Hacker News new | past | comments | ask | show | jobs | submit login
Some Amazon Reviews Are Too Good to Be Believed – They're Paid For (npr.org)
455 points by EnderWT 7 months ago | hide | past | web | favorite | 265 comments

I've been burned multiple times by this; the worst was a pair of remote control off-road trucks for my sons' Christmas presents. They had hundreds of 5-star reviews and both units worked for maybe 5 minutes each. Within days of Christmas, the listing was flooded with 1-star reviews from other conned parents with very similar stories.

The number of products impacted must be far over the 1% claimed. There are whole categories that have been ruined by Fake China and kids' toys is probably the worst. Headphones, as mentioned by another commenter here, are another.

This problem is quickly destroying Amazon's credibility with a lot of customers. I went to a Best Buy for the first time in a decade to buy a pair of iPhone headphones, just so I'd be assured a genuine product. I now buy my kids' toys at Target, just so I can be assured that there's a reasonable chance that they aren't crap.

> This problem is quickly destroying Amazon's credibility with a lot of customers

It has totally destroyed their credibility with me personally. I used to be an avid Amazon shopper. It was just too easy. I now actively find stores locally to shop. Does it make a difference to Amazon's bottom line, not in the least. While it might be less convenient, I feel less dirty than supporting Amazon. There are times where local stores do not carry something, and I still have purchase from Amazon but only if I can't find it anywhere else first.

I still buy a lot of stuff from Amazon but I have found that I cannot trust their review system at all. So all research is outside Amazon or else I'm basically treating it the same as Alibaba.

The thing that utterly destroys credibility in my eyes is that fraudulent sellers are able to target name brand listings by offering fakes at a slightly cheaper price than the original manufacturer. There are apparently some mitigations against this but their use must not be widespread among sellers or something because I've been burned by this.

I can handle getting some crappy set of headphones but a big let down for a child's Christmas is just so upsetting to a point that maybe I've got unaddressed issues. I have a toddler and I think I'll gladly pay the premium to buy Toys R Us stuff I can touch and play with in the store.

> Toys R Us

Um, I have some bad news for you.

They seem to be strong here in Canada. I hope that lasts. Walmart is the only other option and it's not really an enjoyable experience.

Hilariously, ToysRUs are still big in China. Lots of kids and parents shopping. How did the vultures only get the US stores?

They got the UK stores too. :(

What I always disliked about toysrus was that no matter if you went at the quietest time you had to queue forever at the checkout. There'd be just the one till open and they'd only open a second one if people were getting restless.

> pay the premium to buy Toys R Us

Toys R Us is NOT "premium". Even back when it existed a few years ago.

Melissa and Doug is "premium" for toddler toys. Spend a little bit more, and you get quality toys that is even a bit fun to use as an adult.

Plenty of those stores around, and their webstore (https://www.melissaanddoug.com) is well featured.

Based on my experience with buying kid's toys... Toys R Us failed to stock good enough quality for me. If I wanted crap toys, Walmart and CVS had plenty. Target's selection of toys is mediocre, but I find a few dolls here and there worthwhile.

A "major" gift is almost-always a Melissa and Doug splurge moment.

I've never trusted Amazon for kids toys either. But Toys R Us was also in a similar boat for me of low-trust low-quality products.

As of the last decade, the quality of products from Toys R Us wasn't that great either. If there was a name brand they had it and also a cheap inferior clone. If no name brand existed then it was just inferior junk. The only real exception was there stuffed animals which were all pretty nice.

Water toys were mostly all garbage.

My approach has been (1) to filter the reviews by critical reviews and read them quickly - that has tended to show the real problem that the shallow 5-star (or 1-star) reviews hide. The aggregate rating (4.5 stars) is totally useless now. I also look at the product's fakespot rating when I remember to do so.

1: when I have to buy from Amazon, i.e. none of my other options such as Walmart (90 day returns, fast shipping with no membership), Best Buy, etc. have the item

Ever since BestBuy has started price matching amazon, I have started going there a lot. And the purchases have been significant: a modem/router, a laptop and a couple of fitbits.

Nothing better than being assured that the product you receive is genuine and well priced.

The thing is, you really can't buy from anyone but Amazon directly anymore. The marketplace is flooded by chinese off brands that break the moment you touch them. At least that's the situation here in Europe.

Don't know which European country and which product category you're thinking of, but e.g. for tech that's not true. There are plenty of reputable tech online stores in Germany (e.g. Mindfactory, Alternate, Reichelt, Cyberport, Arlt in no particular order). I expect all of these to be able to ship to at least all of the continental EU.

What I meant with "can only buy from Amazon directly" is that I don't use Amazon's marketplace features. Of course, OUTSIDE of Amazon there is competition especially in the electronics category. I actually usually order my electronics at Alternate, if their prices aren't way above market which sometimes seems to happen with them.

Is there a way to actually do that for most stuff? I actually want to buy some real Apple EarPods and the first 30 results seem to say supplied by some random outlet for about 25% the regular price. Dunno if there's a only show direct from Amazon button somewhere?

Of course there isn't, they would lose a lot of sales that way. For the Apple EarPods, they've been out of stock for months now, since everyone wants to get their hands on them and Apple seems to be not be up for the demand, dunno why.

For things like headphones you should definitely try to find domain experts who have put significant effort into researching and documenting the difference between products, and then purchase based on their review until you're able to make the purchase on your own based on trying the product out in person.

I knew nothing about headphones, bought a few based on expert reviews (maximizing quality in my price range), and now I feel confident when trying out "floor models" at my local audio store.

If anyone's interested in the head phone website, I can provide it. Just didn't want to seem like I was shilling a blog. If you search Google "site:reddit.com in ear monitor review blog" you'll find it, and this is a pretty good formula in my experience for researching consumer products that have a high information asymmetry.

When researching a pair I blindly sleuthed around before asking an audiophile friend of mine who referred me to http://head-fi.org which is where I ended up narrowing my selection down prior to finally examining models at stores. Build quality and noise cancelling were significant factors in my purchase.

Does Amazon accept these faulty items back and refund the customers if they ask? Do they pay the return shipping cost? Assuming yes, surely this is costing them a lot as well.

So my understanding is this: as a seller, buy subpar version of whatever item. Send it to Amazon to be comingled with every other item sent to Amazon for that SKU. Sit and wait, Amazon will use some of your stock to fulfill orders from this listing (so other sellers' sales will result in buyers receiving some of your stock). You'll also make sales yourself, and Amazon will send whatever stock from that pool of item (so not necessarily the exact ones you sent in, and most likely the genuine/better quality ones other sellers sent it). You make the sale, keep the money. Meanwhile, the buyer who got your item when buying from another seller will complain and return their item. The seller (not you!) will have to issue the refund and eat the cost. Voilà, you've exploited amazon's comingling to your profit and actually made money off of it.

Amazon accepts returns for 30 days and pays for shipping if you select "Item defective or doesn't work" for your return reason. However, that kind of con may still be very profitable for merchant since many buyers won't bother returning. 1. Pay for couple hundred 5-star reviews before Christmas 2. Get tens of thousands of orders 3. Send people cheap defective items that costs 1/10 of price to manufacture 4. Profit! Even if 80% of buyers return, it could still be profitable.

The seller can also just hit and run leaving Amazon with the losses.

Not if they only get the money after the return period is over.

Exactly - any big company will usually pay for the stock 30, 60 or even 120 days after the delivery. I can't imagine Amazon pays their suppliers immediately.

"sellers", toy was comingled and Amazon doesnt track inventory source.

I guess that's the only legislation in place right now - that is, Amazon can't be sued / held liable for e.g. broken goods or false reviews if they offer a gratuitous return policy. Like how youtube can't be held liable for copyright infringement because they offer a gratuitous takedown procedure.

They do, but sadly, the remote control trucks that I bought had incredibly cheap packaging. Once I unpacked them, the kids immediately put them to use and everything got scattered to the wind (it's Christmas morning, after all) and I couldn't find enough of the packaging to do a return. I should have, though.

"Our approximation is that less than 1 percent of reviews are inauthentic"

Get the hell outta here. Look at literally any product and you'll see a plethora of obviously fake reviews. The worst I've noticed? Check out headphones. 500+ 5 star reviews for cheap Chinese products all over the place. In fact, anything that's around 500 5 star reviews is probably mostly fake reviews.

Worse still, Amazon will delete authentic reviews that contain information they don't like. For instance, I started working my way through all of the batteries on Amazon (top review on this product, for instance: https://smile.amazon.com/dp/B009PPR5OU/ ), and Amazon deleted them from multiple products because of the table with cost/battery data (which is extremely relevant when reviewing a commodity product like Alkaline batteries). I got pissed off and quit reviewing stuff on Amazon unless it sucked (or in rare cases where a chinagarbage product and/or seller exceeded my expectations).

As a seller, I can verify that one of the reasons I can request a review is removed from my product is because of cost. However, I cannot find a reference in their product review guidelines [1]. Seems like they need to update that guide, unless I missed it.

Sellers play a big part in flagging reviews that violate the guidelines. In fact, I think most of the culling of bad reviews comes from customers and sellers complaining and very little from automation as people here might imagine.

