[danpalmer]

This data is really not that useful, or doesn't reflect what the simple title suggests it does.

There doesn't appear to be any attempt to control for job level, and it's not even clear how one would control for such a thing. Public company stock and private company stock appear to be valued the same which may be true on the surface, but is a wild misrepresentation of reality (in many ways). Different companies need different distributions of skills, so would have different median salaries even though they might be paying the same amount for the same job. This doesn't account for geographical distribution, those centered in the bay area are likely going to be higher up this list despite others potentially paying much better in their regions.

[miketromba]

Agreed, added a wall of disclaimers to clarify what it is and what it isn't.

[altdataseller]

Would it not be possible to look at one single level across all company? ie IC4 or something?

[danpalmer]

The definitions of levels vary wildly and there just isn't a good way to compare them. People can be justifiably one level at one company, while being quite a different level at another company.

[leetcrew]

not all companies have the same highly structured leveling system as amazon/google/facebook. look at the comp by level at jane street (https://www.levels.fyi/companies/jane-street/salaries/softwa...), it obviously doesn't make sense.

[altdataseller]

Would accounting for years of experience control for job level somewhat? That's the only info that could be extracted

[danpalmer]

I think that would be an interesting data set to publish, but you're right it would only somewhat control for it.

[dtech]

They normalize the levels between companies on a scale they defined themselves.

[altdataseller]

Who is they?