Hacker News Reader

I think the challenge with finding firsthand reports on HN is that the owners of small brick-and-mortar businesses that Yelp is alleged to exploit are not well represented in this community.

Having said that, I have firsthand experience in having my own review of such a business censored:

6 months ago I wrote a review for the RazzleDazzle Barber Shop in Riverview, FL. Great little shop, I was surprised that it had zero reviews and was gray in Yelp. A month later, I returned for my second haircut and asked about business. I heard the same story about Yelp pressuring for sales, and mentioned that I had already written a 5-star review.

Upon checking Yelp, I noticed that the shop was still gray with 0 reviews, despite having left one a month ago myself. After carefully reading the site, I found that the keyword is "0 Recommended Reviews." At the bottom is a gray link that leads to more than a dozen reviews that are "Not Recommended", all 5-star, including my own. I wasn't the first to review the shop, the other reviews were hidden.

I wrote Yelp support to ask why this review was censored, despite my other reviews for other businesses being accepted. They assured me that an algorithm decides what reviews are recommended for protection of customers and that's all that's happening.

However, within a day my review was the only recommended review, and the barber shop was live on the site instead of gray. A few other reviews have been accepted since.

Go have a look for yourself, scroll to the bottom of that shop and click the "x Reviews that are currently not recommended" link in gray at the bottom. You will even find my original 5 star review, so perhaps they made a copy of it to get it to show.

Curious! I'm looking at it and don't see a duplicate review in the "not recommended" section. I do see one trend which matches my experience with "not recommended" that I mentioned in another comment: 10 out of 13 reviewers have no connections on yelp. It seems like the account's social network is one heavy factor in deciding whether to recommend a review or not.

There is one curious review that shows up in the normal sort that looks like a "follow up" review despite both version of it being identical.

It's their anti-fraud system. Reviews get moved to Recommend after the user passes some minimum anti-fraud Activity threshold.

If that was your first review, it's almost always unrecommended at first. This stops a lot of the "ask your friends and families" for 5 stars sorts of reviews.

The company I work for advertises pretty heavily on Yelp for our 150 locations, and the main perk isn't better ratings but having your bad reviews hidden behind a different sorting system.

If you go on any Yelp business page today you will see that it isn't sorted by Date, it's sorted by "Yelp Sort". So if my business doesn't advertise on Yelp then maybe bad reviews will be on the first page of a Yelp Sorted ranking. Once I starting paying they're typically hidden. Now if you manually go and and sort by Date they will still be there, they're not deleted and my star rating doesn't change - so I guess you're technically right that I don't get better ratings. I do think that most people don't realize that the most recent bad reviews aren't necessarily as visible though if you advertise with Yelp.

Is the extortion scheme as simple as that? Paying businesses get "Yelp Sort", non-payers get "Default Sort"?

If so, that answers a lot of questions.

No. I believe every page defaults to the same sort.

They default to the "same" sort, but that sort acts differently depending on whether you've given Yelp money.

That's not far off of how (Yelp's partner) Grubhub operates, where the default sort for restaurants is in descending order of the cut they give Grubhub.

From my experience, the reviews filtered out are often reviews by folks with small or no networks. I saw my wife's reviews filtered out when she first had an account, before she added friends or connected the account to facebook (not sure which one did it, this was years and years ago). I'm not saying this is a good system, just my observation of one factor.

Yea I'm sure brand new users have lesser ranking on Yelp Sort. My experience is just from talking to the Yelp Account Managers and seeing our bad reviews be Page 1 one month, and pushed to a further page the next month.

Except in this case there's a million anecdotal stories about Yelp doing this. I've been in the restaurant industry a long time, I've taken calls from Yelp sales agents, they're shit. And their sales pitch basically is extortion (pay us to hide your bad reviews, or else). I've seen owners straight up tell them to fuck off.

Thankfully they have zero traction here; my last place had a 4.5 rating on Google with 250+ reviews, and a 3.3 rating on Yelp with about 10 reviews.

JFYI:

http://www.bottobistro.com/YELP.html

This is hilarious and great. The whole website is really funny.

I do not have enough thumbs to adequately show my support for this.

