Optimal Aggregation of Consumer Ratings: An Application to

Weijia Dai, Ginger Z. Jin, Jungmin Lee, Michael Luca

NBER Working Paper No. 18567
Issued in November 2012, Revised in November 2014
NBER Program(s):Industrial Organization

Consumer review websites leverage the wisdom of the crowd, with each product being reviewed many times (some with more than 1,000 reviews). Because of this, the way in which information is aggregated is a central decision faced by consumer review websites. Given a set of reviews, what is the optimal way to construct an average rating? We offer a structural approach to answering this question, allowing for (1) reviewers to vary in stringency and accuracy, (2) reviewers to be influenced by existing reviews, and (3) product quality to change over time.

Applying this approach to restaurant reviews from, we construct optimal ratings for all restaurants and compare them to the arithmetic averages displayed by Yelp. Depending on how we interpret the downward trend of reviews within a restaurant, we find 19.1-41.38% of the simple average ratings are more than 0.15 stars away from optimal ratings, and 5.33-19.1% are more than 0.25 stars away at the end of our sample period. Moreover, the deviation grows significantly as a restaurant accumulates reviews over time. This suggests that large gains could be made by implementing optimal ratings, especially as Yelp grows. Our algorithm can be flexibly applied to many different review settings.

download in pdf format
   (752 K)

email paper

Machine-readable bibliographic record - MARC, RIS, BibTeX

Document Object Identifier (DOI): 10.3386/w18567

Users who downloaded this paper also downloaded* these:
Mayzlin, Dover, and Chevalier w18340 Promotional Reviews: An Empirical Investigation of Online Review Manipulation
Lebergott Annual Estimates of Unemployment in the United States, 1900-1954
Chevalier and Mayzlin w10148 The Effect of Word of Mouth on Sales: Online Book Reviews
Celis, Lewis, Mobius, and Nazerzadeh w18590 Buy-it-now or Take-a-chance: Price Discrimination through Randomized Auctions
Nosko and Tadelis w20830 The Limits of Reputation in Platform Markets: An Empirical Analysis and Field Experiment
NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us