Worth considering the biased data. Levels will only have data from job seekers- which would imply that they either have A) recent offer or B) salaries from employees who are considering jumping ship (biased towards employees who are underperforming or in underperforming business units)

There is probably a missing category of employees who have been successful, gotten raises, and aren't shopping for jobs

I wish it were easier to compare that against other considerations. Comp isn't everything (though not judging those for whom it is) and it can be challenging to weigh the benefits of being somewhere for multiple years. Especially for places with steep learning curves and very large orgs where getting established is a very different day to day type of stress than continuing on and growing in something familiar.

Levels is garbage data. How is this not obvious?

One reason it’s not obvious is that several findings of my analysis aligned neatly with things the recruiter told me about the company’s comp structure.

I know there is bias in who enters data on levels, but is there something else that makes it garbage data? Can you elaborate? Because I have never seen it as 'garbage data' and would like to know more.

This is not obvious to me. I had access to comp data at two different companies and Levels data was pretty accurate across the board.

I’ve never seen it be close to any job I’ve ever had access to data for