Amazon’s Fake Review Economy
buzzfeed.com
buzzfeed.com
I would like to see Amazon raise the barrier to entry to prevent fly-by-night sellers from listing the same crappy products. I give up sometimes when looking for something after seeing dozens or hundreds of pages of the same thing listed under different brands from different sellers.
Is there really so much benefit to having everyone and their uncle selling on Amazon? I can go to Ebay for that. As Amazon becomes no more trustworthy than Ebay I lost my incentive to shop there.
Say I have an idea for a robot, and I need a few dozen battery clips, some NiMH batteries, a dozen servo motors, a bunch of disc magnets, a box of M3 screws/nuts, some PLA filament, a microcontroller board, and a small display panel to show status/errors/etc.
I can buy all of that and have it delivered in a day or two, for less than $5 more per item than I would pay if I ordered it from a Taobao listing and waited on international shipping. And I tend to have less problems with the Amazon listings, presumably because the sellers know I can return them quickly and painlessly if there are quality issues.
There is a huge benefit to being "the everything store".
At this point, I would rather order my socks from Target and my electronics from Mouser. At least i know they have long standing relationships with distributors and I wont be getting fake products.
For the specific parts I mentioned, I'd probably go to Pololu, but you get the idea.
Also, it would lead to more stories like the other one that was popular today, along the lines of "Amazon Shut Down my Store for No Good Reason".
Your comment demonstrates why price-driven as opposed to value-driven economics are essentially irrational.
"I want to pay as little as possible" turns out to have all kinds of unpleasant outcomes, not least of which is the complete collapse in quality in certain sectors on Amazon.
It's a question of what can be done. Again, you're already paying $12.99 for a $9.99 item, and $6.99-$7.99 for a $3-4 one.
For example, a listing had reviews from prior to 1/18 that mentioned how the product was a metallic ruler however the product page had been completely changed and it was now selling USB wall chargers, using the positive reviews previously posted for credibility. I haven't seen this scam mentioned much. I'd imagine it is people selling their old, unused Amazon product pages to someone else.
Amazon's prices are no longer cut-throat.
Chrome's autofill works across devices so, besides entering my security code, I haven't manually entered my card info in quite some time.
I was looking for gaming mousepad for a known brand and item was sold by Amazon. Majority of the reviews (mainly 3 stars or less ones) were mentioning as sensitivity of the mouse is not good or one of the mouse button got broken quiet fast.
I don't think a well known brand would do this trick but apparently there is a problem and Amazon should fix it.
"When a measure becomes a target, it ceases to be a good measure."
I don't know a way around this with Amazon but it is getting increasingly annoying using a site where everything is 4-5 stars with hundreds or thousands of reviews and you need incredible intuition to tell the fake reviews from the real ones.
It's not perfect, but it's something.
Another strange thing about the "Amazon's Choice" label is this: I've seen a product get that label in part due to a low number of returns ... but the seller doesn't accept returns! That should be a con, not a pro.
As you can probably tell, I have not been happy with the results of following Amazon's product choice advice.
It feels a lot like Consumer Reports to me, where the reviews of products are handled separately from the vendor.
https://bardsandsages.com/juliedawson/2018/05/06/amazon-froz...
It will be like de-monization (of YT videos) all over again.
Anecdotal evidence: when my mother and her team were running for student council (in college), the opposing team introduced blatantly fake votes (photocopied) in favor of her team; my mother's team got disqualified for cheating, and the opposing team won.
So I could certainly picture a way in which a mischievous seller could submit fake-looking reviews for another product and get it taken down.
Then you notice the keywords attached to "Amazon's Choice". Oftentimes, they're strange and esoteric. Oftentimes, that flag is put on what appears to be a random off-brand item sitting right next to a much, much better product more deserving of that "Choice" insignia.
> but the seller doesn't accept returns! That should be a con, not a pro.
US? In the UK, the seller has to accept returns, so that removes one variable (and if the product is bad, you can just send it back).