[1] - https://www.amazon.com/gp/help/customer/display.html?nodeId=...

Additionally they delete reviews of the same product if distinct people are deemed to be somewhat connected or deemed to be in one household.

So if you and your spouse like to keep their own opinions and review separately never connect your accounts in any way.

Unless you both have different billing and shipping addreses, your accounts are already "connected" in their system.

This is actually very likely to be true. Over the entire amazon catalog, spanning some completely ridiculous number of items.

That of course, makes it a pointless statistic, and hopefully, not something, they seriously tell themselves matters.

I'm sure it sounds great for the press though!

IE you really want to know STH like "What percent of reviews are fake for the first two pages of items in the top 100/1000 categories on amazon"

or something that is weighted towards things people see/look at, not "amazon's entire catalog"

The fakes are a relatively new phenomena. A book like say 1984 has thousands of reviews over twenty years that are legit. They don’t break down fakes by year or tell you that 1% of new reviews are fake.

Also you have grey fakes with solicited reviews. I have a few vendors who offer significant discounts in exchange for reviews.

Fakes are at least three years old. At least, that's when the great I Was Given This Product For Free Or At A Discount plague started.

Plus these Chinese companies just rebrand and post new product listings free of the bad reviews, leaving the old ones to die.

Agreed -- but keep in mind that Amazon is old -- over 20 years old. There are billions of reviews.

I believe it. It's just very misleading.

Amazon obviously has a huge catalog. There's lots of niche products, where quite honestly, there just isn't enough money for fake reviews.

Not to mention some products like books, where most seem to be real.

The question Amazon isn't answering -- and what I think you're getting at -- is something like for top products in certain products like electronics -- what % are fake?

The answer to that I would wager is >50%.

There are three possible scenarios.

1. Amazon uses good ML techniques, and they don't work.

2. Amazon is doing something else to detect fraud, and it doesn't work.

3. Amazon is lying about the issue.

1. Amazon uses good ML techniques, and they don't work.

A good friend of mine worked on the recommendation engine ML side for AMZN until fairly recently. His review was that it's worse, and more expensive, than having hired people do it by hand.

But the sibling comment is more correct -- all three of these things are currently the case.

All 3 could be possible.

Couldn't it be that 1% is fake site-wide, but that for some products, 85% of 600 reviews might be fake? If I were Amazon and would want to downplay the issue, I would certainly quote the site-wide figure when the fake-review-posters are only paid to hype a couple of products.

1% seems way way too conservative, every product I click has a lot of them. Even if you count all the products people don't buy much, it has to be at least 5 or 10% fake reviews minimum.

The last time I needed headphones, I turned this on its head and bought AmazonBasics brand. Any chance of commingling there? Anyway they were great. I went back to buy more, but they were sadly discontinued.

I've only bought Sennheiser and Audio Technica headphones for the past couple of years. Good quality and good sound.

Lots of people hand out 5 star reviews too easily. They buy the cheapest headphones on Amazon, and if they make any sound at all it's five stars.

I just realized that I do that. It's just some misplaced pity from having worked in customer service where performance was tied into customer ratings.

Isn't that fair? If cheap headphones work you are getting what you paid for. No one should expect a $12 headphones to be as good as $100 headphones. A 5 star review for one means something different than 5 stars for the other.

I mean, I'm incredibly happy with my $24 Chinese bluetooth headphones.

Anecdata, sample size 1.

I bought some toy for a toddler. Upon receiving it, I realized that the description was misleading; many of the "pieces" they had listed boldly were just cheap plastic pieces of crap that had no useful functionality.

So I returned it, and wrote a 1-star review.

A few weeks later, I got an email from them: they were offering $30 if I took down the review. I ignored the email.

Several weeks later, another email: this time offering $40.

And so on.

Their latest email is offering me $70 to take down the review. I still haven't.

I'd forward the email to Amazon if (a) I knew they'd do something, and, more importantly, (b) I knew WHERE to send it!

Basically, Amazon has no interest in making sure that the reviews are not being gamed. Since the email goes through Amazon servers, they should be able to at least make an attempt at catching such people!

Hmmmph. The best I've been offered is a full refund to take down a review. I obviously didn't: a refund does not change the quality of a product.

That being said, I had a few reach out and ask me for a review. In one instance I merely indicated that I had to use the product for longer to offer a genuine review, they gave me a line to the founder for concerns. For a cellphone screen protector, that has actually worked beautifully. I honestly don't know where to draw the line between buying reviews and customer service. If I put myself in the shoes of an honest businessperson (as I'd hope to be), I hope I'd deal with customers personally at early stages of the startup. It's hard to be objective with so many miscreants in play.

You are exactly right, and for a business owner there are some other pretty awful forces at work as well. People will tend to avoid products that have zero reviews and people are more motivated to review short-term product failures than long-term successes.

So, it is in their best interest to create review incentives. Of course, some review incentives are better than others (personal notes telling you how valued a review is and how much they will help with any problems vs. "here's $5, give me some stars").

Holy crap, $70 to remove a review. I'm pretty sure one could find a law under which it's illegal to offer relatively large incentives to silence negative publications.

By the way, care to link the product so I can write a couple 1-star reviews as well? /s

>> Basically, Amazon has no interest in making sure that the reviews are not being gamed

I don't agree. First of all, these fake reviews are bad for Amazon's reputation and thus for their business. Second, these frauds lead to people returning their items to Amazon. I do not know how Amazon settles this with the original seller but I'd imagine that in the end Amazon will not make any profit on the whole transaction, or even a loss.

I still haven't found a feedback mechanism where you can say "I suspect it is fake".

I'm not buying on marketplace anymore, except if I can't find the product anywhere else.

> I don't agree. First of all, these fake reviews are bad for Amazon's reputation and thus for their business.

If that were the case, then Amazon would provide a way to flag fake reviews. But they don't.

I've had vendors on Ebay selling obviously fake products[1] refuse to refund my payments until I removed negative reviews and/or made positive full 5-star reviews. I've also received products where there was a note included, offering either free stuff or 10/20/30% off my next purchase, if I made a full 5-star review and sent them a screenshot.

When you're just one of thousands and thousands of vendors on a platform, I guess any trick is considered fair game.

[1]Such as flashlights that looked only vaguely similar to the pictures in the listings, and were blatantly just cheaply-made knockoffs.

It sounds like you didn't even try. A few days after you've contacted the seller you can escalate to Ebay customer support and they side with the buyer and issue a refund so fast I'd be terrified to sell anything on that site. You can find it written in the Ebay terms that for misrepresented items seller has to pay for a return shipping label. I only ever had one seller refuse to and Ebay c/s said keep the item and we'll issue a refund.

Oh I did get my money back and left negative reviews. Not sure how you got the idea that I didn't. Ebay almost always rules in the buyer's favor.

Have you ever tried reviewing a seller instead of the product? Amazon reviews almost every single negative review.

> Basically, Amazon has no interest in making sure that the reviews are not being gamed. Since the email goes through Amazon servers, they should be able to at least make an attempt at catching such people!

Short-term, no. Long-term, it harms their credibility until they're past being the status quo because people go to alternatives which aren't as popular or infested with this crap. In short, this is good for competition.

But the real problem, as I already wrote in another post, is the fact that Amazon isn't held accountable for this. And if that's currently not possible, it should become by law. After all, they're hosting deceiving misinformation and facilitating a platform for sales. Its easy to get away being the middle man but that doesn't mean its fair.

You can report a seller and it was pretty easy to find via google, "amazon report seller": https://sellercentral.amazon.com/gp/help/help.html?itemID=20...

It's common knowledge among most techies that the majority of Amazon reviews are fake and that counterfeit items are a big issue, but most of the general population is completely unaware. I was searching for an inexpensive pair of polarized sunglasses last week and the "Amazon Recommended" ones had an F on FakeSpot with 70%+ fake reviews.

You really should never be shopping on Amazon for high margin products that are easy to fake. This means perfume, cosmetics, sunglasses and name brand clothing. These items are cheap to manufacture (even the real ones), but get their high perceived value from brand. Those Raybans you love? A $250 pair costs Luxotica around $5 to manufacture. So items like these are common to fake because the fakers can often get a product that is close in quality to the real thing. If they can just trick people into thinking it IS the real thing, they can make as much margin as the big guys without the cost of building an actual brand.

BTW, because of the way Amazon’s FBA service works, even ordering items that are “sold by” a reputable retailer but “fulfilled by Amazon” gets you a high chance of receiving a fake. This is because they mix FBA stock and don’t discriminate what stock it comes from when filling that item, regardless of who the seller is.

Fakes I’m not too worried about, I can usually spot those. The problem is that many categories are flooded with garbage products with inflated reviews. Try to find a bathroom scale. Tons of choices with 4.5 stars and above, but most of the reviews are fake, so it would take hours to sort through the choices and find a decent product.

Same for most things I’ve wanted to buy recently (hiking poles and a daypack are two other examples from this past weekend).

The deluge of low quality, rebranded, foreign built junk is a real problem for Amazon. A search for food scale[0] for example gives you a whole page of similar (many exactly identical with different brand names) pieces of garbage - they have massive, consistent weight drift and drain batteries like it's their job, and yet they're rocking thousands of 5 star reviews.


