Search papers, labs, and topics across Lattice.
This paper investigates the emergence of agentic recommendation markets enabled by LLM-based user agents, where users specify their needs prior to selecting a platform, leading to competitive dynamics among platforms. The study reveals that while this user-centric approach increases the pool of relevant items for comparison, it also creates a tension between access and attention, with platforms strategically manipulating rankings to maximize visibility. Key findings indicate that when user agents provide feedback on platform actions, the effectiveness of exposure improves significantly, suggesting that the design of recommendation systems must consider the interplay of access, attention, and accountability.
User-centric recommendation can increase relevant item exposure but often leads to platforms gaming the system to maintain visibility.
Online recommendation has traditionally taken place after a user enters a platform, which determines the candidate pool and the ranking shown to the user. LLM-based user agents enable a different recommendation process: a user specifies a need before choosing a platform, leaving platforms to compete for the user's attention, which we refer to as an agentic recommendation market. In our controlled LLM-based experiments across three product domains, we find this new setting of recommendation creates a tension between access and attention. Compared with traditional platform-centric recommendation, user-centric recommendation greatly expands the opportunity for relevant items to enter comparison; yet broader participation does not translate directly into effective exposure. Competition directly triggers platforms'strategic play: selectively positive explanations occupy 73--78% of first-ranked positions. When the user agent relates platforms'actions to subsequent user feedback, this share falls to 36--41%, while the chance of a user purchasing the relevant item increases. A user agent is therefore more than a ranker over a larger pool of candidates: its querying, ranking, and feedback mechanism governing who can compete, how scarce attention is allocated, and how earlier outcomes shape the evaluation of platforms directly affect user utility. Designing agentic recommendation therefore requires treating access, attention, and accountability as a joint mechanism design problem.