But now, you're right. It has gotten so difficult to trust any review and I often find myself ping-ponging around competing products trying to intuit what reviews are the least fake. Good reviews have simply become a target that don't necessarily signal a good product.
The socialist solution is some kind of collectively-funded watchdog organization that provides useful review information. The capitalist solution is an organization that straight up charges for reviews, ie, Consumer Reports.
I'm sure someone's working on their startup about it now.
That's more like the liberal social democrat solution.
The Socialist solution is to abolish exchange based production of goods so there's no longer an incentive to lie to increase sales of poor quality products.
While it's definitely a problem, there are WAY more 1-3 star listings than you know, but you'll never see them because the sellers stop selling them or their sales slow down and are dropped from the rankings.
Vendor sells Amazon 500 shirts at $4/unit, retails at $10.
Once in stock, vendor sells the shirts very quickly. They do this by either putting the shirt on sale for pennies, or by using hacked accounts to buy the shirts (or by paying people to buy the shirts).
What happens when Amazon sees that the shirts they ordered sold out in 3 days, and are going to take 30 days to replenish?
They buy more, and although the vendor may have initially been down the $4 * 500, when Amazon places that reorder for 5,000+ units, the vendor is up substantially and bails, leaving Amazon with a ton of inventory.
Just make people research what they're buying elsewhere--which will probably also get swamped with fraud--or just let customers roll the dice without the false sense of security. They're already doing it, but with a bunch of fake 5-stars instead of a bunch of unreviewed items.
I especially hate services that intentionally break their mobile site and disallow forcing the desktop site to make you to install their greasy, slimy mobile app. Like Facebook and Reddit. So scummy.
I got my first question yesterday and I really appreciated being able to give an in-depth answer before everyone else.
If you send that message to random people's inbox as if it's personally addressing them, you're going to get mostly noise in return. Really dumb. And they've seen how it works thus far yet they don't change it.
I saw lots of "I don't know" answers to questions, I guess Amazon realised this also as now they put "I don't know" button to emails.
This is clearly total and utter crap, as anybody who has spent any time on Amazon knows.
What was not mentioned in the article, but is critically important, is that many times you can get more mileage out paying for negative reviews of competitors than you can your own positive reviews. Many times I've been looking at product X only to read in the reviews something like "X was okay...and enjoyed it. But it wasn't near as good as Y"
And over to Y I'm headed. It's an in-site referral.
I've read many negative reviews where it's obvious that the competitors are paying. "Great look, but why can't they have A?" (A is not important, but another vendor offers it and wants to use it as a market differentiator)
I don't see any easy answers to Amazon's problems. As they continue to monopolize all commerce, it'll just get worse.
"Today's Deals" is absolutely infested with brand new products with tons of suspicious reviews, and has been for the last year. Even just checking the date from the first to the last review (e.g. one week) reveals problems for a product with 100+ reviews.
> I don't see any easy answers to Amazon's problems.
There's no "easy" answers, but user reputation is a good start. Right now every review is worth the same as every other review, but looking at the data available we can rate reviews:
- Did the verified buyer actually receive the item before review?
- Is the reviewer's account brand new?
- What does account activity look like outside of reviewing (e.g. Do they browse products? Shop? Check order history? How many non-buy/non-review interactions do we have for every review posted?)
- How did they get back to the product to review it (e.g. direct link? third party site? order history?)
- Did they use a coupon for a discount on the product? Is that coupon posted on Amazon? What % was the coupon? If it is 90% are they even a verified buyer?
- Was the shipment tracked? If not how do we know they even received anything?
Amazon could counter this by offering a big bounty on turning in sellers that do this, and/or financially penalizing them (the sellers) if sufficient evidence is produced.
Machine learning is the obvious hot answer. If you could train a model on how accounts that tend to post bad accounts behave (like another reply said, do they browse a lot? do they tend to hit every product they order via a direct link?), you can set up a decent filter. Even things humans might not necessarily notice, like what time of day they tend to write their reviews or their zip code; AMZN has a treasure trove of data to work with.