This exact issue caused me to use Amazon a lot less.

these days, I check Consumer Reports for a review and usually buy their recommendations. Most of the time, still on Amazon - their delivery is still top notch.

But - not all the time, as there are now viable alternatives that do 2-day delivery well enough. For example, I started ordering some stuff on Walmart.com. A lot of things that are “prime pantry only” on Amazon are available for the same price on Walmart.com, without any restrictions.

> But - not all the time, as there are now viable alternatives that do 2-day delivery well enough. For example, I started ordering some stuff on Walmart.com. A lot of things that are “prime pantry only” on Amazon are available for the same price on Walmart.com, without any restrictions.

There's also local retail. If you don't mind a trip to a store, you can size up the product in person and go home with it the same day. I'm my experience, the prices are often competitive with Amazon, too.

Consumer Reports, Sweethome, Wirecutter. Usually ignore customer reviews and go with Amazon Prime picks for the rest.

Just wait untill these Chinese firms master naming their brands with English sounding names.

That and the thousands of "very good product!" fake reviews, you will not be able to find a non-plasticy, built-to-last product on Amazon at all.

And the fun will continue when you discover the warranty requires you shipping items to PRC, if it's honored at all.

> drain batteries

I'll stick with my ancient one that uses a spring.

> it would take hours to sort through the choices and find a decent [bathroom scale]

Maybe it's different in the USA, but when I needed a bathroom scale for our new apartment I went over to the store and bought the first one they had. If I would have come home and it wouldn't have worked, I'd have gone back and had either gotten repair under warranty or cash back (if they don't want to repair it, you can rescind the purchase). Took maybe 20 minutes. Ordering online would be quicker in terms of buying, but I would have to stay home to receive the package, and if something is wrong, I have to ship it back and get the money back for shipping, plus wait for them shipping a new one... Wouldn't buying from a physical store be easier than looking through hundreds of reviews, both in terms of time as well as in terms of mental effort?

I understand it's just an example and this doesn't apply to everything, but I do hear that in the USA, people order everything from Amazon instead of walking over to the store. I just can't imagine staying home for packages all the time, let alone trying to spot fake reviews for simple things like scales (of any kind), microwaves, etc. They are simple things with simple features (e.g. microwave: power, size, and whether it revolves) that I don't need reviews for. If it doesn't work as expected, also if it breaks after three years of light and correct usage, I'd get warranty (assuming the expectation that a microwave lasts a decade is not wildly off).

> I just can't imagine staying home for packages all the time

I'm not sure why you think people in the USA do that. You certainly don't have to stay home to get package. Nobody does that.

If it's a valuable package, there are ways to make sure it's not stolen, including using an Amazon Locker.

In Germany DHL has lockers that you can use as an address. The next one is closer to me than any store.

Although I agree I have a certain reluctance to return stuff, because now I need to wrap the package, get it to the post office (which is far away) and pay shipping.

In the US, you can just put the package in an Amazon Locker with the return label printed on it. You don't have to pay for shipping.

I don't know if Amazon Lockers exist in Germany, but I'm sure they would reimburse your for the shipping cost if you ask.

It's called living in the suburbs. I don't order everything on Amazon, but if I need some basic item (hangers, laundry detergent, vent cover, bathroom scale etc) it's a huge time saver! It would easily take me 45 minutes round-trip to drive to the nearest, say hardware store, locate what I need (assuming it's in stock) and drive home. On Amazon, I find the item, swipe right and wait two days. No need to stay at home waiting for the package, I live in a safe area.

Honestly I've had zero issues with counterfeits, but then again I don't have any kids yet, so I'm not buying many toys. I also typically pay more for name brands due to quality concerns in general, so that probably insulates me to a degree. I'm not denying people's experiences, but honestly Amazon's reputation is still fine to me. They're a marketplace like any other and they never really marketed themselves as premium. Curation and trust was never part of the deal IMO. I don't trust Walmart or Best Buy either.

I live in NYC and I'm definitely not taking the subway three stops just to get to a place to buy a bathroom scale, I would just order that from Amazon. Getting to certain varieties of stores in urban locales is pretty annoying and time consuming just like in the suburbs.

In my old neighborhood there was one place to buy bathroom scales , however, they were not sold with warranties and were basically getting what you receive from Amazon at a higher price.

Yeah, and the cheap bathroom scale would have been just as awful as the middling ones on Amazon. And that would have probably been just fine.

One of the weird things about Amazon is that its turned us into people that think we need a crowd-sourced review for every $20 common household item.

  I do hear that in the USA, people order everything from
  Amazon instead of walking over to the store. I just can't
  imagine staying home for packages all the time
It depends on your life situation.

Personally, it's hard for me to get to stores on work days - whereas my workplace will accept my personal parcels.

On the other hand, I know people whose workplaces won't accept parcels for them, but who walk past loads of stores on their way to work.

So for me, ordering online is a convenience - but for them, it's an inconvenience.

> Maybe it's different in the USA ... I went over to the store

Disclaimer: I don't live in the USA

Maybe it has something to do with the fact that many cities in the US have residential neighborhoods separated and far from commercial ones? If I had to grab the car and make a 20min drive, and find parking, just to buy a 10$ scale, no thanks, better to order online

They need to give customers a way to sort by review counts in a time-weighted fashion. In other words, a product with 1,000 reviews and a 4.5 average that had been sold for five years would rank higher than a product with 1,500 reviews and a 4.8 average that had been sold for three months. This would help identify the quality products from stable manufacturers.

> They need to give customers a way to sort by review counts in a time-weighted fashion. In other words, a product with 1,000 reviews and a 4.5 average that had been sold for five years would rank higher than a product with 1,500 reviews and a 4.8 average that had been sold for three months. This would help identify the quality products from stable manufacturers.

Major improvements have been needed in Amazon's search ranking for years. The fact that it's little better than it was in 2002 tells me they don't really care to fix any of the issues with their shopping experience.

Dear anyone at Amazon: please read this and implement the suggestions: http://www.evanmiller.org/how-not-to-sort-by-average-rating....

Any ranking system you can suggest can and will be gamed and exploited. The problem is that Amazon is trying to rank products at all, not whatever opaque algorithm they are using to do the ranking. They need to turn off all the ranking and rating, and start actually vetting their suppliers, but that’s hard and expensive so they will never do it.

> They need to...start actually vetting their suppliers, but that’s hard and expensive so they will never do it.

I agree they need to start actually vetting their suppliers. But I disagree with you about the ranking. "Sort by average ranking" is broken obviously brain dead ways. They need a transparent ranking system that correctly balances number of reviews with average rating. I don't want to user their actually opaque "relevance" ranking.

I find that it's usually pretty clear from the bizarre language, excess enthusiasm, and lack of detail which reviews are fake or paid for, without need to refer to any 3rd-party websites. For many products with fewer than 100 ratings, it's clear that all or almost all of the ratings are fake. This often happens when there are several versions of essentially the same product, presumably all from the same factory in China. It's usually best to stick to the one or two variants with the most reviews.

Bear in mind that the existence of obviously fake reviews doesn't imply that all fake reviews are obvious. Maybe it would be prohibitively expensive to do a really good job of producing fake reviews at scale, I don't know -- but surely some companies manage to avoid the obvious tells.

Recently it became clear to me that my Amazon shopping has been limited to stuff I don't really care about that much.

These are product categories (and european sources) I frequently buy from, yet Amazon hardly serves any of them competently:

- studio equipment (Thomann)

- audio equipment (countless niche, yet competent HIFI shops)

- computer hardware / technology (Really, where to start? Alternate, Mindfactory, Apple)

- bike components (Rose, Mantel, ...)

- tea (Teekontor Kiel)

- clothing (Rapha, Uniqlo, Zara)

For most of these categories Amazon is completely useless. Discovery is abysmal. Selection is mediocre. Product data / listings are a mess and show repeated results with slightly different, often incomplete variations from multiple sellers. Alternatively you get one single product page that features countless, sometimes unrelated product variations that may or may not be Prime-eligible. Maybe Amazon is fine with me buying cheap stuff from them and I still value their Prime-powered, reliable next-day delivery, but for important, expensive stuff I expect some degree of curation and will use other stores. At this point Amazon is slightly better than eBay.

For something like hiking poles, I really recommend REI. It may cost a bit more but you get higher quality materials and great customer support.

Faced with hundreds of junk offerings, sometimes a high-to-low price sort gets the quality product onto the first few pages. Doesn't work if there are many "packs of 50" though.

>BTW, because of the way Amazon’s FBA service works, even ordering items that are “sold by” a reputable retailer but “fulfilled by Amazon” gets you a high chance of receiving a fake. This is because they mix FBA stock and don’t discriminate what stock it comes from when filling that item, regardless of who the seller is.

Is this also true with items sold directly by Amazon?

Yes. They commingle all the inventory.

Not true. Sellers can prevent comingling inventory by creating their listing with a unique FNSKU. This doesn't prevent the product from being sold on the same ASIN. If sellers can do that then Amazon can absolutely do that.

They could, but they don't. Why is a bit of a mystery because in the long run this is destructive and it can't be that profitable.

If you sell anything on Amazon and get items back from FBA before they hike up prices near the holidays you will see items returned which were not what you shipped...books for example.

