Hacker News Reader

Comment by strombofulous | Hacker News Reader

I would want to see evidence to show blind is more than just a collection of anecdotes before calling it "data". In my (anecdotal, lol) experience the people who post on blind are a small subset that are way more likely to complain about things, sometimes unreasonably.

View on HN

How is that "small subset" not normalized across all companies, making an equal comparison? You can make a reasonable case that Blind reviews across all companies are worse than the average inside the companies, but it wouldn't explain why some companies are so much worse than others.

Why can you assume it's normalized across all companies?

What feasible reason would cause you to believe two companies' internal morale are the same, even though only one of them has disproportionately negative reviews?

Feel free to only visit 2.5 star restaurants on Yelp because clearly that must just be disgruntled people disproportionately for no reason.

Off the top of my head, one reason could be that internally discussing grievances is more acceptable at one company, so there is less of a mental barrier to discussing it publicly. Or (and this is actually pretty likely, I think) maybe even just seeing people complain about Amazon online all day makes people more comfortable doing it themselves on Blind. I bet I could come up with more reasons if I thought a bit longer.

I'm not saying I think I'm right in this case but I also am not saying I think websites like blind/ratemyprofessor/etc accurately reflect how the general population actually feels

You’re grasping at straws.

Going by the Yelp restaurant review example, perhaps the people who rate a restaurant are either very satisfied or dissatisfied. In the case of Amazon ratings on Blind, it seems that no one is willing to advocate for the company.

Aside from that, the toxic work culture of Amazon and their high attrition rate has been well-known for years and years and famously reported in NYTimes etc.

> I would want to see evidence to show blind is more than just a collection of anecdotes before calling it "data".

Didn't Bezos say something like if anecdotes and data disagrees, its the anecdotes that are usually right?

Reply on news.ycombinator.com