Working Papers
- “How Do Firms Advertise When Customer Reviews are Available: Theory and Empirical Evidence” (Job Market Paper)
(Presented at: IIOC 2015, the 8th Workshop on the Economics of Advertising and Marketing, EGSC Washington University in St. Louis 2015)
Online consumer product reviews have become very popular and influential in consumers’ purchase decisions. I study how competing firms choose advertising and prices when customer reviews are available and when firms may build up loyal customer bases. The model predicts that higher-rated firms are more likely to be dominant in advertising. In other words, online reviews complement firms’ advertising. I also analyze an extreme case of the model: an entry game in which an entrant and an incumbent interact. I find that the availability of customer reviews undoes the “fat-cat” effect of a big incumbent with a lot of loyal customers. An incumbent with a high enough ratio of good reviews can successfully deter entry and maintain a high profit. Comparative statics of the theory model can explain the pattern of advertising response to Yelp rating found in the empirical RDD paper. Intuitively, when the capacity limit of a local business becomes binding, a jump in the display rating will reduce the complementary effect of online reviews on advertising.
- “Advertising Response to A Better Online Rating: A Regression Discontinuity Design on US Local Restaurants” (updated version coming soon)
I analyze the advertising spending pattern of local restaurants with different online ratings on Yelp.com. Rating information on Yelp includes the display rating and the distribution of reviews. Surprisingly, although both types of rating information summarize to different extents how consumers like a restaurant, advertising spending responds in entirely opposite directions to changes in display rating and in average rating (i.e. a summary statistic of the distribution). Given the discontinuity in Yelp display ratings that is created by the rounding algorithm, I use an RD design to identify the effect of a higher display rating on local restaurants’ advertising spending decisions. I find evidence of a significantly negative effect of display rating on advertising spending for relatively higher-rated (i.e. rated above 3) restaurants. On the other hand, when the display rating is constant, the relationship between local restaurants’ ad spending and average rating is significantly positive. The reason for the opposite advertising responses to display rating and to average rating is the capacity limit of local businesses.
- “Delay in Adoption” (Joint with Marc Rysman) (pdf draft available upon request)
Our paper proposes a new explanation for adoption failure, or adoption delay, in markets with network effects. In our model, consumers and software providers play a dynamic two-sided adoption game, choosing between two incompatible platforms/technologies. Consumers are allowed to choose to wait in adoption, but firms need to adopt one platform upon entry. We show that, in a parameter space that only has standardization equilibria in a static setting, when we introduce dynamics, there exists a “market split and adoption delay” equilibrium. In this equilibrium,firms split between two platforms, and some consumers choose to wait in period 1 in order to join the turn-out-to-be-dominant platform in period 2. This “split and delay” equilibrium is inefficient since the market would benefit from immediate coordination on one platform or the other. Our model is motivated by the 56K modem market, in which competition between two similar technologies appears to have led to adoption failure, until an industry standard setting organization coordinated the market on an alternative standard.
