Examining the Impacts of Airbnb's Review Policy Change on Listing Reviews

Abstract

In July 2014, Airbnb, one of the biggest firms in the sharing economy, decided to change the way the guests and the hosts review each other on the platform. Before this change, the guests and hosts were able to post reviews about their experiences asynchronously - the host/guest would be able to see the other party’s review whenever it was posted. The new review policy though, rolled out a simultaneous review system in which reviews are viewable only after both the host and the guest posting their own reviews. Our aim in this study was to examine the impacts of this new review policy on guest reviews.

Using Regression Discontinuity Design, Panel Data Analysis and a variety of techniques in the text analytics domain, we discovered that the new review policy made the reviews more diverse in terms of topics discussed in the reviews, lengthier, more objective, less positive, and more diverse in terms of their sentiment. Interestingly, we also discovered that the review policy had a long-lasting impact on the constructs related to personal opinions (i.e., sentiment and the diversity of sentiment in the reviews). However, the impact of review policy on informational content (i.e., diversity of topics, depth, and objectivity of the reviews) was short-lived.

Keywords: Sharing Economy, Airbnb, Online Reviews, Regression Discontinuity Design (Rdd), Panel Data Analysis, Time Series Analysis, Text Analytics