4 ms·

I infer from your comments that there would be no way to prove based on review display that Yelp was providing review relief if you pay for advertising, right?

by pgrote 14y ago

I infer from your comments that there would be no way to prove based on review display that Yelp was providing review relief if you pay for advertising, right? You basically have to trust their algo?

bickfordb 14y ago
The algorithm isn't published, but you can click "filtered reviews" on the bottom of the page and eyeball them yourself / decide for yourself
- ohashi 14y ago
  Which last I looked was super hidden and required a captcha to even read.
  - derwiki 14y ago
    The CAPTCHA is probably to prevent people from scraping unfiltered reviews and using unfiltered reviews to attempt to reverse engineer the algorithm.
seekup 14y ago
To your first question - By all means, you could prove it. Establish relationships with and systematically monitor the displayed and filtered reviews of a statistically significant sample of businesses. Record when they are contacted by Yelp sales reps, record their decisions on those calls, and classify their filtered review behavior before and after. This is doable, and it would be enough information to determine if there is actually a correlation in review filter activity to advertising offers. Go for it. As a former Yelp engineer, I can assure you that the answer would be negative (no correlation), but I invite you to give it a shot. To your second question, the answer is "yes". A spam fighter simply cannot publicly release their spam detection algorithm - it would make the system too easy to bypass. SEO black hats have been trying for a decade to get around Google's sorting algorithms, and then they cry foul when Google changes things up. The message from Google all along has been: Stop trying to game our system, and instead focus on creating good websites. Yelp's situation is no different.
- espressod 14y ago
  2 other data points from (disclosure) someone with ties to Yelp: 1. Ongoing independent research / A working HBS paper found no connection between advertising and filtering (see last page of: http://hvrd.me/ZNpuOy http://hvrd.me/ZNpuOy) [Why do we see advertisers are more likely to have their positive reviews filtered?] 2. A simple Google test like this: http://bit.ly/15QV1WS http://bit.ly/15QV1WS [Why are these advertisers not using their 'Delete negative review' button? Hint: because it doesn't exist. Which begs question -- If it doesn't exist, is there a rational incentive to sell it?]

by pgrote 14y ago

I infer from your comments that there would be no way to prove based on review display that Yelp was providing review relief if you pay for advertising, right? You basically have to trust their algo?

bickfordb 14y ago

The algorithm isn't published, but you can click "filtered reviews" on the bottom of the page and eyeball them yourself / decide for yourself

ohashi 14y ago
Which last I looked was super hidden and required a captcha to even read.
- derwiki 14y ago
  The CAPTCHA is probably to prevent people from scraping unfiltered reviews and using unfiltered reviews to attempt to reverse engineer the algorithm.

seekup 14y ago

To your first question - By all means, you could prove it. Establish relationships with and systematically monitor the displayed and filtered reviews of a statistically significant sample of businesses. Record when they are contacted by Yelp sales reps, record their decisions on those calls, and classify their filtered review behavior before and after. This is doable, and it would be enough information to determine if there is actually a correlation in review filter activity to advertising offers. Go for it. As a former Yelp engineer, I can assure you that the answer would be negative (no correlation), but I invite you to give it a shot. To your second question, the answer is "yes". A spam fighter simply cannot publicly release their spam detection algorithm - it would make the system too easy to bypass. SEO black hats have been trying for a decade to get around Google's sorting algorithms, and then they cry foul when Google changes things up. The message from Google all along has been: Stop trying to game our system, and instead focus on creating good websites. Yelp's situation is no different.

espressod 14y ago
2 other data points from (disclosure) someone with ties to Yelp: 1. Ongoing independent research / A working HBS paper found no connection between advertising and filtering (see last page of: http://hvrd.me/ZNpuOy http://hvrd.me/ZNpuOy) [Why do we see advertisers are more likely to have their positive reviews filtered?] 2. A simple Google test like this: http://bit.ly/15QV1WS http://bit.ly/15QV1WS [Why are these advertisers not using their 'Delete negative review' button? Hint: because it doesn't exist. Which begs question -- If it doesn't exist, is there a rational incentive to sell it?]