Asymmetric Network Effects in the Integration of Digital Platforms

2021 
We have witnessed an increasing number of mergers and acquisitions (M&A) between two-sided digital platforms, generally involving a unidirectional resource transfer policy from the acquired to the acquiring platform. This scenario may incur an asymmetric network effect that has not been explored in the extant literature. Using a unique dataset from Dianping and Meituan, two digital platforms similar to Yelp, we explore whether and how the unidirectional data integration policy causes a positive network effect, as reflected by the increments of user engagement on each side of the M&A. We measure user engagement in two dimensions, namely, quantity measures (e.g., number of total reviews) and quality measures (e.g., number of reviews with pictures and length of review). The results show that the policy enhances both measures of user engagement for the acquired and acquiring platforms but that its effects on user engagement on the acquiring platform are larger than those on the acquired platform. We propose three mechanisms: the network size effect, self-image management effect, and social learning effect to explain the possible impacts of platform integration on user engagement, and we conduct single-group interrupted time-series analyses to test the effects. We further conduct three sets of supplementary analyses. We find that, first, the policy not only expands its user base but also boosts the contribution enthusiasm of current users. Second, we clarify how the benefits of platform integration are shaped by merchant features such as per capita consumption and whether a merchant has chain stores. Third, we focus on the multi-homing phenomenon of merchants and find that multi-homing merchants benefit more from platform integration. The findings of this study deepen our understanding of the connections between unidirectional resource transfer and the asymmetric network effect in the context of the integration of two digital platforms.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []