I think independent consumer reviews are a very valuable service to the (shopping) community and, yup, to humanity. It seems contradictory to collate "shopping" and "community", but in general people do spend a lot of time and, duh, money shopping to improve their lives.
Counting on peers to help you shop, what you shop for, how safe and reliable the product is ("Could this toy hurt my kid?", "Is this tattoo remover safe?") are not to be taken lightly.
Online product/services reviews are a huge milestone on how we interact with the world around us. It even causes me anxiety to try a restaurant without having checked it online for reviews... Any effort on how we can better correlate opinion and experience is not just helping "billionaires" here. It's helping everyone who want some feedback to help cut through flaky, moronic marketing that we are bombarded with on a daily basis.
Well yes, but it‘s not like the flaky, moronic marketeers don‘t know people like you exist. So reviews just become another topic of optimization for them.
Also why I never saw the point of sites like quora. Why are all these "subject matter experts" spending countless hours writing free content for a platform that is ultimately locked in and going to either a) Be infested with ads or b) Going to charge money (like the publishers' racket in scientific publishing).
At least for Quora (or rather StackOverflow for me) I could understand it as satisfying a need to be in a community of experts and exchange ideas on more fringe topics. Also it‘s probably a good thing to put on your resume if you have amassed sufficient fake internet points ;)
Amazon reviews however... you‘re just a drop in the ocean and there‘s certainly no community. Thus you‘d have to pay me if you want me to engage in this.
It could also be used to improve your technical writing abilities. Doesn't really matter that you have submitted review number 5397, the effort that you put into writing a persuasive argument is valuable to you in itself.
One of the best dutch webshops, coolblue.nl, actually has people responsible for each market section, they test the products themselves and have recommendations for the items in each category they think are the best. It's supplemented with user reviews, but still, it's nice to get a semi-objective indication for which products to look at. The pride themselves in customer service, only sell a select range of products for each category and you can tell real time and effort went into it, this, in turn makes this shop pretty trusted by people. That's both good and bad, because now companies could approach them to get into their favours, but still, from what I've been able to tell, they actually offer some good recommendations.
This - also things like "you can create a PR to change the microsoft doc's site" - FFS microsoft make so much money but the users are expected to maintain the bs documentation.
I don't get it. You have two paths if you know what's wrong with the docs: 1. tell them it's wrong, 2. tell them how to correct it (pr), 3. ignore it and let yourself or your team run into the same issue in the future.
The amount of money ms makes is irrelevant to any of those points. The docs writers will make mistakes - this is unavoidable. The question is then: do you care whether your coworker sees the same error in the docs.
I'd rather fix their docs than leave that to some random question on an unofficial forum that you have to Google with very specific keywords.
If reviews could be trusted, you would be doing more to help consumers than the business owners. Independent reviews have been a staple of commerce long before it went online.
Reviews really help small businesses who don't have millions to spend on marketing and traffic.
For example I sell online video courses and reviews (legit of course) help quite a bit with sales. This is independently, not on Amazon or another platform.
That’s how they exploit us. They tell us we are helping the small business owner or the phone rep and play on our natural fellow feeling while the main beneficiary is the giant, faceless multinational.
I frequently do those surveys because if I plan to use something more than once I'd like it to be suited better to me, and because I feel I rarely get much info on how happy my own customers are.
In these cases I don't care either way about any billionaires, I care about the product being better for me.
I don't do these surveys when the company is irredeemably horrible (e.g. AT&T, Comcast, most airlines etc) with deeper problems than a mere survey can reflect.
I also abandon them when I it appears the survey is poorly designed (ambiguous questions, nothing but marketings, or simply too long).
I wonder what all the negative side-effects would be of rewarding review writing of certified purchasers with store credit, and if there would be a way to avoid those side-effects.
Not much, no. But I’d argue that’s different anyway. Facebook is a quid pro quo. They provide me a service (distribution) in exchange for my content. Verizon gives me nothing in exchange for my survey response.
I mean, if we continue with that logic you're actually providing it to your ISP, who does what they think will profit the most. That is, delivering it to Amazon. Who will then deliver it to others looking to buy that product.
Sort of beside the point though - I don't actually care about the impact, good or bad, on Amazon. If I write a review it is to help a fellow consumer. If I think my review instead hurts the fellow, or has no effect, I won't do it. But how it impacts Amazon simply doesn't matter to me.
I'm not sure who first made this observation, but it's incredible how slowly, everything becomes labor (in the abstract sense of work performed for someone elses profit). With social media, we spend more and more of our free time working for companies for their profit. It is essentially, the ultimate form of labor from a capitalist perspective— where not only the effects, but also the act of labor itself has been externalized from the cost.
