis.

Two interviews with representatives from one of the partner companies were omitted from the data analysis. Both representatives were unfamiliar with implementation of the NAV 2009 RTC system and one of the respondents was furthermore clearly nervous about being interviewed. A third interview with a manager from the company

was arranged instead, resulting in a useful interview. Additionally, one of the representatives from Microsoft had left the vendor shortly before the interview was conducted. However, the respondent had extensive insight into the reflection of roles in the user interface of NAV 2009 RTC and, with an approval from Microsoft; it was thus decided to include the interview in the data collection for the case study of Microsoft.

2.6.2 Documents

Documents may be used to provide “background and context, additional questions to be asked, supplementary data, means of tracking and development, and verification of findings from other data sources” (Bowen 2009, p. 30). The documents collected for the research project filled each of those purposes at different times in the research project. The documents were primarily collected from publicly available sources and from access to Microsoft’s intranet. A variation of the document type data consisted of getting access to an interactive software version of the ‘Microsoft Customer Model’, which was used by Microsoft for communicating the concept of role-oriented ESs to the partner companies in the ecosystem. While the Customer Model was not intended as a “full-fledged” enterprise model, it still provided valuable insight into the how the role concept was viewed by the vendor. A detailed description of the model is provided in paper III and IV. The role of the collected documents thus ranged from purely contextual background information to being central to the analysis.

2.6.3 Observations

Observations were not initially an explicit part of the research design. However, as part of the participation in different events organized by Microsoft and their partner companies, notes were taken of incidents that seemed relevant to understanding various aspects of role-oriented ESs and thus formed a type of participatory observations (Angrosino 2005). Unstructured observations from a total of two conferences, two vendor presentations, and three meetings/workshops were collected.

The role of the observation data was thus primarily as source of inspiration.

2.6.4 Demo system

Early in the research project a demo version of NAV 2009 RTC was acquired with full feature capabilities and capable of running on a regular PC. The observation of the demo version served multiple purposes. First, it became a foundation for establishing basic knowledge about the artifact of the research project, sharpening the other data collection methods (e.g. the precision of the questions in the interview guides). Second, it provided the foundation for analyzing how Microsoft reflected organizational roles in their ESs. Finally, the running demo system provided a source for triangulating the other data types, e.g. information in the documentation for the NAV 2009 RTC and statements from the interview respondents. A demo version of SAP All-in-One was not obtained.

2.6.5 Data from NAV 2009 RTC implementations

A final type of data consisted of collecting data about the implementations of the NAV 2009 RTC in the client companies. The administration module of the system contained data on the number of user logins and which of the predefined user interfaces of NAV 2009 RTC these user logins were associated with. Second, the data in the administration module allowed identification of which user interfaces that had been adapted compared to standard, when combined with confirmation from the respondents in client organizations.

2.6.6 Summary

The data collection for the research was based on interviews, documents, observations, a demo system of NAV 2009 RTC, and data from the administration module of NAV 2009 RTC in each of the implementations. Table 6 provides an overview of the data collected at the different agents in the ES ecosystem.

Table 6. Overview of data collection.

Ecosystem actors

Data

Vendors x 2 case studies x 6 interviews x Documents x Observations

x Demo system (NAV 2009 RTC)

Partners x 10 case studies x 16 interviews x Documents x Observations Customers x 5 case studies

x 5 interviews x Documents x Observations

x Implementation data of NAV 2009 RTC