Article Highlight | 3-Nov-2023

Five practices fight fake online reviews, build consumer trust

Penn State

UNIVERSITY PARK, Pa. — The right online review can win over a new customer or convince them to take their business elsewhere, but not every review is an honest product or service assessment. Bad actors looking to boost their sales have paid for fake reviews or posted the reviews themselves. A team of researchers, including from Penn State’s Smeal College of Business, identified five practices that online review sites like Yelp, Tripadvisor, Edmunds and Trustpilot can (and to varying degrees, do) implement to fight fake reviews.

The findings, published in the Journal of Marketing Research, have implications for consumers who turn to online reviews before making a purchase or booking a vacation.

“Misinformation and fake reviews have become an acute source of concern and consumers are well aware of the problem,” said co-author Stefan Wuyts, professor of marketing and director of the Institute for the Study of Business Markets at Penn State. “This creates a paradox as consumers increasingly rely on online reviews for most of their purchases, but their trust in online reviews is waning. We examined concrete practices that online review platforms can utilize to credibly serve as guardians of trust.”

Consumers can look for the following practices on online review platforms, according to the researchers:

  • Monitoring: Review platforms that use algorithms and human monitors to flag and remove fake reviews. Platforms may also implement generative pretrained transformer (GPT) detectors to identify reviews created by artificial intelligence models like ChatGPT. When review platforms clearly communicate their monitoring practices, they set expectations for sellers and consumers that the reviews that appear on their sites are authentic.
  • Exposure: Reputation matters in online environments. A review platform that publicly calls out firms that engage in fake reviews lets consumers know that the online review platform is looking out for their interests and monitoring for fakery. The threat to the company’s reputation, and its bottom line, disincentivizes bad actors.
  • Community building: Sites that encourage reviewers to interact regularly with consumers in question-and-answer or online forum formats. Such community building disincentivizes fake reviews and can increase user trust in the online platform.
  • Status endowment: Awarding online credibility badges to reviewers introduces competition to gain and maintain status. The competition for credibility badges encourages reviewers to conform to the social norm of posting genuine reviews.
  • Identity disclosure: Requiring reviewers to share their name, location and profile picture can increase perceptions of reviewer credibility. However, it also poses privacy concerns and has been linked to positivity bias, in which reviewers give higher ratings than they would have if they could post anonymously. The researchers found that implementing the previous four practices removes the need for identity disclosure while fighting fakery and increasing consumer trust in the review platform.

“We found in a series of studies that while monitoring and exposure are the most powerful practices for building consumer trust, also community building and status endowment help restore trust,” Wuyts said. “These results hold across studies, and most of these practices are even more important than review and reviewer characteristics. What is more, once these practices are implemented, there is no added value for review platforms to ask reviewers to reveal their identity. That is good news from a privacy point of view. As review platforms implement more of these practices, they experience stronger website performance such as organic site traffic — in sum, these practices benefit not only consumers but also review platforms and honest firms.”

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.