UCL School of Management is delighted to welcome Professor Xitong Li, HEC Paris, to host a research seminar discussing; Word-of-AI:” Revealing and Understanding the Match-Enhancing Effect of Generative AI Reviews Summary. 

Abstract:

Online platforms are increasingly integrating generative artificial intelligence (GAI) generated content, yet its impacts require careful investigation. This study examines how GAI reviews summary (GAIRS), a unique policy recently adopted by major travel platforms and shopping platforms, affect consumer-product matching quality. Three sets of cross-platform difference-in-differences analyses on review information for hotels, restaurants, and physical consumer goods demonstrate that GAIRS enhances consumer-product matching, manifested by increased average rating and decreased rating dispersion.

Two follow-up preregistered online controlled experiments provide direct causal evidence for the match-enhancing effect of GAIRS, as well as comprehension augmentation and consideration set expansion as two key mechanisms underlying the match-enhancing effect GAIRS. Specifically, the two experiments demonstrate that the presence (vs. absence) of GAIRS enhances consumer-product matching by enabling consumers to both gain a more nuanced understanding of product features (comprehension augmentation) and browse a larger number of products (consideration set expansion). This research highlights GAIRS’ potential as a novel GAI application for enhancing consumer-product matching on online platforms.