This is where the nuance of their scumminess comes in and where you're either ignoring or are ignorant of it. Yes, they can say that they're not "deleting" or "altering" reviews because that's not what they're doing. What they're doing is sorting by "Yelp Sort" and listing "Recommended" reviews. If you look at any Yelp business page, you'll see a tiny little link at the bottom that says "See reviews that aren't Recommended". They claim that these reviews are hidden behind this wall because they're from users whose reviews are biased or who don't have enough patterns to be "trusted" but this is literally just their way of hiding the reviews that people don't pay for. Since every single one of those terms is subjective, they can change what constitutes a "recommended" review and be completely subjective about why one review is "trusted" and another isn't. It's completely up to them and they totally line up with whether or not a company pays for Yelp's "services".

It's a lot like the BBB. No one has to pay to be listed by the BBB but you can pay for the BBB to "arbitrate" disputes between consumers and your business when someone posts a strike against you on their site. You're just paying for them to contact the customer and hide their bad review.

> or are you suggesting they're getting away with lying on official communication?

They've probably carefully avoided lying while still being as misleading as possible.

> If yelp is altering reviews for paying clients, please show the evidence

yelp may not alter the reviews themselves, but it does change the sorting of reviews and images for paying customers.

Also the willingness/capability to carry through on the threat used in extortion isn't directly required, just the threat itself. The execution of the threat is only important when you want the consequences of non-compliance broadly known.

> I suspect, at worst, that they either misunderstood the sales pitch or were contacted by a scummy SEO-like company.

Really? Yelp is a business which provides little to no customer support bandwidth. Sales is easiest point of contact for Businesses with questions about reviews. Do you really think none of those sales people might have deliberately mislead businesses about impact of purchasing advertising on reviews? It may or may not have been condoned by Yelp, but it almost certainly happened.

For specific a specific claim that seems to provide evidence that something weird is going on:

> And what is indisputably clear is that Yelp’s claim that 75 percent of the reviews that get posted by Yelp users become recommended reviews is wildly inaccurate in the case of Baraka Café. When Yelp was only displaying 16 reviews and filtering out 30, it was displaying roughly 35% of all the reviews of Baraka Café, and suppressing 65% of them. It recommended less than half the claimed figure of 75%. Most importantly, none of the suppressed views were negative, only positive reviews were suppressed.[1]

[1] https://thetechnoskeptic.com/yelp-extortion-starring-role/

> Do you really think none of those sales people might have deliberately mislead businesses about impact of purchasing advertising on reviews? It may or may not have been condoned by Yelp, but it almost certainly happened.

I really don't disagree with this having happened or it being possible that the culture/leadership turns a blind eye to it. My issue is that there is this oft-repeated claim that paying clients' selection, ranking, or content of reviews is somehow different from non-paying ones. It shouldn't be hard to study this claim and yet there is no evidence of it being true.

I've been to Baraka Cafe. It's great. Take a look at their "not currently recommended" reviews. As has been mentioned elsewhere, almost all (8 out of every 10) reviewers have 0 friends on yelp. Active accounts' reviews get preferential visibility.

Perhaps if I start a business and tell all my friends and family to create accounts and review it with 5 stars, I too will have evidence of yelp "altering" reviews when theirs are inevitably hidden for looking suspicious.

Don't want to edit my previous comment, but I just realized that you're listing one particular scam that you allege they're not running, while they're running another totally different shitty scam.

This is the equivalent of "why are you mad, I didn't take your wallet" as you're riding off on my bicycle.

It's not Yelp directly, it's their outside sales contractors. Anyone who manages the online presence of a small business can contribute corroboratory anecdotes, including me. Yelp encourages this behavior from their contractors.

I'm sorry that no one has formally studied Yelp's shady business practices, because that leaves us without data you find worthy of consumption. Maybe an enterprising law firm will change that once they smell enough blood in the water.

Until then, consider that swaying someone's opinion on Yelp using their own PR resources is a complete waste of everyone's time.

The practice of using a contractor to avoid liability really speaks to the fact that they know what they're doing is wrong.

> www.big-bad.wolf

> Does the Big Bad Wolf blow down houses?

> "Of course I don't!"

In that story, the piggies couldn't file class action lawsuits against the wolf. If sites could lie without consequence in official business capacity, why would the average privacy policy take 18 minutes to read?

If you opened the link, you'd also see a link to an independent scientific study on the subject: http://people.hbs.edu/mluca/papers%20on%20ris/fakeittillyoum...

Perhaps the issue is more with folks falling for conspiracy theories and urban myths without investigating the claims in an unbiased manner.

> In that story, the piggies couldn't file class action lawsuits against the wolf. If sites could lie without consequence in official business capacity, why would the average privacy policy take 18 minutes to read?

