Skip to Content Skip to Footer
What Drives Herding Behavior in Online Ratings

What Drives Herding Behavior in Online Ratings

Sarang Sunder, Kihyun Hannah Kim and Eric A. Yorkston

Online ratings are a double-edged sword. On one hand, they help consumers make informed product quality inferences. On the other hand, they can induce biases with consequences for a brand. For example, the Game of Thrones last episode experienced a precipitous drop in ratings on IMDb and Star Wars: The Last Jedi tumbled in ratings on Rotten Tomatoes as viewers piled on to “troll” the entertainment and amplify the negative reviews.   

While some of the “snowballing” of opinions (online ratings) in the above examples may be attributable to “vote brigading,” a large portion of the issue lies with the way in which online opinion is generated. That is, although raters are all too willing to express opinions about products and services ranging from products on Amazon to even their own doctors, these opinions are not free from bias. In fact, a consumer’s opinion is quite easily swayed by others. However, the extent of and the contingencies that govern this herding remain largely unknown.


A new study in the Journal of Marketing investigates the extent of herding (i.e., how consumers may be swayed by others) in online ratings as well as the distinct herding effects produced by specific in- and out-group networks (friends and the crowd). Further, our research team uncovers key rater-level and firm-level factors that attenuate or amplify herding and examines the role of ‘mixed’ opinions (i.e., disagreement between crowd and friends) on the herding effect.

What we discovered through a rigorous analysis of data from an online community (44,108 board gamers rating more than 5,138 games published by 2,206 firms over a period of 10 years, which ≅2 million observations) is that a more nuanced view of social influence is required. Specifically, we found that:

  • Although herding exists in online ratings, the source matters. Consumers rely heavily on both their friends’ and the overall crowd’s opinions when rating products online. On average, crowds exert a stronger herding influence on the average rater.
  • Brands/firms can influence online opinion through their product portfolios. In this research, we show how firms can leverage themselves in online rating environments through their product line strategy. Firms’ product portfolio strategies act as proxies for firm competence and can significantly attenuate herding both from crowd (out-group) and friend (in-group) networks. That is, the marketing mix (product strategy) can counteract social influence in online ratings.
  • “Mixed signals,” or diverging opinions between reference groups, create herding and differentiation. Our research finds that divergence between friend and crowd ties can create herding and differentiation depending on the experience level of the rater. More specifically, we show that when the two reference groups disagree with each other, experienced raters coalesce more on friends’ ratings than with the crowd.

What do these insights mean for managers?

As an avenue for consumers to express their opinions and evaluations about products, online ratings have become a staple component of the customer experience. This research has important implications for online reputation management, online rating platform design, and product strategy.

Online Reputation Management

Our study suggests that firms can and should take advantage of herding in rating environments. Even in the absence of conventional advertising strategies, firms can leverage reputation effects in ratings by strategically targeting their review solicitations. Further, managers must be careful because herding influences are a double-edged sword. Positive word of mouth results in more word of mouth that is positive, but the reverse is also true.

Online Rating Platform Design

Our research has implications for online rating platform design. Depending upon prior product ratings and the goals of the website, friend and crowd information can be made more or less accessible. Specifically, we found that the herding effect depends on rater experience or the lack thereof. Firms should adopt a “targeted” perspective for review solicitations to get less-biased evaluations.

Product Strategy and Portfolio Planning

This research provides guidance for portfolio planning. We discovered that a firm’s product strategy could have profound positive effects on online opinion, both directly and indirectly. Firms with broader and deeper product portfolios are viewed more favorably and can counteract herding influences from both crowds and friends.

In summary, when it comes to online opinion, all herding/social influences are not equal. Context and contingencies matter!

Read the full article.

Read the authors’ slides for sharing this material in your classroom.

From: Sarang Sunder, Kihyun (Hannah) Kim, and Eric Yorkston, “What Drives Herding Behavior in Online Ratings? The Role of Rater Experience, Product Portfolio, and Diverging Opinions,” Journal of Marketing, 83 (November).

Go to the Journal of Marketing

Sarang Sunder is Associate Professor, Indiana University, USA.

Kihyun Hannah Kim is Assistant Professor of Marketing, Rutgers Business School, Rutgers University.

Eric A. Yorkston is Associate Professor of Marketing, Neeley School of Business, Texas Christian University.