(62 points, 27 days ago, 15 comments) https://news.ycombinator.com/item?id=44184974
I'm sure detection is getting harder as LLMs' writing patterns become less predictable, but I frequently come across reviews on Amazon that are so blatantly written by ChatGPT. A lot of these fake reviewers aren't particularly sneaky about it.
It's pretty easy to spot obviously unrelated reviews that talk about or include pictures of completely different products. What's hard to spot is similar reviews written by bots or people paid to write as many reviews as possible using similar language, especially when there are thousands of reviews.
One issue is that seller warnings would appear on Prime delivered products, which meant that the risk is then pretty much zero for the buyer.
The ratings gradings system wasn't very reliable either. I bought a few things that were rated "F" but were fine.
Today I go for a combination of sales + ratings. Amazon also has a warning for some things that are "frequently returned items" or a notice that "customers usually keep this item." And then I buy Prime delivered items, and a return is not an issue for me then.
I don't know if it's fair for me to armchair quarterback, but still - what was their business model when they decided to do the acquisition? From the outside looking in barely did anything whatsoever.
I browse Amazon using Firefox extremely often and I don't recall seeing any helper UI pop up. Even so, what would have been their strategy to monetize me? User data? Commissions? Some kind of Mozilla+ subscription?
I love FF and cheer Mozilla on where I can, but honestly these decisions are inscrutable.
I'm sure there will be a replacement though, and I'm sure they will go hard with referral links.
Why do you think VPs love new projects / products so much?
I dismissed it quickly because fake reviews is not a problem I have. Maybe I'm not the target market? I do buy a lot on Amazon but can't recall ever thinking I felt burned by fake reviews.
It's a move straight out of Google's playbook, with the glaring flaw of them not being Google, and their user base likes them for not being Google.
Honestly, Mozilla gives me gnome vibes. They're so caught up believing their own spiel that they don't understand why they keep missing the mark.
Before AI, a lot of search result gamesmanship looked more like bad Amazon reviews. But leading-edge fraud is far past “humans pretending to be real, U.S.-based consumers/posters on a website.” The tools don’t generalize anymore.
Just relying on consumer judgement has certainly proven to be inadequate in combating fake reviews, and without incentive, we're not going to get many legitimate reviews.
TheReviewIndex.com I didn't find to be very helpful, as it doesn't index all products and sometimes just refuses to check on listings you ask it to. It seems to have some kind of subscription model, but they don't list the price and offer some kind of enterprise model that doesn't sound like it has anything to do with checking reviews.
SearchBestSellers.com isn't for checking individual products, but it will show you the top sellers for each category so you can get an idea of what could be good in the category you're looking for
Camelcamelcamel.com is a price watch tool that will also show you some historical info on a product & notify you if you sign up and want to be emailed when a price drop occurs
There are a few others on AlternativeTo that weren't there the last time I checked. https://alternativeto.net/software/fakespot/
On Reddit, some people were mentioning alternatives, including asking ChatGPT about the product and it might have some kinda helpful advice, but nothing like Fakespot offered. https://www.reddit.com/r/firefox/comments/1ktm4g4/now_that_f...
If you use something else, have found a good alternative or a particular prompt you've tried in your favorite LLM to get info on an Amazon product, let us know!
I'm actively working on a prototype and have a landing page at https://www.truestar.pro if you want to get notified about when I launch.
I buy extensively from Amazon across a number of product categories. My order history shows purchases as far back as 2005 (though I cannot be sure given I remember buying things in 1998 while in college, probably on a different account). During the intervening 20 years I can count on one hand the number of products I ordered which weren’t legitimate, matched my—admittedly moderate expectations for any commercial product—or included overhyped reviews.
I’d be interested in a service like yours if I could understand how the cost would cover itself in benefits.
It makes sense on paper, if the service helps confirm legitimate reviews for you and convinces you to purchase said product, they should receive attribution revenue for helping generate the purchase.
The reality is much much messier though, because often times the people who award attribution revenue have a conflict of interest against any service that could even potentially expose bad practices happening on their marketplace.
I once worked for a popular deal site that developed a price tracking extension, a certain marketplace threatened to completely ban us from attribution revenue and we had to kill the extension over night despite our users loving it.
IMHO judging these random Chinese products with the nonsense capital letter brands by actual reviews is a lost cause.
If it does end up being a bad buy, Amazon typically has a 30 day return policy for most items. Use that and get something else.
Amazon used to offer this as a service!
This combined with Amazon’s commingling of inventory of Amazon corporate sourced items and third party seller items results in a status quo in which, when purchasing an item on a page operated by the first party manufacturer and/or first party supply chain, the Amazon item picking system may still fulfill that order via inventory sourced by third party Fulfilled by Amazon sellers who knowingly and unknowingly are selling counterfeit products. You never know what you’re going to get with Amazon, and neither does Amazon or the third party sellers. It’s insane.
