Turnover rate calculator: attrition rate, employee turnover rate and retention rate formula

This is the part almost every turnover guide skips, and it is why two competent analysts hand a board two different numbers for the same year. There is no single official US turnover rate definition. Each measure below is legitimate, each is computed from the identical twelve separations and the identical headcount, and none of them agree.

Read the first and last rows together. Turnover of 23.1 percent implies retention of 76.9 percent if you subtract, but the measured retention rate is 82.0 percent. Both numbers are right. Section 03 explains why, and it is not a rounding artifact.

The BLS row deserves attention because it is the row you will be benchmarked against, whether or not you chose it. BLS states the method plainly in its technical note: "The separations rate is computed by dividing the number of separations by employment and multiplying that quotient by 100." Employment, in that sentence, means "persons on the payroll who worked or received pay for the pay period that includes the 12th day of the reference month." A single day, not an average.

There is a sharper version of the point inside the same BLS release. The job openings rate in that publication is "computed by dividing the number of job openings by the sum of employment and job openings, that is, all filled and unfilled jobs," while the separations rate divides by employment alone. Two rates, published side by side in the same table set, on two different denominators. If the federal statistical agency needs two denominators to describe one labor market, your internal definition needs to be written down rather than assumed.

There is no single average US turnover rate, and any page that gives you one number for the whole country is quietly averaging a restaurant and a bank. BLS publishes a quits rate every month by industry through the Job Openings and Labor Turnover Survey, and quits are the closest published equivalent to what most employers mean by voluntary turnover.

The figures below are the seasonally adjusted quits rates for July 2026, released September 1, 2026 as USDL-26-1432. BLS publishes them as monthly rates. The annualized column is our own arithmetic, the monthly rate multiplied by twelve, so that the numbers sit on the same scale as the annual rate your HR system reports. BLS does not publish that column and it is an approximation that assumes stable employment through the year.

Treat any single month as provisional. JOLTS revises the two prior months with every release, and the revisions are not small: the September 1 release restated June 2026 job openings down by 177,000 and June quits down by 19,000. If you are benchmarking against a figure you copied down six months ago, it has almost certainly moved since. Pull the current month rather than the one you remember.

The spread is the finding. Accommodation and food services runs roughly three times the quits rate of finance and insurance, so a 30 percent annual voluntary rate is comfortably good in one and a crisis in the other. Benchmark against your own row, and then against the roles inside it, because a single company rate averages a warehouse floor and a controller.

At the whole economy level, July 2026 ran at 7.3 million job openings, 5.1 million hires and 5.1 million total separations, with quits at 3.1 million and layoffs and discharges at 1.7 million. The total separations rate was 3.2 percent for the month and the quits rate 1.9 percent. Roughly 38 percent and 23 percent annualized, and the distance between those two is the share of departures an employer did not choose to prevent.

JOLTS is not the only free federal wage and workforce dataset worth wiring into a comp process. The occupational wage estimates behind BLS salary data come from a different survey with different rules, and they are what you price a leaver's replacement against once you know how many you are replacing.

You have seen the figures: turnover costs six to nine months of salary, or a third of annual pay, or twice salary for a senior hire. Here is the honest position. No US federal agency publishes a cost of turnover statistic. BLS measures how many people leave, not what their leaving cost. Those percentages come from private surveys with their own samples and their own definitions of cost, and they are estimates rather than official data.

That does not make turnover cost unmeasurable, it makes it yours to measure. The calculator above takes a percentage because the only defensible percentage is the one you derive from your own books. Four line items get you most of the way, and every one of them is already somewhere in your general ledger or your ATS.

Run those four against the nine voluntary departures in the worked example. At $65,000 average salary and a 33 percent assumption, the year cost $193,050. Now compare it to the alternative. A 4 percent correction applied to every one of the 52 employees costs $135,200 a year, and it is spread across the whole team rather than spent on replacing a third of it. That comparison, not the turnover rate itself, is the number a CFO will act on, and it is the arithmetic behind any salary increase you are modeling.

The comparison only holds if the raise is aimed correctly, which is why turnover analysis without market data usually produces an expensive across the board increase that fixes nothing. Splitting a retention budget between a general increase and targeted market adjustments is the decision covered in merit increases versus cost of living raises.