Want to master the CMO role? Join us for GrowthBeat Summit on June 1-2 in Boston
, where we'll discuss how to merge creativity with technology to drive growth. Space is limited and we're limiting attendance to CMOs and top marketing execs. Request your personal invitation here
Have you had the feeling the quality of Groupon deals is going down? If so, you’re right. A report released this morning by Giorgos Zervas, a postdoctoral fellow of computer science at Yale University, shows that the average rating of a Groupon merchant before a deal runs is declining as Groupon matures. Zervas and other researchers had shown earlier that a business’s individual Yelp ratings dropped after running a Groupon. Yelp ratings by Groupon customers were, on average, 10% lower than those of their peers. (Disclosure: I met Zervas after the initial study and offered my theory that the average Yelp rating of merchants featured would be declining.)
Zervas looked at deals in New York, Boston, San Francisco, and Seattle. The analysis was based on nearly 12,000 deals that ran from January 1, 2010 to March 31, 2012.
“The slope of the trend line is negative in all four cases; indicating Groupon deal quality is, on average, decreasing over time,” Zervas wrote. “In all four cases, the correlation between time and expected Yelp rating is negative, and statistically significant.”
The study also looked specifically at low-rated deals, defined as those with three or fewer stars on Yelp. “Low-rated deals appear to be increasing in frequency,” Zervas wrote.
Zervas’ study shows that this is happening; it doesn’t offer a conclusion as to why it is happening. The study cites the following possibilities:
1. The population of merchants willing to run a Groupon deal remains, more or less, constant over time, but as Groupon is expanding the number of deals it offers, it has to work with some lower-rated merchants;
2. The population of merchants is changing; better-rated are merchants dropping out of running Groupon deals, and Groupon has to substitute merchants with lower Yelp ratings who are offering the sames kinds of deals;
3. Yelp ratings are naturally eroding over time.
As a long-time observer of both Groupon and the local-business ratings space, my assessment is that Groupon is suffering from what economists call an adverse selection problem.
In some ways, this is a problem that affects all businesses going after the local market. The very best restaurants never have to advertise. They are fully booked based on word of mouth.
I was talking to Bill Gurley of Benchmark Capital on Friday. He related a story of a locksmith he found on Yelp. (30:16 in the audio recording.) “He had 46 positive reviews,” Gurley said. “I call him, he comes to my house at 7 p.m. and stays there for three hours and changes every lock. And, of course, then I go write this huge positive review. He doesn’t need any other form of marketing. … He’s got a little mini network effect going.”
Although the best businesses had no reason to do any form of marketing, the next group of businesses who had to do some marketing considered Groupon because it was perceived as the next big thing. In the earlier days of Groupon, the company was highly lauded in the press. The horror stories of businesses with bad experiences weren’t out there. So this group of merchants would have been more likely to run Groupon when the risks weren’t as well known. This leaves the least successful businesses.
As a marketing company, this would be a problem. But it’s an even bigger problem for Groupon because Groupon also acts a financier in the U.S. and Canada. It provides money to merchants before all services are delivered. Groupon is, in effect, insuring the services of small businesses.
In economic terms, Groupon is the equivalent of a health insurance company that only insures cancer patients. That’s not a business that is going to be profitable in the long term.
Disclosure: I have short interests in Groupon and Yelp.
[Image credit: Ronald Sumners/Shutterstock]
VentureBeat’s VB Insight team is studying email marketing tools.
Chime in here, and we’ll share the results