Question Can user reviews really be trusted?

[mangaman]

Over the course of many years, many people have gone by user reviews on Amazon, Newegg etc. instead of professional testing results. Couple of my friends from awhile back, said that they always go by the reviews on online shopping sites. If they see a one star review on a well built PSU with many likes, they tend to avoid the PSU. These reviews usually have a few pictures with burned cables or "My PC almost exploded!" as the title.

However, I always go with professional review sites, such as techpowerup, jonnyGURU and [H]ardOCP. If the professionals recommend the PSU, has quality components and passed all of the tests, I'll consider getting it for a build. If not, I'll avoid it at all costs. However, these professional reviews very rarely mention a well know PSU blowing up a system, or "burning cables". I highly doubt the professionals are being paid by the PSU manufactures, to make these testing reviews.

So, can user reviews really be trusted?

 

[WildCard999]

I would think any poor user reviews on high quality parts fall back to the actual user doing something stupid and not the part itself so I typically disregard user reviews for single parts.

I would only really go by a user review for something like a laptop that is a complete system.

 

[saint19]

That depends on you really.

I mean, something like "Wow, the best and amazing thing that you can buy is this one, I tested many others but this is amazing" and checking the profile is a new user or have just a few reviews with almost the same structure.

 

[jonnyguru]

My problem with Amazon and Newegg reviews, and I'm just as guilty as anyone else for doing this, is when a product tends to work as advertised, I just leave it be and don't say anything.

If that product fails, however, I am 100% guaranteed to write a negative review.

So your ratio of good to bad reviews become exponentially lopsided the more units you sell. If you sell 1000 widgets and 990 widgets work as advertised, you may get 100 positive reviews. But if 10 widget fail and you get 10 negative reviews, it looks to the person reading these reviews that the "failure rate" is potentially 10%!!!

I bought a thing to hold firewood the other day. I didn't leave a review. It just worked. I put wood in it and it just sits there. But there were a lot of negative reviews. Why? "Wouldn't fold". "Fell over when I put too much wood in it". "Arrived with scratched paint". All valid complaints, but not the actual MAJORITY for the number of units sold. Out of guilt, I went back and left a 5 star review. But I didn't have much to say so I just typed, "I needed to hold wood. This holds wood. Performs as advertised."

This weekend I bought a Type-C to HDMI adapter and they put something BRILLIANT in the box. A card that says, "leave us a review and send us a screenshot within 4 days of receiving the product and we will send you a $5 Amazon gift card." That's it! A stupid Type-C to HDMI cable I would NEVER leave a review for unless it didn't work! It works. Of course it works. But who would be more likely to leave a review? That one guy it didn't work for! So you better believe I'm leaving a five star review and cashing in on my $5 Amazon gift card!

 

[FALC0N]

Reviews are one data point to consider, but you have to take the time evaluate and weigh them appropriately. There is useful information there. It's just mixed in with a lot fluff.

 

[Wolfshadw]

User reviews need to be taken with a grain of salt. I can't count the number of 1-Star - "Best Product Ever!" reviews I've seen. I've also lost count of the number of 5-star reviews I've seen for products that were released THAT DAY.

For the most part, I don't pay much attention to them unless there are quite a few of them. Less than 100 reviews and I don't even bother looking any deeper. More than that and I'll see how many are 1-star and how many are 2-star. Probably half of all 1-star reviews are likely due to user error. I'll also check the percentage of 4 and 5-star reviews. Over 80% and I'll consider the product.

-Wolf sends

 

[USAFRet]

Shipping was a day late - 1 Star
It works perfectly, but they sent the wrong color - 1 Star
I haven't used it yet - 5 Stars (because they paid me to say this)
Works perfectly with my Canon/Pentax/Fuji/Sony/Nikon - 5 stars (really? You have all those camera bodies?)

Look through the 1 and 2's to see if there is a common pattern of fail.

 

[Elvis Hbpbgrtnk]

My Inglish I still learnings but readings it I knowings what others is sayings. When I is readings reviews if they is soundings just a little too good or bad I is taking salt grain.

When all review sound same or langwage not like people I thinkings it owner writings reviews. My Inglish bad. That differint. They good at it and all sound like they practice good review too much and not sound like real peoples.

 

[DMAN999]

I actually leave reviews of a lot of products that I buy online. I take a few pictures and write my honest opinion of the product whether good, bad or mediocre.
But as stated by everyone above you need to use common sense when reading user reviews.
I also read as many respected professional reviews as possible before buying electronics/PC parts or major purchases.

 

[MoreMoneyThanSense]

I tend to place more value in reviews if there are lots of them, because it helps me get some idea for the distribution overall / law of large numbers and all.

