臺大管理論叢 NTU Management Review VOL.29 NO.1

63 NTU Management Review Vol. 29 No. 1 Apr. 2019 informants’ direct involvement in dual BM decisions. We focused on objective facts rather than hearsay from different informants for the same questions (Eisenhardt, 1989). Thus, we ensured that their interpretations contained meaningful data for this research. It took three months to consolidate internal and external interview data with archival- based data, enabling a better and triangulated understanding of each product phenomenon (Kumar, Stern, and Anderson, 1993). For example, PC makers, which were OEM buyers, gave “ no comment ” on AV-Firm’s retail marketing capabilities. But internal informants’ responses were: “ We are the main player in retail market tuner business .” The conflicting results provided relatively complete and multi-angled thoughts. 3.3 Data Analysis We began by asking why and how a dual BM was chosen for each product line and wrote case histories for each (Eisenhardt, 1989). There were no theoretical preferences or a priori hypotheses. We read each case thoroughly and independently to identify the theoretical constructs and their longitudinal patterns and relationships through the facilitation of graphs and tables (Miles and Huberman, 1994). To ensure accuracy and completeness, the second author provided an independent perspective by reviewing the data and triangulating with the first author. We reconciled by often going back to the data and occasionally to informants to better understand the major capabilities and synergies involved. Finally, we identified the reasons behind each choice. To identify patterns and themes, we first performed cross-product analysis and compared agreements and discrepancies to find insights about each product (Eisenhardt and Graebner, 2007). Agreements were grouped and discrepancies were noted and further investigated by revisiting data and/or informants. We followed an iterative process of cycling among data, literature, and theories to refine our findings. The iterative process ended when “theoretical saturation” (Eisenhardt, 1989) was reached. This data analysis took another three months and resulted in a theoretical model of interpreting why and how suppliers choose dual BMs. 4. Research Findings 4.1 How a Dual Business Model is Chosen In general, our research suggests that successful decision makers adopt a strategic fit and consistent approach to select the right buyer (Porter, 1980). We identified two critical considerations for decision makers in defining a particular type of dual BM so that mutual