CITIC Securities: AI e-commerce cost reduction at the B2B end comes first; C-end transactions are still in the early stage.

CITIC Securities research report states that the industrial restructuring of AI + e-commerce follows a "back-to-front" penetration sequence, with B-end cost reduction and efficiency gains materializing first, while C-end transaction entry points remain in an early exploratory stage. Backend segments have clear assessment criteria and ample fault tolerance, and merchant AI tools on the platform side have already been implemented at scale; leading sellers are increasing full-chain AI investment, with advertising ROI and workforce efficiency optimization already initially reflected on the cost side, while also lowering the barrier to entry for small and medium-sized merchants, gradually returning the industry's competitive focus to product strength and fulfillment capabilities. On the demand side, the transaction scale of domestic AI entry points remains small, cross-platform price comparison lacks industrial soil, and the short-term industry landscape remains stable, dominated by platforms' own AI entry points. Overseas personal agents are rapidly gaining traction by relying on fragmented product discovery scenarios, but mainstream platforms generally adopt a closed strategy. In the long term, it is recommended to focus on three marginal variables: platform interconnection, AI entry point GMV share, and product data structuring. The competitive endgame will extend to the levels of product supply and platform rules.