In another thread just today actually, I had expressed some concerns over a particular PSU's failure rate, after having gone through some reviews on Amazon, and Jonnyguru posed the question to me: "503 reviews with an average of 3.5 stars and you're going to focus on the handful of one stars?"

I don't think I really gave a full answer to it but maybe this thread is a better place.

I don't just look at the one-stars, but rather the whole distribution. I actually care more about the number of 4-5 star reviews in general. If I am shopping for the best of the best, in particular I want strong 5-star frequencies.

When I go to Amazon, the first thing I do is sort by Most Recent. This is because when using Top Rated view, you might have negative reviews that show up on top but are several years old -- maybe the product has been improved or fixed since then.

If I scroll through the Most Recent reviews and see a spattering of low ratings, I do get concerned. What jonnyguru says is true in that most happy customers don't leave reviews, whereas unhappy ones will often complain, but I try to see what people are actually complaining about rather than just look at numbers.

If I see a bunch of one-star ratings from people who type in all caps and have questionable reasoning / very one-sided stories, then I won't take them seriously. Or maybe it's user error, or a problem caused by something else, etc. Those ratings I mentally discount.

But if I see many posts where e.g. units are breaking under normal usage, the first thing I do is to to competitor pages and see if the rate of such reviews is comparable. If so, then I can take these reviews with a grain of big grain of salt. If not, then maybe it speaks to an actual issue. Of course to better compare I would need to see if the units sold in each case are also comparable, but that's part of the research process since AFAIK Amazon doesn't release these numbers.

In general I try to get a holistic view of things by looking at number of reviews, what people are saying, and the distribution of their ratings. I'll try to do some research or ask people questions and get a better understanding for which attributes matter, which don't, the tradeoffs involved, etc.

tl;dr yes I think user reviews are very valuable sources of information, but it's not necessarily gospel. Just one thing to consider among many others.

 

[geofelt]

A very good question and some very good answers.

1. I mostly look at amazon and newegg reviews.
2. Only by verified purchasers.
3. I discard dumb negatives such as late delivery, user error and such.

4. If possible, I look at reviews done by reputable sites.

5. Vendors do not intentionally ship faulty products.
Anything you buy is most likely to fail immediately or shortly thereafter.
If you buy from a place with good customer service, you will not go far wrong.

 

[Elvis Hbpbgrtnk]

Maybe customer buy computer and try to shove it up butt or light it on fire or some other stupid crazy thing and then get mad when it broken.

That them and they blame business. Good business get bad customer sometime. Customer no is always right.

 

[jonnyguru]

Maybe customer buy computer and try to shove it up butt or light it on fire or some other stupid crazy thing and then get mad when it broken.

Goodness knows I do this all the time.

 

[Elvis Hbpbgrtnk]

Goodness knows I do this all the time.

It free country and hope everyone happy. Just bad thing if people write bad review if they broked it.

 

[justin.m.beauvais]

A lot of time you can gauge the validity of a review by HOW it is written. I generally ignore the "it wurked gud lolz" posts and focus on those that people take the time to write properly, with some substance. If I'm writing a hardware review I like to add some background. I got this device to solve this problem. Here is what it did well, here is what it didn't do well, I think you should buy it, or not. That is what I look for in a review.

 

[Endre]

Over the course of many years, many people have gone by user reviews on Amazon, Newegg etc. instead of professional testing results. Couple of my friends from awhile back, said that they always go by the reviews on online shopping sites. If they see a one star review on a well built PSU with many likes, they tend to avoid the PSU. These reviews usually have a few pictures with burned cables or "My PC almost exploded!" as the title.

However, I always go with professional review sites, such as techpowerup, jonnyGURU and [H]ardOCP. If the professionals recommend the PSU, has quality components and passed all of the tests, I'll consider getting it for a build. If not, I'll avoid it at all costs. However, these professional reviews very rarely mention a well know PSU blowing up a system, or "burning cables". I highly doubt the professionals are being paid by the PSU manufactures, to make these testing reviews.

So, can user reviews really be trusted?

There’s always a risk!
Even if both professionals and non-professionals give their thumbs up to a product!
Your experience with that certain product might not be as theirs!
But still, I’d listen to their oppinion too!
Also, your past experiences with that brand should play a factor! (If any).

 

[jsmithepa]

Users' reviews tend to be biased and talk about specifically the stuff that bothered them, but they can be valuable as they are real life experiences. The so called "Pro" reviews these days I don't think go into any details, did they REALLY TESTED 10 PSU on the bench? where are the graphs? and the Pros are not working for free, who fund them? It really depends on the item, and things, update-cycles so fast these day you reading item X reviews then updated item Y is already on the market, did they improved on X? nobody knows because it's so new.

