When Better Choices Feel Worse: The Paradoxical Effect of Personalized Product Recommendations on Objective versus Subjective Decision Quality
(A2026-134133)

AUTHORS

Gerald Häubl, University of Alberta; Benedict Dellaert, Erasmus University Rotterdam; Nahid Ibrahim, University of Leeds; Sarah Wei, The University of Warwick

ABSTRACT

Technological advances increasingly enable firms to personalize recommendations by predicting products’ attractiveness based on consumers’ idiosyncratic preferences. While such personalization aims to improve objective decision quality (i.e., selecting options that better match consumers’ preferences), prior work suggests that it may undermine their subjective decision quality, as reflected in lower choice satisfaction. We resolve this paradox by introducing a dual-pathway model that identifies two competing underlying mechanisms: while a more attractive top option increases satisfaction by enhancing set attractiveness, encountering options that are highly similar in attractiveness decreases satisfaction by making choices harder. Five experiments, including incentive-compatible product choice paradigms across diverse domains, support this framework. Our findings offer insights for designing personalized decision environments that optimize both objective and subjective decision quality.