【公开学术报告-第145期】Humans Favor Messages that AI can Predict

发布时间:2025-09-29

Humans Favor Messages that AI can Predict

Speaker: Dr. Andrew Meyer (CUHK)

Date & Time: Tue. 28, October 2025, from 10:00 to 11:30 (Beijing Time)

Place: Tongji Building A2101 

ABSTRACT

We exploit a fundamental similarity between human reasoning and large language model inference to quantify the word-by-word predictability of written text. We then show that consumers respond more positively to marketing materials with predictable text. First, we examine a dataset of click-through rates from 65,362 article headlines that were assessed over the course of 17,802 A/B tests and more than half a billion user impressions. We show that predictable headlines have higher click-through rates, with economically significant effect sizes. Next, we report process evidence from three Prolific studies. They show that participant judgments of “strangeness” are elevated for unpredictable text. We conclude by examining 226,238 hotel reviews from Tripadvisor.com. We show that predictable hotel reviews are more likely to be voted helpful. In all cases, our results are robust to controlling for various other text features, including length and traditional measures of reading fluency and sentiment.

Keywords: processing fluency, large language models, headline click-through, consumer review helpfulness

GUEST BIO:

Andrew Meyer is an academic with expertise at the intersection of marketing, psychology, and business.

Since 2023, he has held the position of Research Assistant Professor at The Chinese University of Hong Kong Business School. Prior to this, he served as an Adjunct Lecturer at the University of Pennsylvania from 2021 to 2023. His postdoctoral experience includes roles as a Postdoctoral Fellow at the University of Chicago Booth School of Business (2017–2020) and a Postdoctoral Associate at Yale School of Management (2016–2017).

In terms of education, Andrew Meyer earned a Ph.D. in Marketing from Yale University in 2016 and a B.A. in Psychology from Wesleyan University in 2008.

联系方式

地址:上海市四平路1500号同济大厦A楼21楼 | 电话:021-6598 1341

同济大学 版权所有