Given the other story about AMZN on the front page currently (the book publisher's account being suspended), AMZN probably isn't afraid to let the AI ban accounts straight out and let them apply for reinstatement.
But then again, I also saw something today (sorry can't remember where) that the Data Scientists are all 100% engaged in building models that make people buy more, not fixing things. Because honestly, does this problem make that many people avoid AMZN completely? No.
What I learned from that is that there are people very different from me, with a different way of creating and expressing relationships, loyalty, and adversity.
If anyone is interested... it is my dissertation on developing ML ensembles to classify reviews as fake or not, strictly using features derived from textual analysis (deception theory, stylometry, etc). Behavioral type features, like IP address and relationships between spamming accounts are more useful, but the text still has some value. At some point, I will expand this to using actual Amazon reviews, but I had to drop that idea just to get the dissertation done on time.
Several years ago, the percentage of fake reviews was estimated at 2 to 6%, so I'm sure it has increased to (gut feeling) probably like 10% these days. Is that "tiny"? A really good stat to calculate would be the percentage of products that have more than just a few fake reviews - such that the consumer is given a bad impression. That stat probably is not "tiny" and thus creates the perception Amazon reviews are useless.
[0]: https://spacy.io [1]: https://github.com/Ejhfast/empath-client
That's consistent with the algorithmic estimate quoted in the article:
'the ReviewMeta algorithm labeled 9.1%, or 5.3 million of the dataset's reviews, as “unnatural.”'
I would say that's far from tiny, especially considering how many other reviews (at least in my own experience) have inadequate information in their text. A 3-star review that has only the words "It worked just fine but nothing special" is just as useless to me as a fake 5-star or 1-star review, whatever the fake text.
> “It is difficult to get a man to understand something,
> when his salary depends on his not understanding it.”I have become super wary to buy things from Amazon even with good reviews, because I am suspicious the reviews are fake.
It could be that products with fake positive reviews sell better individually but the presence of fake reviews in general reduces sales overall. But the former is really easy to measure and the latter is much harder.
It's an easy problem to solve too... only allow reviews from people who have purchased the product on amazon. Nearly every other major e-commerce site does that. The fact that amazon hasn't required this is a good sign that they don't think that is the most beneficial thing to them. Their system already has this feature, but most sellers don't enable it.
If someone literally reviews over 90% of what they order on Amazon, they're probably paid for it. If someone only buys things via external links, they're probably paid for it.
There's lots that Amazon can do. Hell, putting seller stickers on co-inventory to track the counterfeiters would be a nice start.
Great idea! They'd still have the benefits of mixing inventory but then you can track who is putting fakes in the bin.
There was no music theory ... just an exhaustive list of scales. The "more than 100 exercises on CD" was simply 30 seconds of the author playing each scale.
So I went back and read the reviews. They were "Just perfect!" and "Exactly what I was looking for!".
Comically, a couple reviews of a different music theory book were obviously writing their review in a foreign language then translating it to English because instead of using the phrase "music theory" they used "music hypothesis".
It wasn't about Amazon specifically, but the sophistication of these scams are really intriguing.
Even the honest ones are forced to get fake reviews, because their competitor is winning with fake reviews. This is the cost of replacing humans with algorithms, before it is on par in accuracy.
The problem here is not replacing humans with algorithms, it's with replacing trusted humans with arbitrary humans (and not even checking whether one human is pretending to be many humans). This is the same problem with online forums and comment sections. We need a way to build a web of trust, so I can tap my network for reviews, but have a way to blacklist the shills who pop up.
There's no reason the "average" from the reviews presented need to count them all... the search ratings could easily exclude reviews by those that didn't have a confirmed purchase.
Something about the language trying too hard that the reviews just came off fake. It's a real bummer having to add this to the mental wading through online junk.
On another front I definitely like the idea of Amazon's Choice, that's based on quantitative things like return rate and what not. I'd like to see more of that data trend in that direction.
It has clearly been designed by someone who has seen a few photos of fast road bikes, but doesn't know or care about any aspect of what makes a bike fast. It's the bicycle equivalent of a cheap Honda that some teenager has absolutely ruined with tacky "upgrades". It might not be obviously apparent to a novice, but the product images genuinely made me laugh out loud.
