←back to thread

97 points healsdata | 1 comments | | HN request time: 0.246s | source
Show context
mrweasel ◴[] No.44376983[source]
Online reviews in general are pretty useless these days. We know that sites like Trustpilot will take down negative reviews if you pay them, Amazon reviews are mostly bots and some sites have weird incentives for users to write reviews.

E.g. take reviews of business on Google, there's no link to actual purchases, but you get a star and a "Local guide level 4" or something if you do enough reviews. A family member runs a consulting business, he has a 2-star review, the only review. It's not made by a customer, just some random dude. What it looks like is that this dude just walked around reviewing business after business, based on look of their office perhaps. He's not customer of ANY of them. So now multiple business are trying to have these negative reviews removed, Google doesn't give a shit, so what are these reviews actually worth?

Most people who write reviews aren't exactly the most mentally stable people either. If you're not getting something in return, most people won't write a review, that just leave the nut jobs.

replies(9): >>44377102 #>>44377135 #>>44377197 #>>44377202 #>>44377251 #>>44377478 #>>44378404 #>>44379069 #>>44380149 #
1. Larrikin ◴[] No.44379069[source]
The default mind set for a normal person should be there is no reason for them to waste their time to improve a closed source data set for a for profit company. You can break this mind set if you truly care about the product and they are small enough to matter, like a local restaurant or single dev software, but we should not be contributing our time to helping Amazon weed out the complete crap USB cords from the only kinda crap USB cords. Professional organizations like consumer report and America's Test Kitchen can do it at scale and smaller reviewers (who refuse free products) can handle more niche things in the interim.

But I believe open data sets will become as important as open source for the future. Filtering out the spam, fakes and slop will be similar work to what AdBlock filter people do today.