I'm low on sleep today, so I'm honestly having a hard time parsing this argument. Are you saying that Yelp and other businesses are sufficiently afraid of class-action lawsuits that they're not willing to do shady things? If this were the case, we'd reach one of those dumb school-book economic equilibriums where no class-action lawsuits would take place. Businesses evaluate the risk of doing things, balancing profit against loss. Exposure to a lawsuit is just another number in that calculation.

> independent scientific study

That study is looking at whether fraudulent reviews are being added to the platform, not whether Yelp is calling up businesses and offering to remove bad reviews in exchange for money, which was my understanding of the original complaints against Yelp.

>I'm low on sleep today, so I'm honestly having a hard time parsing this argument. Are you saying that Yelp and other businesses are sufficiently afraid of class-action lawsuits that they're not willing to do shady things?

Businesses doing shady things will dance around the allegation of shadiness with clever language, legalese, and PR-talk to avoid saying the truth. They will rarely intentionally give an unequivocally simple and straightforward answer when there is some truth the allegation.

Or... just informally incentivize employees to do the wrong thing and make sure they have (or refuse to remove availability of) the tools they use to do so.

If yelp has salespeople that can alter reviews or how they show in any way, and those salespeople are incentivized to hit sales numbers, some subset of them will use the tools at their disposal to facilitate their sales.

In this theoretical scenario, is Yelp extorting businesses? No, it's just a few bad employees with an outside influence (and outsize sales numbers likely, too). But if Yelp knows about this and does nothing about it? It's gets more complicated then, and almost impossible to prove.

That's just one possible scenario where Yelp's statement is true, and the people reporting extortion are also being truthful, and depending on the details, Yelp is or isn't complicit in some way. A simple statement from either side doesn't really eliminate myriad possible situations like this, but given the number of reported instances of businesses having problems, I'm not so willing to discount that something is going on.

>If yelp has salespeople that can alter reviews or how they show in any way, and those salespeople are incentivized to hit sales numbers, some subset of them will use the tools at their disposal to facilitate their sales.

I've been threatened by SEO sales people that they can hurt my site's search rankings if I don't sign up for their service. There are robocallers[1] that pretend to be Google employees making these threats. Does that mean they have direct access into Google's database, or even the power to fulfill their threats? Do you think that local businesses, often run by much less technically-literate people, aren't hit by the same kind of scam for their online reviews?

[1]:https://www.zdnet.com/article/google-robocall-scam-were-suin...

I'm sure they probably are at some times. My only point was to say that there are situations where a statement from Yelp to the contrary doesn't really mean something isn't happening, with or without Yelp's informal approval, so it doesn't necessarily put the discussion to bed. It's just one more point of reference, and should be seen as such when people are making their own assessment of the situation.

Someone hearing complaints of Yelp coercing people shouldn't immediately believe it or believe it explains the whole situation, and neither should someone hearing Yelp's response. They are both possible insights into the actual reality of the situation, which may accurately describe part or al of that reality to varying degrees, and should be seen as exactly that.

We're getting off topic; are you suggesting that the people calling small businesses claiming to be Yelp employees are lying about who they work for?

This HBS Associate Prof just seems like a pro-yelper. He has another paper about how ads on yelp really work!

Hmm, I wonder why Yelp would keep mentioning him.

https://blog.yelp.com/2016/10/academic-study-shows-yelp-ads-...

Fascinating. I totally fell for the drama and didn't realize it. Thanks for posting that link.

A good reminder to be critical about what you hear.

I don't use Yelp so my opinions are both irrelevant and formed entirely by what I've heard from others. I'm not sure I believe Yelp hides/shows reviews based on whether businesses are willing to pay but I'm not sure I put too much weight in the above posted link either.

If Yelp's review algorithm has nothing to do with advertising spend then it should be fairly easy to have someone from PwC/Deloitte/E&Y look at the algorithm and issue a letter saying advertising spend isn't a factor. I'd give a lot more weight to a third-party review than I do the Buzzfeed article (not Buzzfeed News) and academic study on the ethics and incentives behind review fraud that Yelp has cited.

In my experience audits done by big 4 are worth nothing. Irregularities are raised, and handled, unofficially, so when the report comes, everything is fine, everything’s legit.

That’s what you get when auditing companies also do consulting.

Thanks for sharing this. I use Yelp for client lunches because I rarely know the area well, and it easily identifies restaurants that deliver. I always felt kind of slimy because of anecdotal rumors. I may have even perpetuated those rumors in passing. Now I feel a bit better, and I have some resources to back it up.