Scammers are somehow using Amazon itself as an A/B test for if your fakes pass muster, from what I can tell, and everyone loses but Amazon and the bad guys. How long must this continue?
I haven’t done this myself, but I have discovered that it is not allowed to ship or mail items with lithium ion batteries that are likely or suspected to be or actually are damaged, which came in handy when I discovered that a previously working device suddenly stopped charging within the return period. Amazon said I had to work with the seller directly through Amazon, which I did, and when they offered to replace the item and I desired a refund, they refused and stopped responding. I elevated the issue to Amazon and they asked me to return the item, which I was unable to in good conscience do, as I could not attest to the shipper that it was safe to mail, as it had a non-removable battery that would now no longer charge. So Amazon said please don’t ship it, but to dispose of it according to my local disposal regulations.
In the interest of public safety, I told a lot of people about this important issue at my own freebooted unaffiliated DEF CON 30 talk outside while a bomb threat or something caused Caesar’s to be on lockdown. At this talk, I gave away the affected device, a Ledger Nano X which would work when powered via USB C but would not charge or work unplugged. All features and functions still worked otherwise.
> You might want to work on your pitch, it comes across as a little crazy haha
It’s funny you mention that, as I really had to almost haggle to give it away, it was really a kind of comedy routine that occurred to me in the moment, and it was hilarious. Think about the tone of delivery of spam emails. The delivery mechanism itself is worded in such a way that it weeds out folks not receptive to the message. The message is the medium. It’s the multi sensory experience of being appealed to which is the payload that runs on vulnerable processors of susceptible minds, if you ingest it in the way presented and intended.
Thank you for coming to my socially engineered TED Talk re-enactment. I had a lot of fun that year and will be going to DEF CON again this year in a month or so too!
You've got stores that would include a $5-$20 coupon/gift card in the item in exchange for a positive review. Sure, this didn't 1:1 translate but if a user did it would look like a legitimate review.
You've got a plethora of LLMs out there just itching to GENERATE.
Then an expensive option I was suprised happened - I bought a Dyson clone vacuum cleaner off of Amazon. A few weeks later, the company emailed me and said 'We have a new model. Buy that one, leave a review, we'll refund the purchase'. So I did it. This happened about 10 more times in 2024. My outdoor shed is entirely stick vacuums.
Feel a bit dirty doing it but that's ok I've got 12 vacuums that can clean my conscience.
I think Fakespot would have difficulty with all 3 of these scenarios.
Plus side looks like the product doesn’t exist on Amazon so guess there’s a victory there somewhere.
Last year I (like a fool) purchased some chunky thru-hole MOSFETs on Amazon. Lo and behold, despite the datasheets promising a few amps with 3.3V at the gate, I only got a few milliamps. Obviously counterfeit - but no matter how hard I tried or how much indirection I employed, Amazon always took down my review warning others of this verifiable fact.
Lots of incomprehensible or useless human ones though.
(And bad machine translations by aliexpress…)
One of these Results of X is still selling the actual quality product, but there is no way for you to ascertain it because you can't trust the reviews, nor the sold amount because they might as well just be good at tricking people.
Aliexpress just fake it themselves. Search for anything, sort by the number of orders, open the product page for the first result.
Next to the number of sales there's going to be a tooltip saying "Sales and ratings are calculated based on all identical products from the platform."
Under reviews there's going to be a message saying "The reviews displayed are from various sellers for similar product in AliExpress."
In other words, they might as well say that these numbers and reviews have absolutely no relation to the specific product you're thinking about buying, they're just there to increase your confidence.
(average aliexpress review on many tech items)
Amazon isn't exactly cheaper anymore, certainly not when you factor in shipping, their shipping times are awful, typically a week or more and you can't trust the reviews. They do have the larger selection of stuff, so if you can bundle a whole bunch of things it might still make sense. The problem is that you can't really find anything anymore and a large percentage of the stuff that you can only get on Amazon does not ship to your country.
Buy $50 something from aliexpress, doesn't work, you can't do anything. Seller wont refund directly, you need to send the item back... to china... and fill out export forms and pay more than $50 for registered mail.
Amazon? Doesn't work? Doesn't matter why, here's the return label, we'll refund you the moment we get the return.
So it really depends where you live.
It's similar for "shark attorneys," who will typically hail from tier 2 and 3 schools. They're the aggressive hustlers.
They have so many flavors of fraud that it’s very hard to get it right consistently at scale.
Not am Amazon fan, and please let’s not do the Reddit “understanding something is the same as excusing it” thing.
It also reminds me of one of the biggest apartment complex management companies, Graystar using a similar method by bribing applicants with $500 off the security deposit for a 5 star review on Google maps.
Having too many things is just abstract unless you had that problem maybe
The OP is effectively taking thousands of dollars in bribes to erode public trust. I think even a child would see that this is wrong.
I know every man has their price, but I hope when the time comes my price will be higher than "a bunch of vacuums I don't need and I can't even be bothered to sell".