How legitimate is FakeSpot though? I find nearly everything has an F with a high amount of fake reviews when I use it.

I checked your comment and it was an F with 83% fake on FakeSpot.

Edit: I was paid for writing this review.

I checked FakeSpot with FakeSpotFakeSpot, a website to check the authenticity of fake-review spotting websites, and it received an A with 99% authenticity.

I would guess that it is legitimate. The fake review problem is real. Here is an example of an item with 2k good reviews:


Notice that it is not a cheap plastic good, and is neither easily duplicated nor ripped off.

Is that a sign of illegitimate fake-spotting, or a sign of just how widespread fake reviews are? Personally I'm quite inclined to believe an extremely high fail rate.

> Is that a sign of illegitimate fake-spotting, or a sign of just how widespread fake reviews are? Personally I'm quite inclined to believe an extremely high fail rate.

I think fake reviews are pretty widespread. This Washington Post story (https://www.washingtonpost.com/business/economy/how-merchant...) talks about how Facebook groups are used trade free stuff in exchange for good reviews. I joined some, and they seemed to have a lot of traffic. FakeSpot seems to put a lot of emphasis on analyzing the reviewers, probably looking for accounts who appear to be participating in one of these review rings.

> FakeSpot seems to spend put a lot of emphasis on analyzing the reviewers, probably looking for accounts who appear to be participating in one of these review rings.

Presumably this (and the fact that Amazon can analyse them too) leads to a wider spread of fake reviews even on legitimate items, as having fake accounts write generic reviews on a random selection of items they don’t care about would make them harder to detect.

Presumably blocking spam leads to a wider spread of spam. Presumably having a zero tolerance policy on drugs leads to a wider spread of drugs. Presumably making firearms legal leads to more violence. Its an arms race, but just because of that doesn't mean the fight is worth it or not.

I'd say however that the lack of accountability by Amazon (et al) is the real problem. They should be held accountable for these fake reviews. We need a lawsuit on this.

It's been reliable for me.

> I find nearly everything has an F with a high amount of fake reviews when I use it

Seems like a reflection of how bad the reviews on Amazon are, not a reflection of the quality of fakespot.

Some categories are simply flooded with fake reviews where you won't find a single honest listing in the first 3 pages.

That said, there are plenty of A rated products in my experience. All of my products are A-rated, and unfortunately many of my competitors are honest too.

It has worked well for me. There are plenty of A and B products, keep looking.

Its hit or miss in my opinion from when I used it.

I personally just do my own homework. Rule of thumb is 2 or 3 star ratings IMO have higher weighted value because people weren't biased enough to leave a 1 or 5 star review.

Always check the most critical review as well, also all the 1 star reviews. If there's a video, its the most authoritative and unbiased source usually. Images are definitely good too.

Do a search on the question/answer section for competing products. The ones that are most unbiased are going to offer viewpoints in two competing products usually, but this sometimes can be faked. Check the reviewers history if there are both negative and positive reviews, as well as the semantics in their wording.

Check also if there's an alternative source of reviews, e.g. best buy & newegg for electronics. Do the ratings more or less indicate the same behavior? Or is there a discrepancy by a large amount?

If its a popular item or something like gym equipment, do multiple youtube searches and find both (1) authoritative well known youtubers and (2) amateur youtubers. (2) are usually unbiased (1) are potentially biased

If its a really popular item (phone, laptop) use other review sites as metrics. Wirecutter/Thesweethome started off as fairly unbiased, but were acquired by a large company so the ratings are now slightly skewed. You can always check other forums like linustechtips, reddit, and do a search across that entire thread for more answers.

Don't bother doing this much research unless its a big picture though. I had to buy a macbook pro so I thoroughly did 5-10 hours of research in which version I wanted (2015), which specs, potential issues I would run across, considering where the 3rd party sources of used macbook pros would most likely be sold at (ebay), etc. Differences in CPU/GPU benchmarks, battery life cycle usages, etc.

Seems to only work with Amazon.com not Amazon.co.uk or Amazon.de. Do those not suffer from fake reviews? Also, I found that buying the full review from "trustworthy" cost 10 USD.

On what do you base your opinion that the "general population is completely unaware?"

I believe that's just common sense.

You'd think that Amazon would have the machine learning resources to detect likely-fraudulent items with high accuracy, should they be interested in preventing such obvious fraud.

They have the resources. What they lack is the proper incentive. As long as fake reviews don't completely destroy the customer experience to the point that people forego any other advantages (mostly convenience) of shopping there, Amazon will continue to reap the short-term benefit of more apparent variety, more apparent quality, and more sales. Any effort beyond keeping the fakery just short of destroying the brand isn't worth it to them.

I'd venture further and claim that it's in Amazon's long-term interest to drive down the perceived brand value of third-party suppliers through its distribution channel.

Costs are lowered and the value of Amazon's brand is relatively increased. Similar to stocking own-brand product in a supermarket. This weakens the negotiating position of big name brands.

Similarly for Kindle Unlimited despite the fact that Amazon clearly has the technology to address Markov Chain generated fakes, plagiarism and upvoting and review cabals, Amazon either chooses not to invest or not to invest enough to prevent all of that bad behavior.

IMHO investing 1/2 of 1% of the KU pool is enough to fund human review via their Mechanical Turk program. It wouldn't be perfect but it would eliminate most of the fraud.

Would this be the same machine learning that was recently in the news for incorrectly identifying 28 lawmakers as people that had been charged with some crime? Maybe it is, and that's why it can't spot the fake reviews?

The 28 were incorrectly indetified, because they ran their tests using default settings, which aren't fit for such an use case.

> You'd think that Amazon would have the machine learning resources to detect likely-fraudulent items with high accuracy, should they be interested in preventing such obvious fraud.

They must not be interested. After all, they still get paid and have a level of legal insulation due to how they've structure Amazon marketplace. I wonder how long it'll take for their overall brand to take a hit over this, though.

There's probably a huge number of shoppers that use Amazon for nearly everything, myself included, that would just toss the cheap crappy thing they got off Amazon and try another brand from, you guessed it, Amazon.

I don't know how many horrible sets of bath rugs and towels I went through on Amazon before I finally bit the bullet and bought an actual expensive brand and the difference is staggering. Before someone asks they're from Frontgate (not on Amazon).

Point being I'm just sending Amazon money over and over again.. It's definitely made me go back to stores where I can see the product for certain things.

Why is this “common knowledge” while almost nobody believes astroturfing exists on Reddit or Hacker News? Serious question. I don’t see a fundamental difference.

I get fake brushheads this week for my Phillips toothpaste. Also Amazon recommend.

So true.. I like the Flying Fisherman San Jose sunglasses and have gotten them from Amazon under <$15 in the past. Not quite as cheap and as disposable at $30 IMO.


There will always be a desire to pay for reviews, since 70%-80% of shoppers check reviews before buying[0][1] (and growing[2]).

Amazon could help fix this problem by creating "sponsored" reviews. This way a company can pay for reviews in a non-scummy way, a reviewer (and Amazon) gets paid, and a customer can tell whether the review was paid for or not.

Or just pass laws to regulate reviews.

0. http://www.peopleclaim.com/blog/index.php/the-review-of-rati...

1. https://www.chainstoreage.com/news/study-81-research-online-...

2. https://deloitte.wsj.com/cmo/2016/11/01/media-device-habits-...

Amazon has done this for years now. It's call Amazon Vine. Select Amazon customers with high review ratings are given free* products for which they have to give unbiased reviews. It's both sponsored by Amazon as well as various manufacturers.


It has mixed results. Amazon is upfront about Vine reviewers getting free product and resulting reviews have a Vine tag on top. Customers tend to not like Vine reviews vs verified purchase reviews. While there are negative Vine reviews, most are positive. It may be due to a psychological quid pro quo, since you tend to get more products from the same manufacturer if you offer positive reviews vs negative ones. It's just really hard to be unbiased when you keep getting "free" stuff from one of your favorite companies.

*It's taxable according to Amazon's cost in most cases so you get to figure out Amazon's margins for a lot of products. If it's manufacturer sponsored, then the item isn't taxable as income. Food and similar products are also not considered taxable.

Why would companies know who's getting their products to review and why would reviewers know which products they're getting? I feel like an easy fix for this is to have the company send the product to Amazon first and then have Amazon handle the distribution to reviewers. There's no "quid pro quo" if there's no incentive to review a product favorably.

That's why Consumer Reports was so successful in the past. Their reviews were completely unbiased because the draw was in people paying to get thorough, objective reviews of products.

Now that revenue comes from ads and manufacturers all over the internet, it's hard to find reviews that aren't paid ads.

While product-for-reviews have a certain conflict of interest, I know quite a few people (including myself) who write reviews of both products we receive for free and products we pay for who give honest opinions. If I agree to let you send me a product, I'll write a review but you won't like my review if I don't like your product.

When I first joined Vine a long time ago I think you had to have passed a certain threshold for number and quality (based upon helpful votes) before you were offered a place. That definitely changed over the last couple of years.

Once it opened up to people that had never left reviews, naturally the quality is going to drop by some amount.

> Why would companies know who's getting their products to review and why would reviewers know which products they're getting?

