Hacker News Reader

Comment by pbreit | Hacker News Reader

The article could have revealed that of the 18 "hidden" reviews, 14 were the only review by the Yelper and 11 were in a one week period. These are two of the main criteria for Yelp's filtering.

I'm surprised people believe that Yelp would be so clumsy and unethical to simply penalize non-advertisers.

View on HN

There is certainly a ton of press around this coming from several reputable media outlets for months. Google for 'yelp extortion'. So you shouldn't be surprised. This is a real problem for them. At minimum, here is a documented case of legitimate reviews that simply aren't shown. They are basically saying 'The computer did that, so fuck off' and end up coming off like a bully, which by all measured terms they are.

This meme hit pretty big a year or two ago, a bunch of lawsuits were filed and nothing came of any of them. Yelp is usually able to demonstrate in each case that what it claims is happening is accurate (advertisers are buying a specific bill of goods, review flagging is systematic, etc.).

I'm surprised people think they wouldn't. They have the perfect hideout. Whatever the case they can always blame it on the "algorithm," and short of auditing their source code, nobody can really prove them otherwise.

Not that perfect of a hideout. I'd love to see someone collect data for cases where they think Yelp is intentionally manipulating reviews and see how many cases aren't easily explained by good filtering criteria.

[Full-disclosure: I'm friends with a good number of Yelp engineers who I expect would leave rather than take part in these such things, but I don't know any more about their codebase than any other non-Yelper.]

Reply on news.ycombinator.com