Data collection

Data collection of each case commenced early on in the research process with the con-struction of the list of identified cases. Once the cases to be included in the study were selected, primary data was collected by organizing semi-structured interviews with each of the provider companies and their selected customers, thus employing a dyadic ap-proach in data collection. Interviewing enables the collection of valid primary data, allow-ing for mappallow-ing of points of interest and meanallow-ings (Saunders et al. 2019, pp. 434-435), thus building the basis of understanding customer value formation in this study. In addi-tion to interviews, case studies often combine different data collecaddi-tion methods, such as observations, archives, or other types of secondary data (Eisenhardt 1989; Yin 2003, p.

83). Incorporating multiple data sources to support the findings with data triangulation will increase the quality and reliability of the case study (Yin 2003, p. 83). This study too supplemented the interview data by various secondary data sources to enhance and validate the datasets. These included primarily non-financial reports, company webpages, news articles, and presentation material. Table 6 gives an overview of the data sources for all four cases, embodying data sources for both the provider and cus-tomers researched in each case. The table is followed by more detailed description of the data collection methodologies.

Table 6 Data source overview

Data source Case Provider interviews 1 interview 1 interview 2 interviews 1 interview Customer interviews

& number of customers 3 interviews /

3 customers 2 interviews /

2 customers 4 interviews /

3 customers 1 interview / 1 customer Reports and

presentations Annual report (1)

media sources Podcast (1)

Webpages (9) Webpages (6) Webpages (8) Webpages (2)

To obtain a focused approach to the data collection, it is important to have initial research questions defined, as well as some tentative ideas of important constructs from previous literature (Eisenhardt 1989). In this study the different value sources play the role of these constructs, and the interview guides were strongly inspired by the ideas from previous literature, in accordance with the abductive approach and Eisenhardt’s (1989) recom-mendations. Table 2 demonstrated the results of this literature review on customer-per-ceived value, which heavily influenced the construction of the interview guide (Appendix C). Literature was explored in a systematical way, with Table 7 showing the literature review methodology for mapping out the recognized sources of customer value. Table 8 shows the same for exploring the research gap by reviewing the extant literature that tackles both customer value and circularity. Systematic approach was selected to obtain as comprehensive understanding of the extant customer value constructs and terminol-ogy as possible.

Table 7 Literature review methodology for customer value source mapping Search

Table 8 Literature review methodology for reviewing extant literature on customer value in circular contexts

Interviews are an essential data source in case study research (Yin 2003, p. 92). Saun-ders et al. (2019, pp. 144-145) recommend using semi-structured interviews in explora-tory studies and in situations where questions may be complex, and their order or logic

varied. Therefore, and because of the good fit with the overall research design, semi-structured interviews were the most appropriate choice for this study. The interview ques-tions were built around topics of circular economy, provider-customer relaques-tionships and most importantly possible sources, drivers, and barriers of customer-perceived value.

Provider and customer companies had their own, slightly different interview guides (see Appendix C for details). The interviewees were selected purposively, with the aim of in-terviewing persons that have first-hand information on the customer interface (providers) or purchasing logic (customers), and most of whom would work at least in managerial roles to also have a good comprehension on the strategic objectives and values of their respective companies. In some cases, an effective way to achieve these goals was to organize group interviews with persons from different levels and functions of the firm.

Detailed data of the interviews is presented in Table 9. Due to the ongoing pandemic situation the interviews were conducted with videocalls in Microsoft Teams. All interviews were recorded with interviewees’ permission. In almost all of the interviews, a second researcher was present to enable researcher triangulation, to ask clarifying questions, and to enable more efficient note taking, which is recommended alongside recording by Saunders et al. (2019, p. 461). The recordings were transcribed for analysis, some by external transcribers and some by the researcher himself.

Before the interviews, the interview themes were shared with the interviewees to enable them to gather necessary information, as well as perspectives from elsewhere in the organization. During the interviews, attention was paid to maintaining an objective stance, not leading the interviewees while still establishing rapport and encouraging rich descriptions by making follow-up, clarifying and reflective questions as recommended by Saunders et al. (2019, pp. 451, 460). Memos were written down during the interviews to support the discussion and getting the subsequent analysis started. After each interview, a short document compiled by the researcher summarizing the key insights was sent to interviewees to support early participant validation of the data and kick-off of the analysis for the researcher. This kind of summarizing is an effective method to minimize especially the interviewee bias (Saunders et al. 2019, p. 461).

Table 9 Interview data

Interviewees Date Organization Interviewee level Duration

I1 15.3.2021 Neste Manager 62 min

I2 26.3.2021 Industrial

Tools

Manager 76 min

I3 & I4

(group interview) 26.3.2021 Touchpoint

Director (I3) &

Manager (I4) 76 min

I5 13.4.2021 Customer T1 Manager 48 min

I6, I7, & I8

(group interview) 29.4.2021 Konecranes

Director (I6) &

Managers (I7, I8) 54 min

I9 6.5.2021 Konecranes Director 52 min

I10 & I11

(group interview) 10.5.2021 Customer T2

Manager (I10) &

Employee (I11) 83 min

I12 24.5.2021 Customer IT1 Manager 49 min

I13 25.5.2021 Customer K1 Manager 55 min

I14 26.5.2021 Customer K2 Employee 43 min

I15 28.5.2021 Customer K2 Employee 40 min

I16 28.5.2021 Customer K3 Manager 49 min

I17 28.6.2021 Customer N1 Manager 80 min

I18 & I19

(group interview) 28.6.2021 Customer N2 Director (I18) &

Manager (I19) 54 min

I20 29.6.2021 Customer N3 Manager 76 min

To deepen and verify the data regarding both the providers and their customers, com-prehensive secondary data were gathered simultaneously with the interviewing process.

This included for example non-financial reports and web page information, with the dif-ferent types of secondary data summarized in Table 6. Because interviews, being verbal records, are always subject to biases, poor recall, and inaccuracies, having additional data sources for verification and supplementary purposes is important (Yin 2003, p. 92).