The Oracles of Online Reviews

Abstract

Online reviews have become an essential part of consumers' search and purchase processes. In predicting purchase outcomes and market demand, the online reviews literature has largely focused on understanding review characteristics and characteristics of a product's full body of reviews. Individual online reviewing patterns, a significant aspect of the modern consumer's behavior, remain relatively understudied. We help to fill this gap in the literature by identifying a segment of online reviewers who systematically review successful products soon after product release. Using unique data of 1.4 million Amazon reviews and sales for 74,113 print books, we create a reviewer score that measures a reviewer’s tendency to review successful books and identify oracle reviewers whose early reviews serve as a signal for various measures of future book success, including life-time revenue. We find that it is oracles' self-selection, and not their taste for popular books, that underlies their predictive signal. Text analyses reveal that oracles write longer, more analytical, outward-focused, and overall helpful reviews relative to other reviewers, possibly explaining their ability to predict market demand. Additionally, male reviewers are more likely to be oracles. Our findings bear important implications for platforms, namely in redesigning reviewer-loyalty programs and leveraging content generated by predictive reviewers. Reviewer-loyalty programs, such as Amazon Vine, incentivize frequent reviewers in order to increase the availability of user-generated content on the platform. Our findings suggest that platforms should tailor designated programs for oracle reviewers, that will maintain the self-selection that underlies their predictive success signal but at the same time allow the selling platform to leverage the early signal, for example by featuring the to-be successful products on its homepage or by improving inventory management.

Keywords: online reviews, predictive consumers, product success, digital book sales