No, but I find myself gravitating much more towards the negative reviews. Not only are they more likely to be legitimate reviews, they tell me the reasons someone didn't like a product, which I find more valuable. Often because a failure for one person might not be an issue I care about: For example, if a reviewer for a book says, "Author spent too much time on the individual people and not enough time moving the plot forward" then I might view that as a <i>positive</i> if I like deep character building. It's a legitimate complaint, but from my perspective it's a feature, not a bug. Or I just might not care: "Item arrived broken" is not something I care much about when it's something that's easy to return/exchange, unless many people are also saying even when it arrived intact, It breaks too easily on use.
I worked for Luma when they were having a review war with Eero. Since Luma is defunct I don't mind straight up telling you.. we were paying a lot of money for positive reviews in order to make our rating match up with Eero's. We had an inferior product with a lot of problems, so it was a losing battle in the end.<p>Probably the most interesting part is that Amazon was one of our lead investors. We came to them for help, and they told us in a wink wink nudge nudge sort of way that our only option was to just keep buying positive reviews.
Lately I’ve been looking at the distribution of reviews. TripAdvisor and Amazon both have bar graphs of rating distributions. In general, I see these distributions:<p>Great products:
Generally a very good product will have between 80 and 90 percent 5-star reviews and then the remaining percentages will tail down with the lowest percent of 1 star reviews. This was how I picked where to go on my honeymoon on TripAdvisor.<p>Fake Products:
All 5-star reviews with zero reviews of any other rating.<p>So so products:
Similar to great products in that the number of reviews decreases as the rating goes down, but the percentage of 4 star reviews is significant and there are less than 70% 5 star reviews<p>Bad products:
Less 5 star reviews than 4 or 3 star reviews. These are surprisingly uncommon and feel quaint when I see them.<p>Mixed supply chain (Fakes) or serious quality issues:
What I refer to as the “1 star bump” where a product has more 1-star reviews than 2, 3, or 4 star review. Similar to “so so products” except you see the 1 star bar sticking out noticeably in the graph. The 1-star bump is my most reliable signal not to buy something. Unfortunately it’s very common on Amazon.
I think reviews have lost most value, but mainly due to Amazon’s co-mingling of inventory, merging similar product listings, and overall massive counterfeit problems. Even if you have a high quality review from a trusted source (Wirecutter? Friends and family?) you have no way to know if what you receive will be equivalent to the thing they endorsed.<p>This happened to me recently with an exercise bike which was an absolute piece of trash compared to what my sister bought and recommended. It makes me want to completely avoid Amazon.
As the trope goes, if the author asks a question in the title of their article, the answer is "no, not really."<p>I'd counter that online reviews still have pretty much the same value they ever have had, and that we're more aware of what the limits of that value is.<p>Lot's of real-person reviews exist on many platforms, and reading through a few often gives you a decent idea of what you're getting yourself into.<p>Paid-for reviews are certainly a scourge, and one business getting loyalists to down-vote a competitor is also a problem. But that's been a thing since the beginning of recommendation features, and we should celebrate a little bit that we're more hip and conscious today to these operations.
One thing I've noticed is that negative reviews, even very specific ones, often just serve to delay my purchase. I was looking to buy an ab wheel recently, and almost every single option on Amazon had some negative reviews which seemed legitimate and did give me pause. "Handle snapped off after just one week, causing me to smash into the floor and loose a tooth", for example. Of course I don't want that to happen, so now I end up spending over an hour looking through products, all which are under $40, none of which are free from some sort of specific complaint - "Handle is too narrow, this is not good for your lats. Can't believe the manufacturer would cheap out on this, I ended up with back pain after less than two weeks of use.", "The additional wheel makes this too stable. I've had many in the past, I clearly do not get as good a workout from this one. Will be returning it." etc.<p>These all have 4-5 star reviews, and otherwise seem perfectly fine. If I saw them in a store I would have purchased them without a second thought.<p>Ultimately I just went back and bought the first one I saw. It seems perfectly fine, I can't see any issues with it.
Yes, but.<p>I don’t Amazon anymore, but the reviews were and I assume still are completely unreliable for many segments. Too many players running to the bottom.<p>For other things, you need to know the market. Restaurants and hotel reviews are full of bad info (fakes, ultra-picky people, etc) but it’s easy to extract useful information. Usually the fakes have details that only workers would care about or are just pure praise without specifics. People on a vacation tell stories and business travelers usually complain about things that cost them time.<p>You also need to consider how you use a review. The stars are there to sell the product. I look for corroborated information about things I don’t like, and mostly ignore the positive. That doesn’t work on Amazon because the fakes are more common and brazen than real reviews, and Amazon facilitates the fakery as more reviews drive transactions.
I've found ReviewMeta (<a href="https://reviewmeta.com/" rel="nofollow">https://reviewmeta.com/</a>) useful for Amazon reviews, though it still doesn't solve the issue of co-mingling products. So generally for buying items I use Amazon or a niche subreddit's recommendation, then buy the product off eBay, the manufacturer's website, or a physical store.<p>As far as "place" reviews, Google reviews in my area haven't let me down yet. Yelp has been hit-or-miss; I dislike their "not currently recommended" review system. TripAdvisor isn't too useful in my local area as there's not nearly as many reviews as on Google/Yelp.
For me, once I became aware of the fact that it is now a commonplace activity to pay for falsified reviews, and there are technology platforms that scan reviews to determine if they are fake or not... almost any review I read on a mainstream shopping platform is suspect to me now. Which is unfortunate, because that's where most of the convenience and value as a consumer is to me.<p>If I now want to make a purchase based heavily on user reviews, I do like many of the other posters here have said: I look at the reviews and low and high middle of the spectrum, have to read each review for obvious, glaring generalizations and seek out language that implies real-world usage vs. hyperbole, and use other features at my disposal, including a verified reviewer stamp (if on Amazon), tools like FakeSpot or ReviewMeta, and reading reviews from competing shopping platforms for the same item if possible (if shopping on Amazon for a widget, look for reviews on eBay or the website of the widget manufacturer). If it's a technology purchase, I'll also look up a bunch of reviews from at least 3-4 technology review platforms, placing a greater weight on those who have included screenshots of their actual usage of the product, which tend to imply a greater deal of rigor to go through for a fake review (but this doesn't mean it's completely trustworthy either).<p>These additional forms of verification, on one hand, could be likened to how we used to shop before internet shopping was a thing (we'd ask a bunch of people for reviews in real life, visit a bunch of stores and window shop, and perhaps read about it in a newspaper or try it ourselves in a real-life demo in the store), but at the end of the day, I choose to shop online for the convenience vs. brick and mortar stores, and this is the unfortunate evolution of that technology.
I now look first at 3 star reviews. These are less likely to be biased and more likely to represent a middle ground between the good and bad aspects of the product. It's sad that we had to get there but the trust is gone with 4-5 star reviews for me.
I was looking for something on Amazon the other day, my searches revealed dozens of odd-sounding brands that I had never heard of before, all of which had lots of 5-star reviews, all of which seemed to be for a different product than the one listed.
Answer is no, as others have said. But I think it depends a lot on which reviews and what you're looking at.<p>I frequently use Google's reviews for finding food places. While it's not 100%, and is HEAVILY biased to the top end (you probably want to aim for 4.2 or higher), it's been very reliable across many locations and even countries for me. Something I like is that a hole in the wall banh mi place is just as likely to earn a high rating as a 3 star restaurant - the ratings reflect the execution of what is being attempted, so cheap eats can rate well and sometimes that's all we want.<p>For games and media, I actually like the aggregators a lot (Metacritic). It's not the final word, and idiots do review bomb some titles (but mostly AAA games that I would avoid), but the "wisdom of crowds" seems surprisingly accurate overall. Steam reviews with their "Overwhelmingly Positive" is similar.<p>For individual items on Amazon or hotel rooms though, as other posters have said the mid-range reviews with something to say beyond the rating are the most worthwhile. I can tell if they're like-minded or whether their concerns mean nothing to me. On this note one amusing thing is that US reviews often focus on "the service" (probably because they're expected to pay extra for it) while I don't really care much about that and don't see it mentioned in other countries unless it's egregious.
Something I've been thinking about recently is that there should be a site that only takes the "good" products from amazon and displays them for different categories - maybe only the top 20 or so. If i'm looking for an oven mitt / phone case / whatever the amount of junk is overwhelming. I envision it would essentially be a semi-curated list of amazon items. Anyone know of something like this? It's on my backlist of projects to build if I can't find any substitutes.
If I'm looking for reviews, I look at results from forums. There's internet communities out there for everything.<p>Forums are usually good for compiling lists of 'approved' items as well.
I have fully shifted away from doing my own research on what to buy to going to recommendation sites. I've always thought I had an astute ability to detect fake reviews (and I was probably right for the first wave of low effort paid reviews). What made me change was the following: 1) Companies and consumers cooperating in huge, fake review rings, 2) product search at places like Amazon becoming totally undifferentiatable, and 3) the lack of trust in tools to combat them (fakespot, etc).<p>I now just go to a recommendation site that earns income through affiliate kickbacks. I use Wirecutter [0], Rtings [1], plus a handful of smaller forum / community sites for categories those don't cover.<p>I actually think this is an enormous opportunity for several huge categories:<p>1) Tool reviews: especially with the rate of battery innovation, there is a huge opportunity to have thorough reviews and category recommendations of both consumer and professional tools.<p>2) Camera / DSLR / video - There's no lack of thorough reviews in this category, but making sense of it is near impossible unless you're a professional.<p>3) IoT gear - This isn't hugely important to me, but similar to cameras and tools there are a variety of systems that lock users in, and getting recommendations based on which system a user already has would be relatively simple.<p>I have found these categories to be lacking not only in trustable reviews, but mostly in updated recommendations. Even if reviews exist, it is up to the effort of the user to read them all and build up a knowledge base of similar options to select from. Each of these categories have a high enough volume and purchase price that affiliate fees could easily support a recommendation business.<p>[0] - <a href="https://thewirecutter.com" rel="nofollow">https://thewirecutter.com</a><p>[1] - <a href="https://www.rtings.com" rel="nofollow">https://www.rtings.com</a>
They haven't lost all value, but you certainly can't take them at face value. I always make sure to read the negative reviews, as others have mentioned. I'll also frequently look at the review history of any review that resonates with me one way or another to see if they appear to be a normal person or a paid shill. Amazon could probably solve this problem if it wanted since it has access to purchase histories, shipping addresses, and review histories for all accounts on its site. I assume that the problem of fake reviews isn't solved because Amazon doesn't care enough about it to take the necessary action and fix it. It may also help Amazon's bottom line to have a bunch of highly-rated products for sale, even if those ratings are fake.
I first and sometimes only read the one star reviews, and look at their percentage. If one-five have a distribution with a big slope discontinuity…<p>If it passes this test, I might read some of the other reviews looking for problems that I might agree with the viewer on.
Considering that anyone can go on Fiverr and pay for hundreds of 5 star glowing (or 1 star negative) reviews for nearly anything that can be reviewed on the web, I would say online reviews are largely BS.
They have not. They've just graduated to a place where you can't lie: Reddit. On Reddit, everyone is sceptical of the guy who doesn't have a long history of comments. People who've only just joined a subreddit will be viewed sceptically.<p>You'll get pretty good reviews if you do "XYZ reddit".<p>The problem for the guy trying to game the reviews is that a real user reviews things too sparsely. There'll be one review in maybe a hundred comments and five submissions.
TrustPilot is just completely ridiculous. For a company who's entire USP revolves around 'real' reviews, the vast, vast majority are so obviously fake.
Same media illiteracy problem as Fake News. Most people are unable to discern ulterior motives. So, "trust".<p>btw; The solution is not improving one's ability to discern (because that is an arm's race the people with power motives (political, financial) are gonna out spend you and vast majority of humans are incapable of beating that) Rather it is by default not to trust and have high bar for earning your trust.
I was getting a screen protector and apparently they're so cheap / specialized that they have no reviews at all. I ended up looking at reviews for other products by the same company.<p>It could also be that they're all dropshippers and change their brand frequently to avoid negative reviews, but the stock graphics were at least slightly tweaked compared to Alibaba's.
I don't trust reviews at all unless I know the "reviewer" in real life, that is if someone I know tells me it's good.<p>Good thing you have 14 day "no questions asked money back" guarantee on anything physical bought online in Europe. I've used that option quite a few times when the product is trash.