Amazon Vine reviewer here, for a handful of years, probably longer. They've changed how things are handled, so I'll say how it is now.

When I remember (they did away with an email of items you could select from years ago), I log into the Amazon Vine portal and see about ten things I can select to review. It seems to be somewhat based upon what I'm shopping for on Amazon, as more than once I've seen something in Vine a few days, weeks, or months after I've been looking for the item online (and sometimes have purchased something similar).

Anyway, I select the items I want to review and then get them within a week. We used to have 30 days (or four week, I forget which) to use and review the item, but now there's no date, at least not one that we can see. To review the item we go back into the portal, leave our review using essentially the same interface you use whenever you leave a review, and then move on.

When it displays on the site our reviews automatically get a badge that it's a Vine review. If the product ends up getting too many Vine reviews (or some other criteria) our reviews could be removed without notice.

So companies don't necessarily have a hand in deciding who gets their product, but I'm sure if they cared they could see who did.

A few times I got something just because, back when we didn't have to pay tax on the items we received, since Amazon could technically ask for them back at any time, as they still owned them, and my reviews were generally lower than my usual reviews because I didn't want the item to begin with.

Now that we need to pay tax I'm much more selective on what I want, and probably give more 4- to 5-star reviews because of it. If I just got something and had to review it, then it would be more likely I'd get something I didn't want (women's wear in a size my other half doesn't wear, or kid toys for a different gender or age range), and it's score would suffer.

From Amazon's perspective, I'm sure they rely upon the fact that reviews tend to be either overly positive or overly negative (when was the last time you left a review on something that was okay, or sufficient?) so that they can move product. Perhaps Vine helped with that. Perhaps their new program (the name of which I've forgotten) does too.

Personally, I can sleep at night because I'm okay with the scores I've given the items I've reviewed. Five-point scales are pretty easy; 1 = hate it, 2 = don't like it, 3 = no feelings either way, 4 = like it, 5 = love it. I've also gone back and changed scores because of what I've experienced later.

Back when they had forums there were clearly some people that just wanted the free products, and would then turn around and sell the item after a few weeks or months. I know I definitely wasn't one of those people.

> since you tend to get more products from the same manufacturer if you offer positive reviews vs negative ones.

Is this true? I'd expect if Amazon had any integrity they'd take steps to prevent this via there position as an intermediary.

I doubt that sponsored reviews would solve it - if they wanted to be shady they would astroturf it anyway. Marketing departments with multimillion dollar budgets still engage in that. As for regulating reviews what could possibly work for that? Real ID would be unduly oppressive in chilling effects in addition to opening up barratry and harassment of reviewers. Even if fake reviews fall under false advertising laws good luck enforcing it.

What about independent reviewers like consumer reports, or similarly disinterested party?

> and a customer can tell whether the review was paid for or not.

Isn't this what the companies are paying for?

Given the opportunity to pay for sponsored reviews or risk getting perma banned from Amazon, I'm guessing a lot of companies would choose the former. Regardless, it's not an idea I have seen tested yet anywhere, so it's hard to know.

I don't just check the average rating and number of ratings. I also check the quality of reviews. If an unscrupulous company wanted to increase the chances of me buying from them, they might try buying bad fake reviews for their competitors.

Yes. A review that explains why the product is good is harder to fake. Plus the reader may be able to verify independently whether the explanation itself is good. In which case the source no longer signifies.

Yes. I find that the positive reviews are often more useful than the negative reviews. If a positive review is five star but still mentions some negative things, it feels authentic and I can go off of that. This is how I deal with the problem where people won't give anyone less than five stars because they don't want to hurt their livelihoods.

The fundamental problem though is that companies want to discretely inject these overly positive reviews. If users can identify these “paid-for” reviews, they will most likely treat them with distrust.

I really can’t see a way around this.

> Or just pass laws to regulate reviews

Australia just today fined a company for manipulating TripAdvisor reviews:


And I've had issues trying to leave less than perfect reviews - seems like only after I spend time typing up the review (and sometimes uploading photos) does Amazon come back and say, sorry can't publish this.

Seems like a dark pattern to encourage only positive reviews.


YES! I have this happen to and notice the same patterns. I honestly think there is grounds for a lawsuit here. Whether or not that would be a successful lawsuit is another question, but the threat of one could get Amazon to change its tune.

“Reviews”, in general, are absurdly misleading on any site; Amazon is just prominent.

I’ve given real reviews on Yelp of businesses that had real problems. These negative ratings were shuffled off into their special bucket of maybe-less-reliable reviews, well under pages of HIGHLY questionable 5-star reviews. I also have a friend with a business who was plainly told that he needs to pay to ensure good reviews. Yelp is a scam, and wins by being a virtual monopoly (as Amazon reviews do).

It is also amazingly hard to get reviews on the App Store from real users, for instance. I imagine you pretty much have to ask 60 friends to “help” you, or pay somebody, either of which creates a pool of fake reviews. It really changes your perspective on any other “ratings” you see.

In the end, I think society will learn that this is another example of buyer beware, and you get out of it what you put into it. Since looking over “reviews” for 30 seconds is low-effort, it is low-value. If you want good results, you need to spend a good amount of time and effort researching products.

This topic keeps coming up in the news, and I've altered my shopping habits because of it. The competition has gotten tougher with Google Express, Walmart delivery and pickup, Target pickup, etc. When I order from one of them or make a trip to the store, I can at least verify the item I'm purchasing is from the manufacturer.

When looking at the reviews on Amazon, countless times I've seen reviewers mention to not buy the item from a particular seller because it was not genuine or was defective. If I'm not sure about the brand, I'll search it to see if its legitimate and if there are reviews/complaints on other sites. I'll also use buying guides when determining what to buy instead of Amazon reviews.

This is a huge issue, several times I've received fraudulent goods when shopping at Amazon. It has made me completely avoid shopping for certain items there and makes me wary of trusting anything on the site where quality/reliability is an issue.

Avoiding particular sellers is no guarantee, since Amazon comingles inventory in the warehouse. You can order from seller A, who is completely legitimate, and receive fraudulent goods seller B supplied.

Makes me wonder how many of those guides have been paid for, or at least what level of influence advertisers have on them. Nevertheless, I agree that a quality guide is well appreciated, however hard to come by.

Many "guides" just end up being affiliate marketing link aggregators. I appreciate The Wirecutter's About[1] page where they tell how they review products.


I'm not really sure why there are any real reviews on sites like Amazon, Yelp etc. Why do people leave reviews; what's in it for them? Is it some strange form of altruism where they're helping their fellow anonymous shopper while also helping out one of the world's most successful megacorporations?

Writing this comment has made me re-think why I'm writing this comment.

I tend to leave a review when something is better than expected or if I'm disappointed enough to do so. I'll also occasionally leave a review of something if I'm prompted, but that more depends on whim. I also am likely to leave a review on products that are occasional purchases - small kitchen appliances. This is because I personally find these helpful when sourced from multiple places.

I rarely leave a review for average products. There is rarely a good way to communicate to folks that the product simply met expectations. For example, if I buy a bag of coffee that I've purchased before and I get said coffee, there isn't much I can say. "I bought this because I find this coffee good enough to buy a second time. It isn't my favorite, but it is good and my average quality of coffee. Package was undamaged and contained the expected coffee".

That isn't exactly a 5 star review, yet if I give 3 or 4 it seems to be a ding on the company. I'd rather not, thank you.

> I tend to leave a review when something is better than expected or if I'm disappointed enough to do so.

And that's the problem with user reviews. They emphasize the extreme ends of the spectrum. There's little incentive to leave a review for the potentially silent majority of moderate users that are simply satisfied with the product.

It is, and it is something that could be solved simply by asking the right questions instead of the overall review on a number (or star) scale. I hardly think simple satisfaction warrants a 5-star review, but I think if I just put 3-star, it might hurt the company... or waitress, or whatever. It isn't sane, and I'd almost argue that there is a disincentive to leave a moderate review. It is the same for store satisfaction surveys and many more - I know folks get penalized for not having overly good results.

But if they simply asked if it was what I was expecting and had reasonable options - along with not punishing folks for simply delivering what I expect - I'd answer yes.

> Why do people leave reviews; what's in it for them?

Me? Because I enjoy helping others - and it is my best way of punishing bad actors and supporting good ones, hopefully helping to balance the market in the long run.

Actually a bit of the same reason as why I'm writing here: help people by providing facts, add some balance and maybe sometimes a new perspective.

> Is it some strange form of altruism

That makes me sad. Altruism is not strange. The urge to do for others what you would have liked them to do for you is common, normal, and healthy. I review only a small percentage of the things I buy online, but I do review some when I feel that I can provide some information I don't see in other reviews. Precise details about fit for running gear, or durability for just about any product, are often hard to come by, so I'll often come back and add a review to share my findings. Maybe it will encourage someone else to do similar reviews, and maybe some day I'll benefit from reading one of those. Does that seem strange to you?

You have to look at the alternatives :

- you can post a review

- you can talk about your disfunctional item to the next person in the room

- you can stay silent

Staying silent is often a mental burden, because you invested time and energy to buy something that doesn't work, and you feel betrayed. You can keep silent only a finite quantity of such "betrayals", before it adds up and starts affecting your mood and social interactions.

