I made a list of 30k LinkedIn profiles listed as current/former employees of 50 companies.

I assume 20-30% are fake or misrepresent experience. 

I assume 5-10% of current/former employees are not represented. 

Can you verify legitimacy and capture the market @ scale?

LinkedIn owns all the data but it’s filled with noise. Payroll company data would be ideal but impossible to retrieve. Gathering data through interviews is a slog and information is gatekept. Looking for some inspiration…

Estimates are conservative- I started with a small sample of ten companies, gathering data manually- error rate was north of 60%.

This is a good idea. If I can run something to scrape for badges at least the manual process of verifying thereafter is a more curated list

LinkedIn has also introduced a concept of verified profile badges wherein you verify with to the official mail id of the organisation you work at. Maybe that’s one good filter