How the data works.
120M+ professional careers, re-checked every two weeks. Big enough to read any company that matters, honest about what it can't see.
The panel
The people who matter at the companies that matter, plus a representative sample of everyone else. It's a panel, not a census, so every count is "in sample." Nobody has a true headcount. LinkedIn's company counts are padded with fake profiles and "affiliated members," whatever those are. We have a damn good representative sample, verified, that you can read trends, mix and flows from with confidence.
Every profile is re-checked on a 14-day cycle. Low-confidence profiles are left out, so our numbers run lower than raw profile counts.
What's in it
Each person's employment history: employer, title and dates, mapped to 18 departments and 8 levels. Education, for alumni. Where they live. From that we build monthly headcount, arrivals, departures and tenure, sliced by company, department, level, school, location, and the employer before and after.
How we count
| A company | The company itself. Subsidiaries are their own companies unless you ask for the whole group (Amazon without AWS, Alphabet with everything). |
| Headcount | People with an open job there at the end of the month. Someone can hold several jobs at once, so a board member counts at each board. |
| Who counts | Everyone who lists the company as an employer, including board members, advisors, investors, contractors and AI trainers. Ambassadors, subscribers and hackathon entrants are left out: those aren't jobs. |
| Turnover | Departures over the last twelve months, divided by headcount at the end plus half the departures. |
| Executives | C-team, VP and director. |
| Small slices | Anything under 100 people gets an asterisk. One person can swing it, so read it as a hint. |
The newest months
People update their profiles after they move. About a third of a month's moves show up in the first month, and the rest fill in over the next few. Charts shade the months still filling in. Never read a dip in those months as a decline. Twelve-month figures (joined, left, turnover) use the twelve months that have settled.
What people do with it
Investors: hiring as a leading indicator, sales churn and tenure as execution risk, exits after tenders and IPOs. Recruiters: where a competitor's people come from and go, which schools feed which employers, how long people stay. Operators and boards: growth, seniority and attrition against named peers. Journalists: check a claim against who actually moved.