Talking about it to the next person in the room is a quick way of getting rid of that mental burden, at the cost of your social interaction with that person. Most people don't want to listen to other's problems/failures, especially if it regards an item they will never use.

Or you can leave a review. That way, you can get rid of the mental burden, and the review targets the most pertinent people : the seller and the future buyers (which include yourself maybe). At best, the business will iterate on a new product, taking into account some of the reviews.

Edit : just writing that comment to convince myself here ;-)

I write reviews usually only in extreme cases if something is either exceptionally good or bad. In the positive case, I want to help the creator/owner having the deserved success, in the negative prevent fellow customers getting ripped off.

The problem is, my review might appear fake.

I'm not sure people know what reviews are anymore. I often see people giving 5 stars to a product they have not received yet, or a game they barely played.

Why do people contribute to open source?

Some of that applies here as well: the feeling of doing good, the hope of "paying forward" works (just the culture of leaving reviews might one day lead to a review for me) etc

I've been using https://www.fakespot.com to flag bullshit reviews for a few months now. After many happy purchases informed by its review analysis, I believe it works.

How can we believe you?

: )

Well don't take it on trust, validate it works. And keep validating it still works, because one day it won't. Everyone seems to sell out in the end.

It's impossible to validate. Further for all we know Fakespot (which curiously has been mentioned multiple times) might itself be a fake that undermines products that are sponsored, etc. It probably isn't, I'm just saying.

? You could manually analyze the reviewers for suspicious non-specificity, typical of paid reviewers for whom it is not cost effective to write a tailored review when they are being paid pennies per post. Also over time you could compare the quality of products that it says have fake reviews vs ones it says does not.

But how can I trust what you're saying? What if you've been paid off by Fakespot?

That's why I'm launching my new website, SpotFakeFakeSpot, which helps spot fake fake review spotting sites. I got the idea from my time as a sales rep selling radar detector detectors to police departments.


I had no idea there's a full scale electronic war on the streets of my city.

You have no idea, man. Radar detectors these days even have radar detector detector detectors built in.

What do you mean by "validate it"?

I've used fakespot for purchases for a while now. I feel that it's helped me avoid bad products. Thus... it's validated (at lease for myself)

Would be amazing if fakespot runs the whole fake review business and therefore can tell you which ones are fake. Wait until this becomes a paid service.

Probably not, though, so more likely the fake reviewers will find a way to avoid its detection algorithm at some point.

How do you determine a review is bullshit? Do you only flag reviews for items you've bought?

5 stars.

Very generic positive language.

Poor grammar and or spelling.


The only sure way is to click on the profile and read the rest of their reviews.

Hmm, I had (apparently wrongly) assumed this had become common knowledge.

Now I am beginning to wonder what percentage of people are aware of this issue.

On a different note, I saw elsewhere (Axiom perhaps? I cannot remember) that amazon claimed it was only roughly 1% of all reviews that were paid for. Which based on my past experience in E-Commerce, seems significantly lower than would be estimates.

> Hmm, I had (apparently wrongly) assumed this had become common knowledge.

me too. maybe NPR is behind the times. also: Amazon is a sponsor of NPR (just sayin')

> amazon claimed it was only roughly 1% of all reviews that were paid for.

yes. that's right. the 1% figure also appears in the NPR article:

"Our approximation is that less than 1 percent of reviews are inauthentic," says Sharon Chiarella, vice president of community shopping at Amazon.

Even if we make the (dubious, imo) assumption that the low 1% figure is accurate, it can still be a huge problem.

There are 480 million products on Amazon, many in somewhat obscure categories. Seems that if the fake reviews are simply concentrated in a popular 1% slice, we've got almost 5 million popular products with ~~100% fake reviews.

Obviously oversimplified, but the likelihood is that the reviews are sufficiently concentrated to be a problem using their own figures, and the real figures are likely higher

I would not assume that. The impact of a paid review decreases as the number of reviews increases. ie it probably better to pay for reviews for a low volume or new product

Last night I finally abandoned a book that was absolute trash, but had 83% 5-star reviews on Amazon. Looking at the reviews now, I see that most of them are written by people who only write 5-star reviews (it has an F-rating on Fakespot). Feeling like kind of a dummy right now.

Name and shame!

Not the OP, but this reminds me of a similar experience I had with a book called Space Opera a few months ago. Don't remember where I first heard of it, but it was being touted as comparable to the Hitchhiker's Guide. All the reviews on Amazon were 5 stars, and since the kindle version was only $5, I bought it without checking further. It turned out to be a pretty disappointing read, and nowhere near the caliber of anything in the Hitchhiker's series. I remember while reading it wondering how could all the reviews have been so good. I just checked fakespot, and it had an F rating with 92% low quality reviews as of April 25 [0]. I then had fakespot update the analysis, and it's now a D, with 41% low quality reviews [1]. So it seems like the fake reviews were bought to get things jump started, and now real reviews are catching up. Now that I think about it, it's probably a fairly common way for authors and/or their publishers/publicists to get the early sales and hype going.

[0] https://imgur.com/a/PWFTTbV

[1] https://www.fakespot.com/product/space-opera-7ea7cc4e-4bae-4...

There are groups where ebook authors get together to essentially cross-review books of other members.

I remember reading an exposé recently, which I suspect was posted here, but I can no longer find it. (An other dropped out of the group(s) due to ethical concerns.)

Self published ebooks also fall under the category where a high proportion of the reviews are likely to be useless for evaluating the quality of the book even when they're noncommercial and sincere, because the initial reviews all come from close friends and family who are impressed by how creative Jonny is or hardcore sub-subgenre enthusiasts, and unless the book sells really well or is utterly atrocious rather than merely a bit mediocre there will be comparatively few reviewers interested in contradicting them.

Do You Realize? https://www.amazon.com/gp/product/B06VX1QQHX/ref=oh_aui_d_de...

It was only 99 cents, but I wasted at least a couple of hours on it, which is what really annoys me.

The Art of the Deal has an F, with 57% low quality reviews.

"Our engine has profiled the reviewer patterns and has determined that there is high deception involved."

Hmmm...can't edit, but: I just looked at this again, and it's interesting to note that the Kindle version has an A rating and the print version has an F rating, even though they appear to share the same set of reviews.

Scroll down to see the first two entries in the product list here: https://www.fakespot.com/company/donald-j-trump

If you are solely posting just to make a political jab then there are more appropriate places to make such comments that are not Hacker News

Shouldn't it be around 49%?

Or is it just bumped up due to an old electoral algorithm Amazon uses?


One of the many reasons I let my Amazon Prime subscription expire last month.

Amazon is selling convenience. But it's not convenient to:

1. Filter out products full of Vine program reviews, which I believe are completely skewed due to the reviewers receiving the products for free

2. Read between the lines to filter out paid/bias reviews that are NOT part of the Vine program

3. Research to ensure the product itself is not fake or counterfeit

Feel like I'm shopping at the shadiest flea market in town.

I usually just ignore 5-star and 1-star reviews. Many of five stars are either fake or people who just bought the product and are too excited to leave an unbiased review. Many of one stars are from people who are not satisfied with the purchase process (slow delivery / fake reviews / broken item), so they leave bad review that doesn't really reflect product quality itself.

I rarely leave reviews, but there has been a case where I had to threaten to sue a company before they'd hold up their end of an agreement. And 'threatening' is not just mentioning "I'll sue you!" but actually researching applicable law, what prerequisites I needed to be able to sue (e.g. I needed to hire a debt collector first, which would have cost me money as well), applicable court, what the bailiff would cost, court fees, etc. Some fourty phone calls and a few emails and letters later, spread out over half a year, I finally got the money back on the day before I would have had grounds to sue them. That was more than half a year of stress and them having both the product and my money. To be honest it was kind of disappointing that they paid up in the end, as I was finally well-prepared and wanted them to have to pay the fees that debt collection etc. would have incurred.

You bet that after it was all over and done, I wrote that one-star review. While writing it, I noticed that other people had the same issue. One person also had to cite court fees before they paid up, and some others didn't know enough about applicable law or had amounts that were just too low to go to court over (like $50-$250). I usually don't put much stock in such reviews, but it turns out I should have read and considered them.

There are definitely legitimate cases for one-star reviews.

According to outside auditors like Fakespot and ReviewMeta, more than half the reviews for certain popular products are questionable. Amazon disputes those estimates.

"Our approximation is that less than 1 percent of reviews are inauthentic," says Sharon Chiarella, vice president of community shopping at Amazon. She adds that "sometimes individual products have more suspicious activity."

The disparity in what outside services observe and users experience vs. what Amazon claims, and the increasing press coverage and online discussion of the problem, will force Amazon to take increasingly drastic measures to crack down on bogus reviews.

There is also the issue of increasing levels of returns for 5-star crap. But, while Amazon can absorb the costs of returns, Bezos truly hates negative press, and the Amazon PR machine will go to great lengths to get high-profile news orgs repeating the official Amazon line.

If they can't tamp down negative coverage, the best hope for generating positive press is to announce action - a crackdown, lawsuits against fake reviewers, or other evidence that they are "doing something" about the issue. Unfortunately for Amazon, they've done this in the past (1, 2) and the problem appears to have gotten worse.