The reviews are an embarrassment to Amazon, but so is the fact that this product is even for sale. It's an atrocious bicycle that should have never been built in the first place, designed with the sole intention of deceiving naive customers. A quick browse of the bicycles category on Amazon suggests that this is an endemic problem - the customer reviews suggest that many are dangerously defective.
What ends up happening a lot for me is fraud fatigue: I go to buy something and after 20 minutes of studying I don't buy.
There's no quality control on anything anymore it feels like sometimes.
Once you give a gift to a kid, you can't send it back without a fuss and I think that the sellers know this. The return rate for kids stuff must be a fraction of that for other consumer electronics.
Can we even take the Internet seriously re: anything that deals with commerce, politics and etc?
Even with books, especially technical books, I have fallen to the trap of unwittingly buying Indian version of the book (I am an Indian living in India) where chapters are culled, contents are 'adapted', preface and other such items are missing etc. But this could be my fault as I failed to read the fine print, which does exist.
(It enrages me that Indians are forced to use substandard versions of internationally acclaimed books for their reference. I am fine with using gray scale instead of 32M color images, or using thinner paper, smaller fonts etc but altering content, skipping content etc is bad. I am sure there is an economic constraint behind this - Indian versions are usually 1/5th to 1/8th the international price - and often think why don't they use substandard presentation instead of substandard content, but what do I know.
Many times, I buy older version of a book which will be far cheaper, and in 'Indian price range', than the freshly minted, latest and up to date version, especially when it deals with 'static' topics like maths, physics, linguistics etc because I have come to realize that the book companies whip out newer versions with negligible changes just to keep their business running, especially when its supposed to be a text book.)
This was less than six months ago and clearly after Amazon had outlawed the practice in 2016.
"Shutting down disingenuous reviews is tricky for Amazon. By doing so, the company risks alienating sellers, a core part of its business."
In the next paragraph:
"Amazon is the place to be for e-commerce sellers and is increasingly seen as a utility. In 2017, Amazon accounted for nearly half of all e-commerce sales in the US (44%) with $196.8 billion in sales."
If Amazon is the place to be, alienating the bad actor sellers isn't going to be an issue.
I've been one for years. It's an odd insider culture, which is pretty hard to talk about with anyone else. There's a lot of competition among the "Viners" as we call ourselves. And there's also a lot of decision-making. E.g., is the tax-value income and time required actually worth it to me? Many, many times, it's not.
It also seems like there are many more products needing reviews than reviewers. I've never worried about being negative.
And finally, the item is definitely not free:
* Amazon displays the tax value of each item we're offered for review, and reports it to the IRS if we choose it. So literally, the product is taxable income.
* Some items, like books, take a lot more time and effort, compared to their monetary value. But all items require time.
* Some items have collateral costs. Just a couple of examples: I once reviewed an AC/DC converter for the car which blew a fuse when I tried it out. That cost me a trip to the repair shop. Another time I reviewed a floor lamp which arrived destroyed in the box and left my living room speckled in glass shards and styrofoam pieces.
tldr; To me, it's a contractual relationship that I'll continue doing as long as it makes sense to me.
Amazon has a real opportunity to legitimize and enable sellers to offer incentivized reviews on their platform. I’m not sure how Amazon Vine ever worked, but I envision an interface almost similar to AdWords, where sellers can try and narrow down their target audience and offer a select sample an incentive to review, inline with search listings.
Since amazon would know about serial reviewers as well as how critical they are based on prior reviews, they could finely tune who is offered the incentive. I can easily imagine a world where these Amazon ordained incentivized reviews become far more trustworthy than ordinary ones. After all, if a product is good why wouldn’t the seller get a few incentivized reviews through Amazon?
An example of a fix could be take into account of reviews to sales ratio and flag those products with high reviews to sales ratio.
But I read that vendors will also fake bad reviews of competitors, so that might not be a good solution, either.