This paper explores the impact of the “Verified Purchase” (VR) label on online review systems, with a focus on how platform-level policy changes influence both reviewer behavior and reader perception. The study is motivated by a common practice adopted by major e-commerce platforms over the past decade: tagging product reviews with a “Verified Purchase” mark to indicate that the reviewer has actually bought the product. While this mechanism was designed to increase transparency and trust, the actual effects of such a policy have not been fully understood. This paper addresses that gap by examining how the VR label shapes review content and reader response.

The authors take advantage of a natural experiment—Amazon’s introduction of the VR label in 2009—to analyze data collected from 2007 to 2011 across a wide range of product categories. Using a difference-in-differences (DID) approach, combined with textual analysis and machine learning tools, the study reveals nuanced behavioral changes among reviewers. Specifically, those who lacked verified purchase status began to write longer, more detailed reviews in an effort to compensate for their perceived lack of credibility. In contrast, verified reviewers tended to adopt more personalized and expressive language, signaling both their authenticity and individual judgment. This divergence highlights the adaptive strategies of reviewers in response to new platform norms.

The findings regarding reader perceptions are particularly striking. Contrary to platform expectations, the VR label does not universally enhance the persuasive power of a review. In fact, in some contexts—particularly when the review is highly positive—readers become more skeptical of VR-tagged comments. They may interpret these reviews as biased or overly favorable, assuming they reflect buyer loyalty rather than objective assessment. As a result, verified reviews often receive fewer “helpfulness” votes than their non-verified counterparts. However, in specific scenarios—such as reviews written by experienced users, for higher-priced items, or for newer products—verified reviews still carry greater weight, suggesting that reader trust in VR labels is highly context-dependent.

The most significant contribution of this study lies in uncovering a counterintuitive effect in platform governance: that transparency-enhancing features like VR tags can sometimes undermine trust rather than build it. The research shows a disconnect between how reviewers adapt their behavior and how readers interpret those changes, revealing a complex dynamic of expectation and perception. This insight holds valuable implications for platform designers, highlighting the importance of anticipating behavioral feedback loops and unintended consequences when implementing policy tools. It also enriches academic understanding of user-generated content ecosystems and the psychological mechanisms underlying digital trust.

About Xiaohua Zeng:

Zeng Xiaohua is a Tenured Associate Professor of Management at the Peking University HSBC Business School. She holds a Ph.D. in Marketing from the University of British Columbia, Canada. Her research focuses on social networks, user-generated content, and the impact of new technologies on marketing. Her work has been published in leading academic journals such as Marketing Science, Information Systems Research, Journal of Management Information Systems, and International Journal of Research in Marketing.