Amazon derives a lot of value from the reviews on their site. They can put people who are skeptical about a product at ease and lead to a purchase, for example.
I tend to ignore good reviews and rely on bad ones mostly to make my decisions. Will also never leave a good review when I'm satisfied, they are indistinguishable from fake ones anyway, I tend to rather just leave bad reviews when I'm not satisfied.
The problem is that bad reviews can be just as fake as good ones, as they could be written by the product's competitors.
Also, if a product is popular enough, odds are that even with a great product there'll be a small fraction of people who'll just happen to get a dud, while most of the rest will be satisfied.
Because of this, I tend to read both good and bad reviews, and give more weight to the longer reviews that go in to great detail about the product and why they rated it the way they did, with comparisons to other products being especially helpful. Also helpful are reviews by professionals who really know the field thoroughly. Unfortunately, such reviews are rare.
I also look at the ratio of good reviews to bad reviews, and look for products with many reviews, though on Amazon this strategy is not quite as effective these days as it once was. One of the big problems with Amazon's model is that it seems that most people buy the products shown on the first page or two of search results. Products on later pages might be just as good or even better, but they may never get bought or reviewed because they weren't on the first couple of pages.
It's become tedious fighting through the endless two word 5 star reviews, 1 star reviews because it was damaged in shipping and so on.
Add that Amazon are determined to show thirty no-name brands of cheap (and nasty) copies of whatever item you searched for, sometimes making the actual market leader really difficult to find, and it becomes more trouble than it's worth. I don't have hours to micro-manage and investigate every £10 purchase for fake product, fake review or not the brand I really wanted.
As a hugely keen early adopter of online shopping, these days I just go to the shops more and more often. It's just a shame there's so many fewer of them!
I love reading bad reviews on Steam, because they usually explain pretty well what the reviewer thinks is wrong with the game. I can then decide if that matters or not to me.
They stopped adding new content to the game? Maybe not too bad. It has a buggy always-online DRM? Deal-braker. Etc.
For me, it's the 4-star and 3-star reviews, and distribution of them compared to the rest (this can be gamed though). I see no reason that somebody would pay for 4 or 3 star reviews, and also competition would pay for 1 or 2 star ones instead. Primitive logic, but worked +-OK for me even on Amazon.
The chap on BBC radio yesterday said that he deliberately gave lower marks (4 star) than 5 because of exactly that - it looked too obvious.
He also said that some of the products that he was paid to review he didn't even use (hemorrhoid cream being one item).
I have also noticed the tit-for-tat reviews on Amazon between competitors. Its quite common on the cheaper products, in my experience. I've bought one or two awful things due to being caught by fake reviews, but I do tend to go back and leave a dreadful review.
I used to do the same, and still sometimes do for bigger investments like good headphones. Other than that i only use Amazon for things im knowledgeable about. For other things where I'm not sure about quality etc and i would have to rely on reviews to make an informed decision i usually don't bother anymore and source it from local brick and mortar stores, either go in person or use their online store.
Blows my mind how amazon is not paying more attention to this issue, i used to get almost everything on amazon - not so much anymore.
This happened to me once, I added my product to a sales channel and a competitor instantly gave it a fake review and lowest possible rating. I guess that was enough to keep people away, as that was the only review for several years.
Yes, bad reviews are often more useful - particularly when they explain clearly enough what the reviewer found unsatisfactory, so that I can make my own judgement as to whether that issue is important to me.
> Will also never leave a good review when I'm satisfied ... tend to rather just leave bad reviews when I'm not satisfied
This is basic consumer behavior. Receiving a good experience (either service or product) is not notable because you paid for it and expect it, whereas a bad experience is offensive and makes you feel cheated so you retaliate by taking your time to leave a bad review.
What this means at scale is that most positive reviews are fake except for the truly extraordinary products/services that are far above all their peers in terms of quality or novelty.
In order to get decent reviews you have to be able to verify that the consumer actually paid for the product/experience and then you have to apply some sort of sampling methodology and statistical analysis to arrive at a meaningful relative score to other products/services in the same industry.
No review site has any interest in doing this because they are just using reviews to generate free content for SEO, to put ads on, and to extort businesses into paying them to "manage" negative reviews in various ways.
When I see a good review, I check other reviews of this person. If there are none or very few, it's a major red flag. If there are also positive, they're not necessarily fake, but maybe this person has very low expectations overall. I imagine someone like the Alec Baldwin character in an episode of "Friends". Of course relying on this isn't fool-proof, the reviewer could be mixing their genuine, "private" reviews with the ones they get paid for, but it's something.
