5 ms·

Edit: I would have loved to add to the conversation, but people do love their pitchforks and I should be working instead of commenting on HN. :-) Admitting you

by irskep 3y ago

Edit: I would have loved to add to the conversation, but people do love their pitchforks and I should be working instead of commenting on HN. :-) Admitting you worked for Yelp on HN: not even once

BonoboIO 3y ago
What can you say about these accusations brought up in this thread? https://news.ycombinator.com/item?id=36329248 https://news.ycombinator.com/item?id=36329248
FireBeyond 3y ago
> It's amazing to me how many people have conspiracy theories about Yelp reviews, when in reality it's just really hard to determine good vs bad actors in review systems and it's possible to do a bad job as an engineer. When cards are cloned, it's exceptionally easy for a bank to determine which merchant is compromised by looking for common denominators. Take three actors. What are the odds that two of them have reviewed one business, and only that business? Pretty high. All good. Same for all three. Now, what are the odds that two of them have only reviewed two businesses, and both businesses are the same, on both accounts? Still decent, sure. All three actors, same two businesses reviewed, and only them? Getting a little dicier. Two actors, seemingly independent, have each reviewed three businesses only. And they've reviewed the same three businesses, in the same order. And perhaps they each reviewed the first business, then two weeks later, the second business, and then the third. Odds? Getting pretty slim now. A third actor, doing the same? You're stretching the bounds of credulity for sincere reviews. This stuff is basic probability theory. Looking across reviews for sets and unions of reviews and reviewers, and increasing the number of reviews, but still seeing commonality between them points to SOME malfeasance worthy of review.
- emodendroket 3y ago
  I can see some ways that might go awry (what if I like reviewing every new restaurant that opens?). But more importantly, bad actors will figure out and work around each heuristic like this you make.
  - FireBeyond 3y ago
    Of course they will, but that shouldn't negate the need to identify and shut down interactions which are quite definitively problematic.
    - emodendroket 3y ago
      Another way of saying what I’m trying to say is that at any given time, no matter how much they keep heaping on new heuristics, there will be some missed cases that should be obvious to a human observer.
- mtlynch 3y ago
  This comment is fairly dismissive of GP's experience and intelligence. If you think you can solve a problem just by spitballing it, and GP's team at yelp couldn't figure it out for years, either you're much smarter than everyone on their team or you're underestimating the problem. I worked on an anti-fraud team in a similar role, and I agree with GP that it's a harder problem than people realize. It's easy to come up with a simple metric that would identify the fraud that you observe, but you don't know what false positives it would catch as collateral damage. Like this example that you find wildly implausible: >Two actors, seemingly independent, have each reviewed three businesses only. And they've reviewed the same three businesses, in the same order. And perhaps they each reviewed the first business, then two weeks later, the second business, and then the third. Odds? Getting pretty slim now. A third actor, doing the same? You're stretching the bounds of credulity for sincere reviews. I think it's highly likely that if you used this metric to find patterns, you'd find fraud, but you'd also find lots of legitimate reviews (e.g., friends who go to the same restaurants together, friends who recommend restaurants to each other).
  - kelnos 3y ago
    > This comment is fairly dismissive of GP's experience and intelligence. I don't have the full text of that comment, as they've edited it our, so I don't know specifically what they said. But perhaps they just don't know everything about how Yelp operates; it's not a small company where everyone knows everything. I also know someone who previously worked at Yelp, and they've talked about some of Yelp's shady internal practices. Everything is not all above-board there.
    - BizarroLand 3y ago
      I'm willing to bet there are more people who could solve that problem working for yelp than you would fairly estimate there are, but the suits are actively preventing solving that problem because they know that shaking down businesses for Yelp support contracts is infinitely more profitable than being a reliable source of truth in the world will ever be. Gold Plated Bugatti's do not buy themselves, after all.
- 8note 3y ago
  > When cards are cloned, it's exceptionally easy for a bank to determine which merchant is compromised by looking for common denominators. Is it actually? My bank calls me to ask which transactions were mine and which ones werent, thus requiring outside feedback
  - FireBeyond 3y ago
    That's your perspective. That's when -your- behavior doesn't match your previous behavior. What they are looking for in a different context is "Customer A reported unauthorized transactions, and so did Customers B through D. Each of these customers, prior to these reports, all had a transaction occur at XYZ gas station, so is there a skimming problem at that gas station?"
emodendroket 3y ago
Yeah, the only way you could have it work perfectly would be actually manually reviewing each review. Which would make the business model not work.
winrid 3y ago
It doesn't matter what you did as an engineer, and in fact automation isnt really a big part of the discussion. If sales people etc have special access to add or remove reviews, which they do, or at a minimum leave negative reviews themselves when you don't pay, it is corrupt. It is well documented at this point that they do this.
- emodendroket 3y ago
  That’s not what the article describes at all so I’m curious what documentation you are referring to.
  - winrid 3y ago
    There are so many people complaining about the sales people doing this stuff. https://youtu.be/C67Lh4LE5LY https://youtu.be/C67Lh4LE5LY
    - emodendroket 3y ago
      I would believe individual salespeople might do unscrupulous things like enlist their friends to leave bad reviews or make promises they can't keep about altering someone's score but that's a totally different accusation than saying Yelp as a company has a policy of doing this.
      - winrid 3y ago
        I doubt they explicitly have such a policy. That'd be suicide.