1. October 2015: https://www.theguardian.com/technology/2015/oct/18/amazon-su...

2. April 2016: https://www.seattletimes.com/business/amazon/amazon-sues-all...

> Bezos truly hates negative press

I'm not sure that's true. I think they pride themselves on not worrying about being understood, in other words, not pandering to the press or Wall Street.

My own anecdote hints that it's even worse than just fake reviews... I actually just wrote a detailed multi paragraph 1-star review of auto parts I received from Amazon, with several photos, too.

They, Amazon, would not post it because they claim it doesn't "adhere to the guidelines".

The review was exactly as follows:

These are "SURTRACK/TRAKMOTIVE" brand axles, part numbers BM-8040 and BM-8041. I only installed the left side but since it didn't fit properly on my 2006 BMW 530xi I'm returning them. The OE part number is: 31607570275 and MFG part number is: BM-8041.

The problem was that the inner boot fits improperly and deforms/creases from being stretched when installed (see images).

Looking at the product images for this product on Amazon, as well as elsewhere online, you can tell the boot is very different from the product I received.


Can anyone find anything that goes against their review guidelines in above?

There was a great Planet Money podcast on the genesis of ReviewMeta which I've been using on Amazon reviews since. https://www.npr.org/templates/transcript/transcript.php?stor...

Since listening to that podcast a few weeks ago I've also started checking both ReviewMeta and FakeSpot for every Amazon listing I think about buying.

It's really made shopping on Amazon a frustrating experience -- well over half the products I've checked (in a couple specific product categories) have had F or D ratings according to both sites. I've wondered if some of these are false positives (i.e. ReviewMeta flags a product's reviews as fake but they're actually legitimate), but when you look at their reasoning in the breakdown they give, it's usually pretty convincing that many/most of the reviews are indeed fake.

I feel like this might be a case where ignorance is (mostly) bliss ... when I naively thought the vast majority of reviews on Amazon were genuine, shopping was much faster and simpler, and I mostly had good experiences buying well-reviewed products despite that many of those reviews were probably fake.

Consider that if you go buy an equivalent item at a box store, there aren't any reviews, the company (sometimes companies) with products on offer just happens to have enough money to get the stores to stock them over any potential competition. Similarly with fake and paid reviews, those companies happen to have money to do it (and do it better than competition, they can't all be using the same github script). A company with money to do things like that surely has some amount of money in the product itself, so what you get will probably work out as well as what you'd get from the store...

The bigger problem for Amazon I think is fake products and inventory co-mingling, rather than fake reviews per se. But while others are complaining about the 5-star system not being very good (there's a fancy Bayesian system I've seen before that's better..), at least Amazon shows you the distribution, reviewers can upload photos, and it lets you filter by star count to get a sense of the things said for various ratings.

I'll have to start trying out the fake review spotting tools. But I don't think my current behavior will change that much...

> Consider that if you go buy an equivalent item at a box store, there aren't any reviews, the company (sometimes companies) with products on offer just happens to have enough money to get the stores to stock them over any potential competition.

Well, I know that there aren't any reviews in a B&M store, but (1) I've been checking online reviews (usually Amazon) for non-trivial purchases at B&M stores for a while now (either before I go or at the store on my phone), and (2) B&M stores do usually have at least one human employee in the loop signing off on a product going onto the shelves, and at the very least the "big name" ones (almost) never sell blatantly bogus/fraudulent products (fake Apple-branded MacBook chargers, etc.). Amazon has pretty much been the Wild West ever since they started allowing third party sellers, which I used to be fine with, back when I was under the impression that the reviews were generally pretty reliable, and sure that something that said "Shipped from and sold by Amazon" (or whatever) was a safe bet.

I've been using https://www.fakespot.com/ to check review for fakes on Amazon.

And I've started using BOTH reviewmeta and fakespot. Good results so far.

Reviews on sites like Amazon or Yelp are a monoculture: everyone sees the same set of reviews and the same ratings. This creates leverage that makes it worthwhile for shills to spend large amounts of effort "infecting" those sites with bad information, since once they find effective ways of doing so it affects everyone.

So the solution must be a review system that works like Twitter, where each user has a unique "view" composed of sources they've selected, directly or indirectly. This diversity would make infection much more difficult and less rewarding for shills and other attackers.

I'm still trying to figure out the best design for such a system. The requirements and usage patterns would be quite different from Twitter's, and I'm not aware of any existing attempts that I can learn from.

Yes, tie the review more personally to the reviewer. I guess this is more like the influencer culture we see a on Instagram and YouTube.

> I'm still trying to figure out the best design for such a system.

Easy answer: distributed consensus, à la Bitcoin.

Star ratings in general are just poor. In a 1 to 5 star rating, 3 is average but most people would perceive 3 stars to be above average. People tend to leave reviews only when they are displeased with the product, really pleased with the product or get something in return for making the review, in this case free products.

I tend to just filter on reviews that are 2 or 3 stars and use those to work out the cons of the product. These reviews are the most likely to be honest and can often give you the information that other reviewers forget to mention. Unfortunately, 2 and 3 star reviews are the least likely to be posted so you can only find them on products that have been reviewed hundreds of times.

Really? I think most people perceive anything less than 4 stars (or even 4.5 stars for some) as "bad". I for one won't buy anything that are "merely" 3-stars.

True, I gave a tech book 3 stars once and the author contacted me to request that I reconsider, explaining that anything less than 5 stars could doom a brand new tech book on Amazon. He offered personal support and training in return for raising the score. It didn't work, but it also convinced me that he has a good point regarding the score: Reviews are difficult ground due to their subjective nature, and a lot of reviewers are simply qualitatively bad at writing reviews. If an author just sits back and does nothing, they are risking their brand on the individuals who take the time to write reviews, and that set of individuals could be--even randomly--pretty bad.

For example: I've seen many, many 1-star "product was damaged in shipping" and similar e.g. "received wrong product" one-reviews for products that were absolutely stellar. Getting into even more tricky ground, I've seen 1-star "product is unscientific" reviews left on books containing information that is mainly intractable to science, like studies in history.

So, when you were giving the book 3 stars, how did you judge the book? Did that match what you then learned (3 stars dooms a book). Did you intend to doom the book or feel it should be doomed?

Upon reconsideration I felt that for some readers it should be doomed, while for others it was probably a worthwhile purchase, and I made that clear in my review. Book recommendation is not as straightforward as people think.

Agreed. I have no problems with books being doomed by reviews, I want to avoid dreck as much as the next reader.

Conversion rates in competitive categories on Amazon usually drop by 50% or more when star ratings drop from 4.26 to 4.24. You could have a product converting at 10% go down to 4% or less after the rating drop. The perception difference is absolutely gigantic even if the actual difference is inconsequential.

I think the issue, for me, is not that 3 stars is bad. I probably use a number of products each day that are legitimately 3-star worthy. It's that 3 stars on Amazon et al are an amalgam of 1 and 5 star reviews. There is then a risk, is this product bad but propped up by low reviews? Can I expect poor quality control? Etc

From the distribution of reviews scores, even when separated by category, the majority of ratings given are 4-5 stars (from one of my 2017 analyses: https://minimaxir.com/img/amazon-spark/category_breakdown.pn...)

My full analysis of Amazon rating values is here (https://minimaxir.com/2017/01/amazon-spark/), although it may be obsolete and it's not possible to get better/more recent data anymore.

When I write a review, I want to give 3 stars to most products, products that are good. I want to reserve 4 and 5 stars for things that exceed expectations. But when I read a review, I want to see 5 stars. Anything less than 5 stars means something was wrong. Why did the reviewer hold back? I think it's mostly because there are so many products with 4-5 stars. It makes products with just 3 stars look bad, even if it's 3.5.

For what it's worth, my interpretation as a writer is what the websites meant. Here are Amazon's labels for stars:

  1. I hate it.
  2. I don't like it.
  3. It's okay.
  4. I like it.
  5. I love it.
Even closer to my interpretation is Yelp's:

  1. Eek! Methinks not.
  2. Meh. I've experienced better.
  3. A-OK.
  4. Yay! I'm a fan.
  5. Woohoo! As good as it gets!
I don't think I've ever had anything where afterward I said, "This is as good as it gets." Still, if I leave fewer than 5 stars, I'm sure the proprietor would feel hurt and wonder what went wrong.

This is one reason I don't usually write reviews.

I don't remember where, but I read from a blog article that 5-star reviews are unreliable, when used by the general public, due to ignorance on how to properly rate.

The proper way would be to ditch 5-star reviews in favor of just "thumbs up" and "thumbs down", youtube-style

I'm glad the media seems to finally be paying attention to a problem most people know/have suspected for a long time, though I wonder if Amazon will ever be held accountable in the same way FB has with Cambridge Analytica...seems like it will take a major crisis (someone dying or getting seriously injured buying a fake product on Amazon) for something to actually happen.

The trick on Amazon is to read the two star reviews. The one star can be PEBKAC, competitor stuffing and a lot of nonsense. The real deal very often is in the two star reviews.