> Will also never leave a good review when I'm satisfied, they are indistinguishable from fake ones anyway
A review can have a certain quality and/or amount of details, that makes it obvious that the reviewer actually owned and used the item. If I think it's worth sharing the quality (or lack of) of an item, I put effort into producing a review that it's clearly real. Reviewers also have a history.
So I think it's not correct to state that the good reviews are always undistinguishable from the fake ones. Of course, with enough effort, a fake review can be made real.
There are some useful good reviews. When author goes into details, comparisons with other stuff, some pros/cons, have photos etc. It's just too much work to do for the sake of fake review. Also if the author could be traced to such good reviews, their other reviews might be more trustworthy as well.
This. I tend to match the reported issues with a product across bad reviews to see if they are just glitches or something particular about it.
When a product has way too many good reviews I usually try to see if similar products show the same reviews. It's outrageous how much copy paste is used.
Let's face it, reviews are useless. Even when they come from professional product critics. Those critics can be compromised (bribed) by big business just the same. A good example of this were reviews for 'Star Wars: The Last Jedi'; most reviews either against and in favor of that movie, by critic or causal viewer, were either fake or compromised.
I tend to read around a number of respected sources and magazines. It's a source of fun for friends and family, but I'm far too cynical with one review or source of reviews.
What I find more informative is bad reviews that get an answer from the seller/company. I guess these interactions will be faked in the next iteration of reviewogenesis.
Fake reviews are a game of cat and mouse. If writing good ones for their own products no longer works, they'll start writing bad ones for competing products.
which is why saas review sites like TrustRadius, G2Cros, etc are useless. Almost every single product has at least a 4/5 rating and there are hardly any negative reviews. Which tells me only people who had a paid incentive left a review. Useless!
I'm starting to recognise that the Internet —despite all its potential benefits— is a platform to streamline the process of lying to each other. Occasionally political, occasionally commercial, and very often social white lies. The instant access it affords us is being exploited by everybody to affect how we think, what we buy and who we like.
It's very rapidly becoming much worse. Machine learning will adapt to people faster than legislation can protect them. And most of all, I don't know how we reverse all this.That's what really scares me. I usually have an answer for everything. All I've got here is: unplug all the things.
We shop in an environment where we are being told what other purchasers thought about it. "Fake news" (made-up clickbait, not the Trumpian term for non-news) spreads like wildfire, polarising anybody who takes it face value.
You can argue that the exposure to these lies is entirely self-inflicted, but that's how society has moved on. The real question is if we can successfully wean ourselves onto something better.
There are hundreds of groups on Facebook in which the seller will ask you to buy the product give them 5 star review and they will return your money (some time with the tip) on Paypal. And you can keep the product.
I only write reviews when my experience is on either extreme.
Expectations exceeded or the disappointment is too much to ignore.
It's kind of like work, innit? Who is going to pat your back when you arrive and leave on time and get your work done? No one. Why should they? It's your job.
Go above and beyond or go in the opposite direction and slack off and you deserve all the feedback coming your way.
I have been thinking about the review problem for a long time and while I think I have a practicable solution, I am not sure if there is any money to be made. If truthful reviews cost more than fake ones is there anyone out there to willing to pay the extra cost? Customers don't appear to want to pay anything nor sellers.
Consumers will pay for reviews they believe are truthful, like Consumer Reports.
But it's far more lucrative for sellers to pay for fake reviews, e.g. celebrity endporsements, magazine articles, youtube videos etc.
So much content out there is just lying about products for sale that I think a for-profit model for honest reviews will inevitably decide to open itself up to fake reviews or paid content. The pressure to do so is just to great and the rewards are hard to turn down.
Are you sure consumers will actually pay? Assuming Consumer Reports is truthful, what percentage of consumers actually pay for it?
My solution actually can't be opened to fake reviews, but it is expensive and I haven't been able to find a work around for that. Unfortunately quality costs.
One thing I've noticed recently is that the number of reviews that are for a completely different product seem to be skyrocketing. This was probably the final straw for me when it came to trusting Amazon reviews, and now I usually look to YouTube to get info. Granted, most of the YouTube reviews are sponsored but it seems like the reviewers are almost always up front about this and it helps to see someone physically using the product.
Given that the review problem seems to be growing worse, I wonder what the priority of fixing this is internally at Amazon. Even if I turn to another source for reviews I'll still probably purchase from Amazon if the price is lower, so I wonder what effect this actually has on their bottom line?