by irskep 3y ago

Edit: I would have loved to add to the conversation, but people do love their pitchforks and I should be working instead of commenting on HN. :-) Admitting you worked for Yelp on HN: not even once

BonoboIO 3y ago

What can you say about these accusations brought up in this thread? https://news.ycombinator.com/item?id=36329248 https://news.ycombinator.com/item?id=36329248

FireBeyond 3y ago

> It's amazing to me how many people have conspiracy theories about Yelp reviews, when in reality it's just really hard to determine good vs bad actors in review systems and it's possible to do a bad job as an engineer. When cards are cloned, it's exceptionally easy for a bank to determine which merchant is compromised by looking for common denominators. Take three actors. What are the odds that two of them have reviewed one business, and only that business? Pretty high. All good. Same for all three. Now, what are the odds that two of them have only reviewed two businesses, and both businesses are the same, on both accounts? Still decent, sure. All three actors, same two businesses reviewed, and only them? Getting a little dicier. Two actors, seemingly independent, have each reviewed three businesses only. And they've reviewed the same three businesses, in the same order. And perhaps they each reviewed the first business, then two weeks later, the second business, and then the third. Odds? Getting pretty slim now. A third actor, doing the same? You're stretching the bounds of credulity for sincere reviews. This stuff is basic probability theory. Looking across reviews for sets and unions of reviews and reviewers, and increasing the number of reviews, but still seeing commonality between them points to SOME malfeasance worthy of review.

emodendroket 3y ago
I can see some ways that might go awry (what if I like reviewing every new restaurant that opens?). But more importantly, bad actors will figure out and work around each heuristic like this you make.
- FireBeyond 3y ago
  Of course they will, but that shouldn't negate the need to identify and shut down interactions which are quite definitively problematic.
  - emodendroket 3y ago
    Another way of saying what I’m trying to say is that at any given time, no matter how much they keep heaping on new heuristics, there will be some missed cases that should be obvious to a human observer.
mtlynch 3y ago
This comment is fairly dismissive of GP's experience and intelligence. If you think you can solve a problem just by spitballing it, and GP's team at yelp couldn't figure it out for years, either you're much smarter than everyone on their team or you're underestimating the problem. I worked on an anti-fraud team in a similar role, and I agree with GP that it's a harder problem than people realize. It's easy to come up with a simple metric that would identify the fraud that you observe, but you don't know what false positives it would catch as collateral damage. Like this example that you find wildly implausible: >Two actors, seemingly independent, have each reviewed three businesses only. And they've reviewed the same three businesses, in the same order. And perhaps they each reviewed the first business, then two weeks later, the second business, and then the third. Odds? Getting pretty slim now. A third actor, doing the same? You're stretching the bounds of credulity for sincere reviews. I think it's highly likely that if you used this metric to find patterns, you'd find fraud, but you'd also find lots of legitimate reviews (e.g., friends who go to the same restaurants together, friends who recommend restaurants to each other).
- kelnos 3y ago
  > This comment is fairly dismissive of GP's experience and intelligence. I don't have the full text of that comment, as they've edited it our, so I don't know specifically what they said. But perhaps they just don't know everything about how Yelp operates; it's not a small company where everyone knows everything. I also know someone who previously worked at Yelp, and they've talked about some of Yelp's shady internal practices. Everything is not all above-board there.
  - BizarroLand 3y ago
    I'm willing to bet there are more people who could solve that problem working for yelp than you would fairly estimate there are, but the suits are actively preventing solving that problem because they know that shaking down businesses for Yelp support contracts is infinitely more profitable than being a reliable source of truth in the world will ever be. Gold Plated Bugatti's do not buy themselves, after all.
8note 3y ago
> When cards are cloned, it's exceptionally easy for a bank to determine which merchant is compromised by looking for common denominators. Is it actually? My bank calls me to ask which transactions were mine and which ones werent, thus requiring outside feedback
- FireBeyond 3y ago
  That's your perspective. That's when -your- behavior doesn't match your previous behavior. What they are looking for in a different context is "Customer A reported unauthorized transactions, and so did Customers B through D. Each of these customers, prior to these reports, all had a transaction occur at XYZ gas station, so is there a skimming problem at that gas station?"

emodendroket 3y ago

Yeah, the only way you could have it work perfectly would be actually manually reviewing each review. Which would make the business model not work.

winrid 3y ago

It doesn't matter what you did as an engineer, and in fact automation isnt really a big part of the discussion. If sales people etc have special access to add or remove reviews, which they do, or at a minimum leave negative reviews themselves when you don't pay, it is corrupt. It is well documented at this point that they do this.

emodendroket 3y ago
That’s not what the article describes at all so I’m curious what documentation you are referring to.
- winrid 3y ago
  There are so many people complaining about the sales people doing this stuff. https://youtu.be/C67Lh4LE5LY https://youtu.be/C67Lh4LE5LY
  - emodendroket 3y ago
    I would believe individual salespeople might do unscrupulous things like enlist their friends to leave bad reviews or make promises they can't keep about altering someone's score but that's a totally different accusation than saying Yelp as a company has a policy of doing this.
    - winrid 3y ago
      I doubt they explicitly have such a policy. That'd be suicide.