sample size
(midwest.social)
This meme demonstrated the importance of buying enough fake reviews to make it seem real.
You know how XKCD sometimes draws fantasy maps of like, the internet or something? I'm picturing a map like that of internet marketplaces, and there's a dark amorphous cloud whose nature I'm not entirely sure of that strongly compels sellers to need 5-star reviews like a smack addict.
In a healthy, honestly run marketplace, 5.0 ratings will functionally never happen. Because idiots exist. For example, I bought a small inverter that runs off of drill batteries. It's powered from a 20v lithium ion battery pack meant to run power tools, it has a standard American Type B power socket on it to power things that run on AC, but it's limited to 150 watts. The reviews were divided into two camps: 4 and 5 star reviews from those who understood that last sentence, and 1 and 2 star reviews from those who don't.
I really don't want to buy a product that has a few 5 star reviews, because that almost certainly means it's dropshipped Chinese garbage, the reviews are either fake or coerced.
Online reviews are tough. I've noticed a lot of what I call buyer's bias (it probably has a real, actual name, but whatever). People tend to review products they bought higher than their actual impression of the product. People don't want to feel ripped off by any measure, so it's not uncommon to accept flaws just to feel good about something. This might only affect some people, but it definitely feels like the ratings are biased to higher scores due to this.
I usually evaluate reviews by looking at the negative ones and seeing what they are complaining about. If I'm looking at Amazon reviews for a product sitting on a shelf in a store in front of me and most of the complaints have to do with the seller rather than the product itself, then I figure it's a safe bet.
Nah, both ratings are crap. What everyone should to is to go read 1-2 stars reviews and 5 star reviews. If all 5 star reviews are same/similar in description, disregard 5 stars. If 1-2 and sometimes 3 stars reviews complain about the same thing which is something actually about the product and is not "delivery too slow", disregard the purchase.
These 3k 5 star reviews can be transferred from another product or are bot reviews. Unless 1-2 stars are complaining about something trivial or can be classified as independent issues, you should seek a seller and a product somewhere else.
Absolutely agree. 5-star reviews are cheap to buy in bulk anyway so I don't trust them.
8 Best Places To Buy Google Reviews (Updated List)
My own preference is to go through the 1 and 2-star reviews looking for patterns. Any company can have a bad product slip through quality control and reach a customer. But a company that lets 15 products with the same problem get to customers likely has poor QC. I also check those low-star reviews for replies from a customer service rep. Like I said, any company can have a bad product reach a customer. What the company does when that happens makes a world of difference.
I once had an air fryer that had a switch stop working within a month of buying it. When I contacted the company, they had me cut the power cord and send them a picture of the fryer showing the cut cord, and immediately sent out a new one which has worked perfectly ever since.
It's also crazy that "5 stars" has become "meets expectations", and anything less than 5 stars is immediately suspicious. But, anything with all 5 stars is also suspicious.
Something I remember reading years ago was, if it’s an important purchase, go read the 3 star reviews. They’re likely more nuanced and offer better pros and cons.
I remember getting an oil change or something and the service advisor asking for me to give a review and “anything less than 10/10 is considered a failure” which really pissed me off. In my mind a 10 out of 10 is “above and beyond amazing” which is simply not something I can do for something like an oil change. I would say “I got out of there driving the same car I went in with in a reasonable amount of time” is the best I would expect and would rate it as 8//10.
Yeah a 10/10 for an oil change would be something like they fixed a problem free of charge since they were already in there. A 9 is "great price and surprisingly fast"
One of the many many shitty things companies have managed to do with review culture is make it be about the bottom rung employee instead of the company or their overall product. They actually found a way to feed extra off the resentment people hold for them by kicking downwards even harder.
Reminds me when I got my first car 15 years ago I got a call to answer a survey about the experience with the sale.
I was giving 3s and 4s and just a few 5s (out of 5) and the rep was very worried that I had a very bad experience. I told him the 3 is good, 4 was very good and they did a few excellent things.
He explained that anything below 5 was very bad feedback to the sales people and everyone involved and that they would get reprimanded for the 3s. I declined to continue the survey
There was some kerfluffle in the gaming press like this; a lot of magazines used to score games out of 10, with a 7/10 being effectively the lowest possible score. "When I booted the game up, my C64 caught fire, as did my Spectrum, which was turned off at the time and not relevant to this review. Then the developer kicked in the door, gave everyone in the building AIDS, then went on a worldwide tour kicking every single puppy. 7/10."
There was a magazine that was in the habit of actually using the entire scale with a "meh, s'alright" being a 5/10, and some developers outright blacklisted them for it.
It's like how grades of meat are all positive sounding. Which is the worst grade of meat: Prime, Select or Choice?
I once saw a documentary with the engineer who worked on the toilet equipment such as it was on the Apollo spacecraft. Urine was collected essentially by a condom with a hose at the end, which came in three sizes. Of course none of the astronauts would be caught dead ordering anything but a large so they were labeled "large, gigantic and humongous."
I like to ask if it's a corporate survey or if they want my honest opinion.
I hate those surveys. I’ve always had them emailed though. My last car I ended up ordering and they were relentless with asking me to complete the survey. Sure, some of that survey is about the dealership experience but a lot of it is about the vehicle itself. I don’t have the vehicle yet. It’s on order. So I docked them heavily on the sales part of the survey after that. Don’t badger me to do something I can’t even complete.
One of the interesting differences about reviews in Japan is how people give ratings for restaurants close to a normal distribution here.
Like, most places tend to hover around 2.6 stars on Tabelog, so anything from 2 to 3 stars is usually average or decent.
You'll see this on Japanese apps/websites, but maybe less so on Google Maps because foreigners/tourists tend to give 5 stars a lot more freely.
Tabelog was such a great resource when I visited Japan! It took a minute to adjust to a normal rating scale instead of “anything less than 4.5 stars is bad” but once I did it was great. I noticed less obsession with portion size on Tabelog too, which I really appreciated. I’m not usually eating out to save money, I want to know if the food is good!
Vice versa, Americans tend to give very high scores (4 stars is “not great” instead of “pretty good”), so in touristic places you’d see the tourist traps get bigger ratings than the local spots. I noticed that a lot in Europe.
It's more about the level of satisfaction for Americans rather than a true-to-scale rating I suppose.
I always noticed on metacritic or rotten tomatoes ~70 is a pretty well received rating, though. Maybe media is more scrutinized than food, but it's interesting that we have different "modes" of assessment across subjects even within one culture.
I start from the bottom up, with the one star ratings. Sometimes a lower rating may be because of a shipping error or something silly like a person not liking the color. The ones that I find most concerning are the DoA products, ones that break quickly or have some other flaw that affects usability. Then move up to the two and three star reviews. If a product has enough of the mentioned issues then it may not be worth digging further.
I think I'd still have the same reaction if the number of reviews was reversed. A 5.0 rating nowadays suggests some review farming shenanigans.
Well yeah, that actually totally fits into a third panel of the meme. You would never expect a perfect score with 1000 of reviews, there are always some reviews of people being dicks in the least, just like you would never expect this statistically in your data.
There's always that one guy giving a seller a one star review because the local delivery guy fucked something up.
There’s also the other person who bought the wrong part—1 star didn’t fit my PC
Yeah, I like my free samples big!
