
Headcount and turnover reporting that builds itself
Every quarter someone rebuilds the headcount spreadsheet, and the turnover number changes depending on who built it. An agent produces the report from your HR records with definitions you wrote down, and a named person reviews it before anyone else sees it.
The short answer
An agent produces headcount and turnover reports by reading your HR records on a schedule and applying definitions you wrote down: who counts as headcount, what counts as a hire, a departure, voluntary or not.
It hides any breakdown small enough to identify a person. A named person reviews each report before it goes to anyone else.
Why does the turnover number change depending on who built it?
At quarter end, someone exports the employee list, filters it, counts, and pastes the result into a slide. Next quarter someone else does it slightly differently. Contractors are in one time and out the next. A person who moved teams counts as a departure in one version. The board sees a turnover number that nobody can explain, and the next question, which team, sends someone back into the spreadsheet for a day.
Who is this for?
An owner, finance lead or people lead who produces headcount numbers for budgets, board meetings or planning and rebuilds them each time. It suits companies with an HR system or payroll system that records start dates, end dates and roles. It does not interpret why people leave. That conversation belongs to the leadership team.
What changes when the report runs from written definitions?
| Part of the report | Rebuilt each quarter | Built from written definitions |
|---|---|---|
| Definitions | In the builder's head | Written once, applied every time |
| The data | An export, filtered by hand | Read from the HR record you authorize |
| Breakdowns | Whatever someone has time to cut | By team, location or role, as you choose |
| Small groups | Shown, even when one person is identifiable | Combined or hidden below the size you set |
| Sign-off | Sent straight to the deck | Reviewed by a named person first |
How it works, step by step
The definitions are yours. The agent applies them the same way every run and shows its working.
Write the definitions down
Read the HR record
Apply the definitions and count
Protect small groups
Human gate: a named person reviews before sharing
Flag records that look wrong
Keep each period's report
What does the agent never do with workforce data?
Interpret departures
It counts. Why people left is a leadership conversation.
Expose individuals. Small groups are combined or hidden.
Change records
It flags problems for the record owner.
Share without approval
Every report waits for its reviewer.
Predict who will leave. No attrition scores, no flight-risk lists.
Where this fits
A recurring report from one data source is a single process, the shape of Tier 1, Automate a Workflow. Your definitions are what the readiness checklist will probe first. The same pattern of a report that runs itself on a schedule is described in recurring reports that spawn themselves, and a finance version is cash position reporting that builds itself. The rest of the library is at the use case library.
How does a headcount report differ from a payroll report?
| Question | Payroll report | Headcount and turnover report |
|---|---|---|
| What it answers | What did we pay, and to whom? | How many people, and how is that changing? |
| Who reads it | Finance and the payroll provider | Leadership and the board |
| Privacy risk | Individual pay | Identifying someone in a small group |
| What it depends on | Pay records | Your written definitions |
Related reading
- Recurring reports that spawn themselves: the scheduling pattern behind any regular report.
- Cash position reporting that builds itself: the finance counterpart, built the same way.
- Pipeline reporting without the Friday spreadsheet: the sales version of retiring a hand-built spreadsheet.
