Can 'Top Reviews' Save the Online Review Systems? Evidence from Empirical Analyses and a Quasi-Natural Field Experiment on Amazon
53 Pages Posted: 29 Jun 2018 Last revised: 29 Mar 2019
Date Written: June 21, 2018
User reviews are an important element in reducing market frictions, both online and offline. These reviews, supplied and later consumed by users, arguably constitute a market for information that has non-trivial implications. Naturally, platforms that host such content focus on continually improving the efficiency and effectiveness of such a market. One challenge, however, is the sheer number of reviews that subject users to information overload. Possibly, in order to facilitate efficient information consumption by customers, many retailers now prominently feature a selected set of reviews, primarily based on crowd feedback. In this study, we investigate the influence of these crowd-endorsed featured reviews, or “superstar” reviews, on Amazon in determining the efficiency of its user-generated information market. We also leverage a quasi-natural field experiment to unravel the nuances in gaining a richer understanding of the power of those “superstar” reviews. We show that “superstar” reviews supersede the signal that average rating provides and strongly impact the efficiency of the market, be it in the form of subsequent customer expectation or in the form of demand. We also show that the power of “superstar” reviews lies in parsimony, more of them is not better. This study contributes to the understanding of the functioning of crowd-sourced information market and the effectiveness of “superstar” reviews.
Keywords: Online Reviews, Information Overload, Featured Reviews, Online Retailer
Suggested Citation: Suggested Citation