 

[mangaman]

Maybe customer buy computer and try to shove it up butt or light it on fire or some other stupid crazy thing and then get mad when it broken.

That them and they blame business. Good business get bad customer sometime. Customer no is always right.

I'll admit, your comment made me chuckle a bit.

But basically everyone is saying:

1. Ignore those with user error/silly mistakes.
2. Use common sense when reading reviews. If it sounds stupid, it most likely is.
3. Many people don't leave reviews for something that works. People tend to leave reviews when it negatively affects them.
4. Go by verified reviews. Non verified reviews go straight to the garbage bin.
5. Anything that deals with late shipping or an incorrect color, should also be ignored.

 

[Elvis Hbpbgrtnk]

I tellings story. By end you will knowings why.

Years ago I had bad terrible horrible job deliver pitza. Tips good rich area, still job terrible company hate employees. I call company Checkers for now. I glad I no do that no more!

Sometime customer cheap mean evil. Orderings every item on menu and they saying "Keep change." Change 2 sents. If you no bow down in praise and sing song of worship they call 800 number. Puttings 800 number all over box just give complayners ideas. I no like it when companys do that or give too much free shipping or always kiss customer ass. It give bad ideas to bad peoples and make them greedy.

Other times customer say pitza make them sick have to go to hospitul. Yet they eat whole thing and then give me more more more for free or I is sayings bad things about business! Wait a minit! You say it nearly kill you!

Customer no is always right!

 

[FALC0N]

It free country and hope everyone happy. Just bad thing if people write bad review if they broked it.

It happens all the time, but not intentionally. A lot of people just don't know what they are doing. My favorite such review story occurred in late 2008.

I was building a system for a friend and looking at Socket AM2+ motherboards for the build. I came across a $80 Biostar board. A lower end board with a simple 3 phase power setup and no VRM heat sinks. The reviewer claimed the board fried on him three times, with Biostar RMAing him a replacement each time. He finally just decided that it was a terrible board and wrote a one star review bashing Biostar as garbage, but was really his own fault.

What he was doing was overclocking a Phenom 9950 to 3.5 ghz, which results in a 300+ watt CPU power draw on the first gen Phenom architecture. The board wasn't even remotely equipped to handle that kind of load. The outcome was predictable. I wouldn't be surprised if he was running a Powork PSU as well, The guy just didn't know what he was doing.

However, I still learned something from it:

1) I learned the reviewer didn't know what he was doing

2) I learned that Biostar honored their warranty

 

[Elvis Hbpbgrtnk]

It late no can sleep no reason, just no sleep for me.

That good story. I no quote whole thing because it long but story havings many good points.

It makings me thinkings if you bisniss owner you better havings good profit margin. OK, you honor warranty and in like story customer push it too hard break it many times and you give them new one many times when it customer fault.

You better havings good profit margin to pay for it.

 

[jankerson]

Over the course of many years, many people have gone by user reviews on Amazon, Newegg etc. instead of professional testing results. Couple of my friends from awhile back, said that they always go by the reviews on online shopping sites. If they see a one star review on a well built PSU with many likes, they tend to avoid the PSU. These reviews usually have a few pictures with burned cables or "My PC almost exploded!" as the title.

However, I always go with professional review sites, such as techpowerup, jonnyGURU and [H]ardOCP. If the professionals recommend the PSU, has quality components and passed all of the tests, I'll consider getting it for a build. If not, I'll avoid it at all costs. However, these professional reviews very rarely mention a well know PSU blowing up a system, or "burning cables". I highly doubt the professionals are being paid by the PSU manufactures, to make these testing reviews.

So, can user reviews really be trusted?

The short answer is no.

Look at the Professional reviews if you want accurate data.

Almost all customer reviews and Amazon or Newegg are not even worth the time to read.

 

[jonnyguru]

It's funny, really. That Type C to HDMI I was talking about? I decided to go back and look at the one star reviews on Amazon. Two of them say the adapter doesn't work with their Dell XPS13. Want to know what I'm using the adapter for right now? A Dell XPS13.

So am I lucky, orrr.......

 

[DMAN999]

It's funny, really. That Type C to HDMI I was talking about? I decided to go back and look at the one star reviews on Amazon. Two of them say the adapter doesn't work with their Dell XPS13. Want to know what I'm using the adapter for right now? A Dell XPS13.

So am I lucky, orrr.......

LMAO

 

[Remeca]

I don't trust user reviews exclusively, but I do rely on them to an extent. If a product was recommended by a trusted pro review, and has a couple of negative reviews that have poor grammar, are obvious user error, or the like, I'll lean towards trusting positive user reviews. If the product is recommended by a trusted review but has a lot of well thought out and comprehensive negative reviews and fewer positive ones, I wouldn't trust the pro review, and move on to another product.