Search Refinement and the Design of Recommended Choice Sets
(A2026-132504)

AUTHORS

Chi-Ying Wang, Fordham University

ABSTRACT

Firms increasingly rely on recommendation systems and search refinement tools to guide consumers through large assortments. This paper develops a game-theoretic model in which consumers strategically decide whether to refine their search and firms design recommended choice sets. We show that refinement tools can either improve or harm both consumer welfare and firm profit, depending on preference uncertainty, prediction accuracy, and profit-margin imbalance. The model identifies when limiting refinement tools increases consumer welfare and firm profit.