It seems like the only real solution is to make it so time consuming to write a review that the economics of paying someone to do so no longer make sense. I'm not entirely sure how this could be done. Maybe have a longer length requirement? Force a reviewer to upload photos? Someone once commented here that making the reviews ephemeral or decay in importance over time could be a solution, which I thought was an interesting idea.
When researching the first thing I do is A. ignore 5 star reviews and B. read the 1 star reviews first. That's probably the best way to figure out if what you're buying is a piece of trash or a legit product that works. One star reviews are usually written by pissed off customers. I also don't usually leave reviews unless it's a one star piece of garbage that really pisse dme off.
Amazon is plagued by this. I just looked up earbuds on Amazon and clicked on a pair on the first page of results. It's a brand I've never heard of with a design that looks silly, perfect. 473 reviews, ALL 5 star, ALL posted April 13th, multiple from the same users. The laughable part is many are for different products like screen protectors and "lightning wires" (lol). What a joke. How Amazon expects to stay in business is beyond me.
I wouldn't look at the one star reviews either. I usually start with the three star reviews and glance at the four and two stars unless there is a preponderance of one.
What is with HN always deleting the word "Why" from the beginning of titles? I get that it can be a filler word, but it can also significantly change the meaning. For example, the blog post that introduced Julia to the world was titled "Why We Created Julia" [1] and if the "Why" had been deleted it would have completely changed the sense of that title. In this case, it's a fairly benign change, but I've seen this drastically affect the meaning of posts recently. It seems like a bad practice that is at odds with HN's excellent policy of respecting original titles of articles.
For me, the whole "rate product/experience by X amount of stars" system has always been broken. Especially when it seems to be taken much more seriously than it was originally meant. Ok, I'm not a UX expert or a data analyst, but as far as I know, the statistics that are drawn from these star reviews were originally only meant to be some sort of rough pointer to approximately evaluate how good a product is. And the result should in its turn, determine if a product needs to be further reviewed. But on the other hand I get why almost everyone who offers a product or a service utilises this system, it's easy. Is this product good or bad? 1 or 5 stars? Yes or no? The web isn't really compatible for maybes, yes buts, and no howevers.
Wikipedia's article on "reviews" [1] is surprisingly sparse.
I expected to find an overview of approaches to combat manipulation of reviews, e.g. through AI or collaboratively through reputation systems, but alas.
A digital driver’s license/ID card type of thing for using the Internet(by country), especially considering deepfakes. You can still post anonymously but it won’t hold much weight vs. using your real name/ID. Something that if posting junk your reputation/credit/trustworthiness takes a hit!
Yes, that limits the number of identities a review can come from, but it cannot tell if they're lying. It's powerless against approaches like: "Buy my product from Amazon, leave me a 5 star review and I'll pay your fees, a $10 bonus and let you keep the item".
Well if your caught lying your online and offline identity/reputation will take a huge hit. It would be extremely detrimental financially and professionally to your life ... like having to go bankrupt. Be moral and just online or you don’t matter! Otherwise in time the Internet may not matter as much.
Further re deepfake stuff that’s even worse ...threat of going to jail!
How do you vet my published opinion? If I order something from Amazon and give it a five star review, how do you, or Amazon know that the opinion I'm submitting in my review is genuine or not? You can't even vet my criteria at the point of review.
I think that is a key point here. We have started to trust aggregate "opinions" as a source of truth, assuming high numbers will obviate any tampering, assuming tampering could not be done at scale.... Something that just doesn't exist for most things on Amazon.
Well why don't you and millions of others commit crimes... cause there are consequences.
There is zero consequences now for getting paid, bribed, etc for writing fake reviews and little by little helping to destroy the Internet.
There are laws about pollution... why not have laws against polluting the Internet with bribery/racketeering junk that is detrimental to the health of the Internet?
The title should be "Why bbc reads Hacker News comments to get ideas for its next article." I could swear I just read a discussion on HN (just hours ago) in response to an article about Amazon getting flooded with fake reviews. Some suggested how they write fake reviews for money and how one could monetize fake reviews. Then here comes this BBC article. Coincidence. I think not.
I noticed that too. Particularly articles based on reddit and HN posts/comments. Maybe it's because those two are the primary social media platforms I use or maybe those are popular with media workers. The one that comes to mind recently was the last blockbluster article from The NYTimes ( I think ). It featured prominently on reddit and then a few months later, there was an article about it. Also, I noticed a lot of content from reddit in movie and tv show dialogues. I'd be watching something and a dialogue would remind me of a reddit post or comment and then I'd think, I bet the writers of the show uses reddit.
Let amazon hire a product review staff. Let Verizon supervisors monitor the quality of the phone staff. Let uber hire secret shoppers.
These companies are all exploiting your good will to enrich themselves.