Posted November 5, 2013 4:50 pm by with 0 comments

Tweet about this on TwitterShare on LinkedInShare on Google+Share on FacebookBuffer this page


Yelp knows they have a problem with fake reviews. They say they have it under control thanks to a sophisticated filter system which removes around 25% of all incoming reviews. But a new university study called “Fake it Till You Make It” (a weird name for a University study but whatever. . . ) pretty much says that as long as Yelp exists there will be fraud – not because people are mean and unethical – but because the incentive to get a high rating is worth the risk.

In addition to a study of positive fake reviews, Boston University’s Georgios Zervas and Michael Luca of the Harvard Business School studied incidents of negative review posting against competitors.

They did the math, (I don’t understand it but I’m going to assume they know what they’re doing) and found that an increase in local competition causes an increase in fraudulent negative reviews on Yelp.

“Overall, our results suggest that increased competition by similar, nearby restaurants has a statistically significant, economically substantive impact on negative review fraud.”

This was not true when a restaurant serving a different type of food opened nearby or a similar restaurant opened a significant distance away.

This also wasn’t the case for chain restaurants. They say that since chains rely on branding and promotion from the parent company, they’re less effected by online reviews and thus aren’t incentivized to post fake reviews.

What’s a fake review worth? Zervas told BU Today that an extra star on Yelp is worth a 5 to 10% increase in revenue for a small restaurant. That’s a big enough bump to make any business owner consider gaming the system.

As expected, the study found that bad reviews and only a few reviews both increase the incentive to post a phony positive review. It’s a numbers game. If you have 100 reviews, the level will find itself. If you have two reviews, one good and one bad, most people are going to believe the bad.

But as Zervas points out, the hardest thing about review fraud is separating the legitimate reviews from the fake ones. They started by reverse engineering Yelp’s own review filter so they could compile a list of qualities of a likely fraudulent review.

Then this happens:

mathAnd from that they figured out that you really can’t believe what you read on the internet.

The takeaway is that as long as people keep using sites like Yelp to make decisions about a business, businesses will continue to post fake positive reviews, slam their competition with phony negative reviews and blatantly encourage (coerce? pay?) customers to post their own “unbiased” review.

Zervas says reading between the lines will help you find the truth:

“I have a lot more faith in a business with 3½ stars and 100 reviews than I do in one with 4 stars and just 3 or 4 reviews. That’s common sense.”

He also notes that sometimes a bad review can be beneficial to a business. For example, if a reviewer complains about a hotel being unsuitable for children, that might be the perfect place for a mom looking to get away from it all.