Media Abuzz with New Marketing Research Results by Institute Fellow

in Institute News
April 16th, 2012

Could a Groupon deal ruin your business’s reputation online?

This is how a Huffington Post article started its discussion of the most recernt findings by Institute Fellow and Computer Science Professor John Byers, his student Giorgos Zervas (now a Postdoctoral Fellow at Yale), and his collaborator Professor Michael Mitzenmacher of Harvard University.

These new findings, made public in February, came on the heals of the researchers first study which made the news last fall, when Groupon was getting ready for its IPO. These new results found that Groupon subscribers are in fact more moderate reviewers on Yelp than the average consumer, giving out fewer 1 and 5-star ratings and more 3 and 4-star ratings than their non-Groupon counterparts. Byers and his co-authors theorize that “reviews from Groupon subscribers are lower on average because such reviews correspond to real, unbiased customers, while the body of reviews on Yelp contain some fraction of reviews from biased or even potentially fake sources.”

These findings make the results from the first study — suggesting that Groupon reviews hurt Yelp ratings — even more significant.

The research work by Byers and his collaborators highlights the role that “Big Data” plays in modern marketing research. To perform this study, the authors mined large data sets harvested from the Internet using innovative techniques.  Byers has recently convened a Hariri Institute Roundtable on “Big Data” to discuss opportunities for collaborative research at BU along these lines.