Could AI agents reassemble shopping baskets across multiple retailers?

Bain & Company argues that the more consequential shift in agentic retail may not be which agent a shopper starts with, but what an agent assembles once it starts shopping. The report calls out "the emergence of agent-assembled baskets" — "multi-item, multivendor carts" that "would make agents far more valuable to shoppers with complex missions, such as planning a full family trip across lodging, transit, dining, and activities." Bain warns that "if agents can fragment the basket across vendors, the concept of a retailer-owned basket could start to disappear, especially in categories such as grocery."

The mechanism Bain gives is that multivendor baskets shift the unit of competition from the product to the basket itself, and in doing so increase price transparency across vendors at once. Bain states that "greater price transparency could intensify price competition and put further pressure on margins." Set against the three-agent taxonomy in the same report, this is a plausible endpoint of the third-party "objective" agent's logic: an agent that compares and recommends across vendors for a single purchase can, in principle, extend the same comparison to an entire basket — and at that point the retailer's basket is no longer the retailer's to control, but an assembly the agent performs.

This extends Why do consumers delay adopting agentic search?: that note's platform-revenue logic describes why an intermediary might want to control search, and Bain's basket-fragmentation claim describes what that control looks like downstream once the intermediary assembles the purchase itself rather than just the comparison. It also sharpens the stakes behind Will brands succeed by building their own AI shopping agents? — if the basket itself can be reassembled by an agent, being the agent of choice for a single product no longer guarantees the retailer keeps the sale. How are different AI agents reshaping the retail shopper journey? is the sibling note on the taxonomy this threat sits beside.

Bain frames this as one "bookend" of a range of outcomes and says explicitly, "we'll likely land somewhere in the middle" — the report does not claim multivendor baskets are already happening at scale, only that they are a plausible consequential shift retailers should prepare for. It names no specific agent or platform already assembling multivendor baskets, and gives no figure for how much margin pressure such fragmentation might produce; the claim is a strategic warning rather than a measured trend.