Staff turnover: the figure that means nothing on its own

Staff turnover is leavers over the period divided by average headcount, expressed as a percentage. The arithmetic takes a minute and the interpretation takes judgement, because the same figure can describe a healthy business replacing a few people and a department quietly collapsing. A single number is almost always the wrong way to look at it.

How to calculate staff turnover

• Leavers in the period divided by average headcount in the same period, not by the headcount on one date.
• Average headcount as the mean of the start and end figures, or monthly if the year was volatile.
• Counted consistently on contractors and fixed-term staff — include or exclude them, but do the same thing every period.
• Separated into voluntary and involuntary, which is the split that carries almost all the meaning.
• Broken down by team, tenure and manager wherever numbers allow, because the aggregate hides where it is actually happening.

Why the aggregate misleads

Twenty per cent turnover spread evenly across a stable company is unremarkable. The same twenty per cent concentrated in one team, or entirely among people in their first year, is a specific and fixable problem — and the aggregate figure reports both identically. The most informative cut is tenure: people leaving within six months indicate a hiring or induction failure, while people leaving at three years indicate a progression problem. Those two require completely different responses, and the headline percentage cannot tell them apart.

Some turnover is good and a target of zero is a mistake. Nobody leaving at all usually means nobody can afford to, which is a different problem wearing a flattering number. The useful question is not how much but who — losing the people you would have promoted matters more than the percentage.

What replacing somebody costs

1. Recruitment cost: advertising, agency fees if used, and the hours of everybody involved in screening and interviewing.
2. Vacancy cost: the work not done, or done by colleagues who were already busy, between the leaving date and the start date.
3. Ramp-up: the months before the replacement is as productive as the person who left, which is the largest component and the one nobody counts.
4. Lost knowledge, particularly of customers and of the undocumented way things actually work.
5. The knock-on risk, since resignations cluster — one departure in a small team makes the next one more likely.

Reading it honestly

Compare it to your own previous periods rather than to a published benchmark, because benchmarks mix sectors, regions and definitions of who counts as a leaver. Look at the trend and the composition together: a falling percentage made up entirely of people you wanted to keep is worse than a steady one made up of planned changes. And where the figure prompts action, the action belongs with the managers of the teams concerned, not with a company-wide initiative aimed at an average nobody experiences.

Ettex Sheets is enough to hold the calculation — leavers, average headcount, the voluntary split and a tenure band per leaver — recalculated each quarter rather than once a year. What people said on the way out belongs with the exit interview, and the two are only useful read together.