Strategic choices in live-streaming e-commerce: Balancing live-streaming selling mode and return-freight insurance strategy

Live-streaming commerce is swiftly rising. While collaborating with Key Opinion Leaders (KOLs) enhances sales through product information sharing, pricing constraints and high commission fees erode the e-tailer's margins. Currently, self-streaming powered by artificial intelligence (AI) offers cost-efficient alternatives by reducing product unfit risks. However, persistent mismatch concerns still make the e-tailer offer return-freight insurance (RI) popular. We explore how the e-tailer strategically selects live-streaming modes (top KOL, regular KOL, or self-streaming with AI) and RI strategies. Using a two-stage game-theoretic model, we have following key findings. Offering RI will increase the retail price, with KOLs increasing commission fees only when their role in reducing product misfit risks is limited. Compared to a regular KOL, the top KOL may leverage the bargaining power to drive down the retail price. Additionally, the e-tailer is most likely to offer RI when contracting with a regular KOL (due to limited misfit reduction) and least with self-streaming with AI. Crucially, offering RI can complement the top KOL collaboration but cannot substitute it. The findings further indicate that if unit product salvage is low, or unit salvage is high while the top KOL shows high fan effects, the e-tailer may prefer a top KOL; otherwise, choosing a regular KOL is the optimal solution. In particular, self-streaming with AI thrives if there are low AI adoption costs and weak KOLs' fan effects, especially when unit salvage is low. These results can help e-commerce platforms optimize commissions, enhance streaming partnerships, and improve AI tools for better consumer-product matching.