It wouldn't solve everything, but I think that the problem could be significantly helped with the following changes:<p>Only allow reviews from people who actually purchased the product.<p>Make it fraud to pay spammers to review your product.<p>Show the country where a seller is located.<p>Make Amazon (and similar marketplaces) liable and on the hook for returns if a seller disappears.
Failed due to Homogeneity, Centralization, Division, Imitation and Emotionality as per <a href="https://en.wikipedia.org/wiki/The_Wisdom_of_Crowds#Failures_of_crowd_intelligence" rel="nofollow">https://en.wikipedia.org/wiki/The_Wisdom_of_Crowds#Failures_...</a>
The star-ratings are nearly worthless, and to the extent there is any value, it is in 2-to-4 star reviews (less likely to be fake).<p>However, the explanation, because harder for a bot to churn out believably, is still often worth something, not least because things that bother other people might be a plus for me.
the reviews are hosted by amazon. they exist to facilitate profitmaking by selling stuff. the company never even pretends to have goals at odds with that. some part of that requires barely pretending to not encourage this behavior... but this is inevitable.
I look for practical information in reviews such as 'works well but the plastic grommet holder snapped off after a week of use, look out for the weak electric dimmer switch' etc etc.
The subjective reviews are mostly a waste of time
For popular things? Yes, not enough of the information is reliable. Unpopular and niche things still benefit from reviews. It's almost as if the quality and usefulness goes up as the review count stays low.
When it comes to entertainment, largely a high level of positive reviews no longer tells me anything because the internet has saturated the planet. Average people are upvoting everything.
A big percentage of reviews on Yelp are fake.<p>Sure you can buy fake yelp reviews on <a href="http://freepage.io" rel="nofollow">http://freepage.io</a> and other sites.
Nope. They just require the same critical thinking and situational awareness that is required to spot phishing, obviously malicious links, dark patterns, etc.
I post frequently here with an anti-advertising stance. Briefly: advertising is often just unambiguously lying, and when it's not, it's still a one-sided collection of facts. Advertising breaks one of the core ideas of capitalism: that the best products or best prices will win, because a better-advertised product can beat a cheaper, higher-quality product. If one company spends money on advertising, all their competitors have to start spending money on advertising, which means that the entire market has to take money away from researching and manufacturing quality, or pass that cost on to the consumer.<p>A common response I get is, "But how will we find out about products?" My stance is that professional review sites supported by users (examples: Consumer Reports, OutdoorGearLab, LabDoor) are the solution.<p>So why not consumer reviews?<p>1. The obvious thing is that it doesn't solve the problem with advertising. Advertisers can pose as consumers and post reviews. Sophisticated methods of detecting this (i.e. ignore all 1- and 5-star reviews) just fall prey to more sophisticated advertisers (i.e. computer generate 30 5-star ratings to keep your star rating up, and then post one carefully-crafted 4-star rating written by a human).<p>2. Even if you do manage to filter advertisers, you still tend to hear from the extremes. Irrational positivity isn't better than irrational negativity: a person who posts all 5-star reviews because they want to be nice isn't better than the furious customer with a grudge.<p>3. Consumer reviews encourage what I'll call second-level advertising: if you can't advertise directly, then you try to get your consumers to advertise for you. For example: spamming your users with "If you like our product please leave a review on iTunes/GooglePlay/Amazon" or the YouTube classic, "Don't forget to Like and Subscribe!"<p>4. Consumers generally aren't experts, which means that consumer reviews are more susceptible to Parkinson's Law of Triviality[1]. A good example of this are the reviews on products that say things like "The parts feel cheap and flimsy." Well, did they break? Probably not, or you would have said that. Titanium parts are thinner and lighter than steel, which sometimes gives them the impression of being cheap and flimsy, when they are in fact more expensive and stronger. But in a more general sense, consumer reviews tend to focus on the look or price of something rather than harder-to-understand things like performance and durability.<p>[1] <a href="https://en.wikipedia.org/wiki/Law_of_triviality" rel="nofollow">https://en.wikipedia.org/wiki/Law_of_triviality</a>
I prefer to read the 1-star reviews. Often, there's nuggets of information written if it's not entirely someone's complaint about a product/thing.