Abstract
Online review platforms like Yelp are an increasingly dominant force in consumers’ decisions to patronize a restaurant. A slew of negative star ratings can ultimately force a restaurant to close. However, little is known about the cultural mechanisms that explain the influence of online ratings on business success. One candidate is the Matthew Effect, a phenomenon whereby initial advantage uniquely predicts long-term success. Using a large dataset of US restaurants and their reviews on Yelp (Nrestaurants = 52,286, Nreviews = 3,355,714), we tested whether: (1) star ratings on Yelp are associated with the success or failure (measured in terms of closures) of restaurants; (2) whether this relationship is independent of content-level features of reviews, such as sentiment; and (3) whether early initial advantages in terms of star ratings would predict continuing operation versus closure years later. Using hierarchical logistic and autoregressive models, we find evidence supporting all three hypotheses. We take these results as evidence that online review platforms play a role in explaining success in the US restaurant sector and that this relationship is partially determined by a Matthew effect. We make recommendations for restauranteurs hoping to maximize success in this volatile sector.