The plan was to flip them on FB market place but I've just hoarded them.
1. The whitelabeled junk is getting very good. In some categories, the brand name stuff has degraded to the point that the aliexpress version is better and cheaper.
2. The IoTification of everything means that a lot of traditionally long lasting items are as durable as their WiFi board - i.e. not very. This also plays into number (1) - where cheap, Chinese, items either lack IoT features or provide them only locally instead of requiring an online account.
anyway, I can imagine some small territory in time where fakespot can accurately deal with the flood. But then..
I had to leave a video review component (No face). I wonder if any shoppers ever wondered why the same monotone man was constantly buying and reviewing vaccuum cleaners if he's always leaving positive reviews?!
It doesnt even need to be that complicated. I worked reputation management for an ecommerce place for a while a few years back. I literally asked very politely against a random sampling of all orders if they would consider leaving us a review, and significantly more actually did than I would ever have expected, with no reward or value in it for them doing so.
I got 100s of reviews this way in the span of a month or two. Enough on a geographically important centralised reviews location to raise the average rating signficantly.
Uh, this is how it's supposed to work? Make a good product, get good reviews for free.
"Make a crappy / mediocre product and pay people to write good reviews" is completely different.
Note: I did not last long in this business before hitting the eject button.
This sheds (no pun intended) some light on why they think there are avid vacuum collectors.
Thank you for the laugh.
But why keep them all? Why not give some away to friends or neighbours, or even sell them?
Why would you want 12 vacuums? What are you going to do with them? Isn't that a senseless amount of redundant objects to horde? Don't you want room for other things in your shed?
Unfortunately, they haven't really countered the "keep creating new accounts" drop-shippers. Some categories are especially bad about this- if you find a back massager that you like, buy it in bulk right away, because the model and probably seller won't be around by the time you want another.
> Mozilla integrated Fakespot's technology directly into Firefox as the "Mozilla Review Checker" feature, making it easier than ever for users to verify product reviews without installing separate extensions.
If it was integrated directly into Firefox, it's funny that I don't recall ever seeing it. I wonder if it gets disabled if you set your security and privacy settings too high, or if you use the Firefox ESR versions (Extended Support Release).
Searching the product and reading about it from different review sites seems more reliable. Also can combine this with marketplace reviews considering reliability.
If there is no review than have to trust the brand and if there is no brand then it is a gamble
Although at least the platform can know if the reviewer actually purchased the product(?)
> Searching the product and reading about it from different review sites seems more reliable
Unless they use affiliate links, which is a great big red flag that the incentives are already stacked against you.
“Musk is that which smells by itself, not what the perfumer says (about it).”
This line is from Saadi Shirazi, the classical Persian poet which has become a proverb in Persian speaking world. Reviews are at this point what the seller wants you to read.
As long as Amazon is the seller, and host of the reviews there is no way to trust Amazon would be fair in hosting those reviews.
The only way to know about a product is to read about it elsewhere like New York Times which is not selling the product themselves.
https://johnnydecimal.com/20-29-communication/22-blog/22.00....
So you know what we do now? Ignore the overall rating: it's worthless. Instead, go directly to the 1*. They're the only true indicator of a product/place/service.
I'm not saying take them all at face value. You still have to put in some work. But all the data is in the one-stars.
https://www.fakespot.com/privacy-policy
Look at Section 2B
B. Personal Information Collected Automatically
We may collect personal information automatically when you use our Services.
Automatic Collection of Personal Information.
We may collect the following information automatically when you use our Services:
Contact Information:
Your email address
Identifiers:
User ID: Such as screen name, handle, account ID, or other user- or account-level ID that can be used to identify a particular user or account. This information could be provided via your Fakespot account, Apple ID, Google Account, or other accounts you may use on the Services. User ID also includes your account password, other credentials, security questions, and confirmation codes.
Device ID: Your device information which includes, but is not limited to, information about your web browser, IP address, time zone, and some of the cookies that are installed on your device.
Usage Data:
Product Interaction: How you interact with our Services and what features you use within the Services, including Fakespot’s sort bar, highlights, review grade, seller ratings, alternative sellers, settings and popups.
Other Usage Data: Individual web pages or products that you view, what websites or search terms referred you to the Service, and other information about how you interact with the Service.
Browser Information: Information your internet browser provides when you access and use our Services.
Application Search History: Information you provide when you perform searches in our Services.
Purchase Information: Your purchase history or purchase tendencies which we may use to recommend better products and sellers.
Location Information. We may collect your location information, such as geolocation based on your IP address in connection with your use of our Services.
Publicly Available Information. In providing our Services we may collect data (including personal information such as profile names of reviewers) that is made publicly available via the internet on the websites analyzed and crawled by our Services.I also got offered some money over telegram to review a hotel from a large chain and leave a positive review.
doppio19•15h ago
rasz•13h ago
i80and•13h ago
doppio19•13h ago
ashoeafoot•12h ago