Search for a command to run...
In recent years, live streaming e-commerce has become an essential channel for brand merchants to expand online sales, with brand self-live streaming and influential streamer live streaming emerging as two mainstream cooperation modes. The choice between these modes directly affects brand merchants' profit levels and the efficiency of supply chain coordination. However, the mechanisms through which platform information sharing and traffic data investment jointly influence brand merchants' mode selection remain insufficiently explored. To address this gap, this paper constructs four game-theoretic models under scenarios of platform information sharing and non-sharing for both brand self-live streaming and influential streamer live streaming. The study systematically analyzes how factors such as demand information and traffic data investment affect the strategic decisions and profit structures of supply chain participants. The results show that: (1) the cost of traffic investment and the commission rate are key determinants of the brand merchant's cooperation mode preference-when both are high, adopting the influential streamer live streaming mode yields greater benefits for the brand merchant; (2) the platform's incentive to engage in information sharing varies significantly across cooperation modes, with a stronger motivation to share demand information under the brand self-live streaming mode; (3) and within a certain parameter range, the strategic interaction between information sharing and cooperation mode selection can foster a win-win outcome for both the brand merchant and the platform under the brand self-live streaming mode, thereby enhancing overall supply chain performance. This research elucidates the optimal cooperation mode selection logic of brand merchants in heterogeneous information environments and provides theoretical foundations and managerial insights for platform enterprises in designing effective information-sharing strategies and commission policies. To further verify the robustness and generalizability of the proposed models, several promising avenues for future research are also suggested.