Turnover Rate
Turnover rate is the share of your workforce that left during a period. Divide the number of separations by the average headcount for the same period, then multiply by 100.
The formula is trivial. What matters is that the headline figure combines events with almost nothing in common, and separating them is where the number becomes worth calculating.
The formula
Turnover rate = (Separations ÷ Average headcount) × 100
Average headcount = (Headcount at start + Headcount at end) ÷ 2
Annualized from a month: = monthly rate × 12
Worked: a team starts the year with 180 people, ends with 200, and 34 people left. Average headcount is 190, so turnover is 34 ÷ 190 = 17.9 percent.
Two conventions matter and are often applied inconsistently. Use average headcount rather than the start or end figure, or a growing company will understate and a shrinking one overstate. And decide whether internal transfers count as separations: usually they should not, since the person has not left the organization.
The two splits that make it useful
A single turnover figure combines a retirement, a dismissal, a relocation and a good engineer leaving for a competitor. Those are not the same event and should not share a number.
| Split | Categories | What it tells you |
|---|---|---|
| Voluntary vs involuntary | Resignations vs dismissals and redundancies | Whether people are leaving or being asked to |
| Regretted vs non-regretted | Would you rehire them? | Whether the turnover you have is costing you anything |
Regretted voluntary turnover is the figure worth reporting. A company at 18 percent overall where almost none of it is regretted is in a different position from one at 12 percent that is losing the people it most wanted to keep, and the headline number rates the second as healthier.
Two further cuts are worth running: turnover in the first twelve months, which measures hiring and onboarding rather than retention, and turnover by manager, which is usually the single most concentrated pattern in the data.
Why benchmarks are mostly useless
Turnover varies so widely by industry, role and labor market that a cross-industry benchmark carries almost no information. Hospitality and retail run structurally high; specialized professional roles run low. Comparing your figure to a national average tells you which industry you are in.
Two comparisons do work. Your own trend over several periods, with headcount growth noted alongside it. And the same figure between teams in your own organization, where the industry, pay bands and labor market are held constant and the differences are therefore about something you can act on.
Context matters too. Gallup reported in 2026 that global engagement declined for a second consecutive year, with the sharpest fall among managers, and in a market where people are reluctant to move, low turnover can indicate people feeling stuck rather than settled. That pattern has its own name: staying while disengaging.
What to do with the number
Turnover is a lagging indicator. By the time it moves, the causes are months old, which is why acting on the rate itself rarely works.
- Look at first-year turnover separately. If it is high, the problem is selection, expectation-setting or onboarding, not retention.
- Look by manager before looking by policy. Concentration under one manager is the most common and most actionable pattern.
- Ask leavers what would have changed their mind, which produces better answers than asking why they left.
- Watch the leading signals instead, which our guide to retention strategies covers.
Key points
- Turnover rate is separations divided by average headcount, times 100.
- Use average headcount, or a growing company understates and a shrinking one overstates.
- The headline figure combines a retirement, a dismissal and a valued engineer resigning. Those are not the same event.
- Regretted voluntary turnover is the number worth reporting.
- First-year turnover measures hiring and onboarding rather than retention, and belongs in its own line.
- Turnover by manager is usually the most concentrated pattern in the data.
- Cross-industry benchmarks tell you which industry you are in. Compare to your own trend, or between your own teams.
Frequently asked questions
How do you calculate turnover rate?
Divide the number of employees who left during the period by the average headcount for that period, then multiply by 100. Average headcount is the starting headcount plus the ending headcount divided by two.
What is a good employee turnover rate?
There is no cross-industry figure worth quoting, because turnover varies enormously by industry, role and labor market. The useful comparisons are your own trend over several periods and the differences between teams within your own organization, where industry and pay bands are held constant.
What is the difference between turnover and attrition?
They are often used interchangeably. Where a distinction is drawn, turnover covers all departures including those you replace, while attrition refers to positions left unfilled, so headcount shrinks. Because usage varies, state which you mean rather than assuming the reader shares your convention.
What is regretted turnover?
Departures of people you would have rehired. It is the most useful single turnover figure, because a high overall rate that is almost entirely non-regretted is a very different situation from a lower rate concentrated among the people you most wanted to keep.
Should internal transfers count as turnover?
Generally no, since the person has not left the organization. What matters is applying the convention consistently, because switching it between periods creates a change in the number that looks like a change in behavior.
Why is first-year turnover reported separately?
Because it measures something different. People leaving within twelve months usually reflects selection, expectation-setting or onboarding rather than long-run retention, and mixing it into the headline figure obscures both.
Is low turnover always good?
No. In a labor market where people are reluctant to move, a low rate can mean people feel stuck rather than settled, and disengagement replaces resignation. That pattern is sometimes called quiet cracking, and it costs the same as turnover while showing up in none of the same numbers.