I've actually very interested in finding a solution to this problem as well. I think the issue is that regular users who are more likely to give honest reviews currently don't have an incentive to write one as there's nothing to gain, but potentially something to lose. If everyone is rewarded instead of just the ones who give positive reviews (i.e ones who receive free stuff, getting paid for etc,), then the ratio of fake/legit reviews might balance out. And we just need a better UI to show the most relevant + recent reviews for each star rating. For example, in https://www.amazon.com/dp/B07DNTWTKH#customerReviews someone mentioned "All the reviews are non verified!! 5 starts with 100+ reviews...... something is fishy!!". It also shows that 54 people found it helpful. So think that's a neat feature.

I think blockchain might be a good use case for this. Cencorship resistant, decentralized review, not controlled by a single company, and anyone can contribute, and see the reviews (i.e via browser extension). And everyone is rewarded by currency when they provide a review. Just not sure how to prevent spam, and deal with people gaming the system. Perhaps a staking system could be introduced wherein a user would lose their stake if the review proved to be inauthentic (proving it comes with its own problems). Anyway, these are just some of things I've been pondering about before.

Not sure about blockchain but contacting the odd person who has the buying pattern of a regular customer and asking them what they think might work.

I've found that it's more useful to sort by negative reviews. It's reasonable to assume that the product is what it appears to be at face value, but customers tend to speak up if they feel like they got cheated.

On the other hand, I believe I read that some fake reviews are actually negative, as a means to funnel users to competing products. So who knows?

Can't trust anything I read on the internet concerning a product or service.

> some fake reviews are actually negative, as a means to funnel users to competing products

Oh, right. I often look for a few positive ones (to find what people thought was positive enough about this product to mention it) as well as the negative ones (to see if they complain about defect units and other one-off issues, or if there are actually real issues with a product), sorted by newest in the hopes that I don't happen to be looking right after a batch of paid reviews were added. But you're right, there is a reason why negative ones would be fake as well. Seems to be less common, though, but still.

I think this issue is most interesting to look at from the other side. Imagine your goal was to create fake reviews for a product you created. How difficult do you think it would be to achieve given you expect a substantial economic reward for success?

This is why I find it difficult to blame Amazon for things like this - though that is not to say they shouldn't try, but I don't think anybody is saying they're not. It's just that this is something you might call a game of cat and mouse, but I think it may be borderline impossible to detect fake reviews if done well. Imagine I somehow get 500 confederates from geographically diverse regions, and pay them each a substantial amount of money to purchase a product and give a positive review. How do you prove those reviews are fake?

I generally avoid looking at good reviews when shopping due to this. Instead I assume that whatever doohickey I want to buy works as advertised, and look for reasons not to buy. Bad reviews are much more helpful in this regard, as they pinpoint weaknesses or other issues.

It's also much easier to filter the bad reviews. There's fewer reviews to sort through, and you can dismiss the PEBKAC or DOA (except as a data point to see if there's a whole bunch of DOAs).

Just bought a rotating bike wheel light from Amazon Marketplace. It's a novelty item that draws images as a bike wheel rotates using persistence of vision.

I don't think anyone has every heard of the "brand" but it was sent from "the UK" allegedly.

It came with a US plug (I'm in the UK) and software on an 80mm mini-cd. That software is riddled with malware. It also has one of those comedy auto-translated user manuals.

Some of these trade offs I can forgive for the price. I can't forgive the malware at any price.

This is why I am starting to do almost all of my shopping by googling "Wirecutter [product category]". Or CNET for electronics. Amazon's reviews have become so useless and corrupted that I question why anyone would even bother with a genuine review. Luckily, Wirecutter gives fantastic, detailed guidance in just about any category you can think of. I'm finding them so valuable that I believe this would be a great space for more sites to compete in.

ReviewMeta has a great Chrome plugin that shows the "real" score (after removing all even slightly suspicious reviews) right in the toolbar https://chrome.google.com/webstore/detail/reviewmetacom-revi...

The funny thing is I think Chrome plugins suffer from a similar problem and quite dangerous too because they can steal your information. I reported a fake clone of Privacy Badger a while back - positive reviews with the only difference being the fake one had a very low number of reviews/downloads. I'm very conservative with the extensions I use now

I only use chrome extensions in a containerized fashion (i.e. on an instance where chrome is used for nothing else)

News from Australia today: Australia's largest apartment developer Meriton fined $3 million for misleading consumers on TripAdvisor by preventing guests Meriton suspected would give negative reviews from receiving TripAdvisor’s ‘Review Express’ prompt email.


I'm surprised the article doesn't mention the Amazon Vine program which sends products to well-standing reviewers in order to seed reviews.

On a positive note, Amazon has been wiping out thousands of reviews this month. Whether blatant manipulations or legitimate reviews, dozens of sellers have lost their reviews. Amazon has been cracking down hard this month, and erring on the better safe than sorry side as usual with their wide nets. The only problem is that Amazon can't keep up with China.

Jeez, I don't think it's possible for this writer (Ryan Kailath) to be any more milquetoast if he tried. Amazon's reviews are completely bullshit. I never buy anything from Amazon (or Yelp, or...) without running the product page through Fakespot.com to check for fake reviews.

I'm honestly starting to believe that the "wisdom of the crowds" that so many startups and businesses rely on to reach "internet scale", is exactly the wrong approach. Between recommendation algorithms that can be gamed, content that can be paid for, and social media sentiment that can prove to be entirely manufactured, we need to come up with a new paradigm for our products (where "our" is the tech industry at large).

The days of allowing what are essentially anonymous sources to dictate what we do, where we go, what we buy, and how we think are coming to an end.

Interesting to ponder that half of the population is of below average intelligence, and that the other half is generally too busy to be writing reviews.

I hate it too but rather than just complain is it a solvable problem?

Reply All had an resent episode about it


first they had fake reviews. Amazobn required actual purchases so companies would pay people to purchase and return. you can't ban returns otherwise there would be no negative reviews.

how can this be solved? some kind of reputation system? how would you lose rep?

I love to hear your ideas?

Tag the review with the info if the item was returned.

How does one protect oneself from these? Are there secure / private browser extensions that might cross-reference those reviews to other sites? Is everything fake news now?

"We have built a lot of technology to assess whether or not we think a review is authentic”

Really now? Has amazon considered firing those teams? Because that tech is beyond broken.

I find that even books are plagued with this issue. Some obscure technical books selling for more than 80$ will come in a very poor quality print, like they had been printed from a scan. Clearly the book has ran out of original prints, and somewhere they print it cheaply just to make the sell.

Actually these books are likely counterfeit, being produced overseas and sold on Amazon as the real deal. It's easy because neither the publisher nor Amazon (nor the customer for that matter) can know that someone listing a product is shipping a fraud.

What would be the right way to check that? Perhaps sending a mail to the publisher?

That sounds reasonable, sure. You might also try contacting the author directly. Authors tend to know way more about their books than the person answering your email at the publisher, and certainly would have a good idea what their print editions are supposed to look like. I've actually found emailing book authors about their books in niche fields to be usually really easy and interesting in its own right.

Simple hack to this problem is forget about buying stuff online.. go to the store and get it right away.

The Internet can not believed for much and is easily manipulated for the better of those smart or immoral enough (however you view it) to use it to line their pockets.

What if you just removed the stars and only had text? Makes buying based on "reviews" harder...

Could also do something like upvotes but then that just ends up with the same problems as stars.

Always always always, read the negative reviews, NOT the positive reviews. If the worst you find is "shipped broken" or something, it's probably ok. If you're reading negative product reviews instead, stay away.

As a marketer I'm already familiar with fake/paid reviews. It's the way to manipulate and it will be used whenever there is an opportunity for that. On the other hand review.network works on solution for fake reviews

Similar discussion about the fake Amazon review economy a few months ago: https://news.ycombinator.com/item?id=17022215

I have found https://reviewmeta.com/ to be an interesting place to look at the reviews and get a bit more insight into them.

https://text.npr.org/s.php?sId=629800775 - plain text article (sans tracking etc.)

Fakespot.com is your friend. I'm fast finding their service indispensable. Also their iOS app is super convenient - you just share a link in the Amazon app with it, and you unmask all the fake reviews.

There was a great Reply All podcast on this topic. Between it and this NPR article, and recent personal experience, I'm seriously questioning whether I want to continue my Prime account.

Amazon really lost me when they shadow banned me for writing a (very) negative review.

They sell me crap and I can't even complain about it?

Google SEO, Amazon Reviews, AppStore Reviews, Facebook News. Fakes are everywhere. Fake it till you make it [0]. If you really think about it, everything around us has a grain of fake. Ads, startup pitch desks, TV, people relationships, politics. Everything! I almost feel like we're fighting the wrong problem.

[0] https://en.wikipedia.org/wiki/Fake_it_till_you_make_it

Why don't they make a review system where the products with fewest bad reviews gets rated the best.

If the reviews have to go through Amazon's system it should be hard enough to fake. The "true" customers will make bad reviews if they didn't like the product. The "fake" reviewer can't rate their own products to make them get a better review, an they have to buy the competitors product x times to give x bad reviews.

And to further secure it, the amount of buys can be compared to amount of negative reviews. Because who will in their right mind buy a product with a lot of negative reviews, to later give a negative review?

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact