The Effect of Consumption Behavior on Online Commerce Membership Adoption: Econometrical Evidence from a Korean Commerce Platform Membership Service*

Purpose: This study examines how consumers’ past purchasing behavior influences their likelihood of subscribing to paid e-commerce memberships. By analyzing spending patterns, including platform-specific expenditures and bundled service consumption, the research provides insights into consumer decision- making during the pandemic and post-pandemic periods. Research design, data, and methodology: Using panel data from 4,597 consumers over four years (2020–2023), this study employed four logistic regression models to analyze the relationship between consumer spending behavior and membership adoption. For the period-specific analysis, separate models were constructed for the pandemic period (2020–2021) and post-pandemic period (2022–2023), examining how spending patterns in the first year predicted membership adoption in the following year for each analysis period. For the integrated analysis, two additional models were developed: one controlling for period fixed effect using a binary variable (AfterCovid), and another incorporating an interaction term between the variable AfterCovid and a variable representing offline leisure consumption behavior of each consumer to test whether leisure spending effects differed significantly between periods. Key independent variables of interest were calculated to represent total spending and retail, specific commerce platform, digital contents, and offline leisure activities expenditure proportion of each consumer. This approach allowed for a comprehensive examination of how spending patterns influenced membership adoption across varying consumer behaviors. Results: Higher total spending and specific commerce platform expenditure proportion significantly increased membership adoption likelihood. General retail expenditure proportion did not show any significant effect on membership adoption. Bundled service expenditure proportion, such as digital contents consumption, also had no statistically significant impact. Offline leisure spending proportion positively influenced membership adoption during the post-pandemic phase but remained insignificant during the pandemic. The period-integrated analysis with interaction terms further examined this period-dependent effect of offline leisure spending proportion, revealing that the interaction term was not statistically significant at a conventional significant level while differences in leisure spending effects were observed between periods. Conclusion: This study offers significant academic and practical insights into e-commerce membership adoption. First, it shifts the analytical focus from the effects of membership adoption on consumer behavior to the influence of past consumer spending patterns on membership adoption, providing a novel perspective on the dynamics of the e-commerce industry. Second, it examines the role of bundled service in membership adoption, addressing an underexplored area in existing research and shedding light on the growing trend of service bundling across industries. From a practical standpoint, the findings enable e-commerce platform firms and marketers to identify high-value customers based on their spending patterns and design targeted marketing strategies to effectively encourage membership adoption.