Employee turnover rates

One turnover number for the whole company decides nothing. It is a headline that tells you to look, and every useful conclusion comes from a cut of it: by tenure, by manager, by joining cohort, by hiring source. Below are the calculations, in the order of how much they tell you, and then the part that matters to anyone with an open role: how much of your turnover was already decided on the day an offer went out.

The calculations, from the least useful to the most

Five calculations. The first is the one everybody reports. The fourth is the one that tells you whether you have a hiring problem or a management problem, which is the only question a turnover number can usefully answer.

01
Overall turnover rate for a period

Separations during the period, divided by the average headcount during the period, multiplied by one hundred.
- Average headcount means the average across the period, not the closing number. Using the closing number understates turnover in a shrinking team and overstates it in a growing one
- Count every separation, then report the categories separately. A probation exit and a resignation at month eighteen are different events with different owners
- State the period on the number. A monthly rate and an annual rate both called turnover rate is the most common reason two people in the same meeting are looking at incompatible figures
What it decides: Very little on its own. It cannot tell you whether the number is high, because there is no external figure whose mix of function, seniority and market matches yours.
02
Annualising a monthly figure

Monthly rate multiplied by twelve gives a rolling annualised estimate, and it is unstable on small populations.
- On a team of a few dozen, a single exit moves the annualised rate further than most real trends do. Below roughly a hundred people, use a trailing twelve-month count of actual separations instead
- Do not annualise a month that contained a restructure and then compare it against an ordinary month
03
Voluntary, involuntary, and regrettable

Three buckets. The one that matters is regrettable voluntary: people you wanted to keep who chose to leave.
- Involuntary exits inside probation are a hiring signal rather than a retention signal, and they belong in a different conversation entirely
- Regrettable is a judgement, and it has to be recorded by the manager at the time of exit. Reconstructed six months later it is always flattering
- A rising overall rate with a flat regrettable rate is usually a performance process that has finally started working
04
First-ninety-day and first-year turnover, by joining cohort

Group people by the month they joined, then track what fraction of each cohort is still there at ninety days and at one year.
- A cohort that thins at ninety days points at selection or at onboarding. A cohort that thins in the second half of year one points at the job itself, or at what the job was said to be during the process
- Track it by hiring manager as well as by department, because the finding is almost always a concentration that a departmental average hides
- Track it by source: referral, inbound application, agency, internal move. The source with the highest first-year exit rate is quietly setting your real cost per hire
What it decides: Whether to change how you hire or how you manage. No other cut answers that question, and it is the only question worth putting a turnover number in a board pack for.
05
The exits your turnover rate cannot see

A candidate who accepts an offer and never joins is not a separation, so none of the arithmetic above counts them.
- They do not appear in turnover, and they do appear in your capacity plan, your recruiting spend, and a seat that is still empty
- Track offer-to-join as its own rate, by role and by month, and read it alongside turnover rather than inside it

None of the numbers on this page are benchmarks. They are calculations. Any figure you compare yours against should come with its own definition of separation attached, because the definition moves the result more than the underlying reality does.

The attrition that happens before day one

In markets with long notice periods, a substantial share of hiring effort is lost between acceptance and joining, and none of it reaches a turnover report.

“About 37% of offered candidates drop the offer, and two-thirds of them do that just before or on the date of joining.”

nasscom community

A second reading of the same market puts the dropout at one in every three offers made, against a typical rate of one in five. Whichever figure matches your market, the population is large enough to change how you plan headcount, and it is invisible to the calculation everybody runs.

Both citations, with the sources they came from, are on the proof page. If you want to know which of your live offers is exposed, the offer acceptance risk model is linked below.

What a turnover number can and cannot tell you

It cannot tell you whether yours is high: Not without your own trailing figure and your own mix. A support function and an engineering team of similar size have different natural rates for reasons that have nothing to do with how well either is run.
It can tell you where it concentrates: Nearly every actionable turnover finding is a concentration: one manager, one location, one joining cohort, one hiring source. The aggregate hides all four, which is why the aggregate is the number that gets reported and the cut is the number that gets acted on.
It cannot tell you why: Exit interviews are conducted by the organisation the person is leaving, in their last week, while they still need a reference. Weight the pattern across many of them rather than the content of any one.
It can tell you what it costs: Replacement cost, lost output during the vacancy, and the ramp of the replacement are all estimable, and the estimate is usually large enough to change a decision that a percentage never would.

Turning a rate into a number somebody will act on: the attrition cost simulator.

How much of your turnover was decided at the point of hire

A measurable fraction of it. Split your last two years of exits by tenure. Anything that left inside the first year was a selection failure, an expectation failure, or an onboarding failure, and all three sit upstream of the manager who was holding the person when they resigned.

Selection failure: the person could not do the job at the level the role required. It surfaces as a probation exit, or as a quiet underperformance that ends around month nine.
Expectation failure: the person could do the job, and it was not the job they were sold. This is the largest of the three and the most preventable, because it is manufactured during the hiring process.
Onboarding failure: the person could do the job and was never given the conditions to start. It concentrates by manager and by whether the hire was remote.
Exits well past the first year are a different conversation, and treating them as a hiring problem is how organisations end up adding interview rounds to solve a management issue.

The levers that move it, strongest first

The brief. A role defined as a list of tools attracts people who match the tools and leaves the actual job undescribed, which is how expectation failures are mass-produced at the top of the funnel.
An honest process. Everything softened during interviews to win a candidate becomes a reason to leave around month seven.
The manager. Turnover concentrates by manager more reliably than by any other cut, and it is the finding organisations are slowest to act on because acting on it is a difficult conversation rather than a policy.
The first ninety days, which is the only lever on this list that sits entirely inside one manager's control.
Compensation review timing rather than compensation level. People leave over discovering that the number moved for somebody else first considerably more often than they leave over the number.

How Continuity1 runs this funnel

The screening above is the job. These are the numbers it produces when a function owns it end to end, set against the published benchmarks for the same market.

Turnover starts at the offer rather than at the first day. These are the numbers on the front half, where a hire is either lost before it exists or arrives already fitting the role.

1 in 3: Shortlisted candidates you meet who become the hire; Aligned engagements run nearer 1 in 2, distant ones nearer 1 in 10. The market takes about 180 applicants to make one hire, and that sifting lands on your team rather than ours.; Continuity1 tracked engagements
~3: Interviews your team sits in, per hire; Ashby puts technical roles at 17.6 interviews per hire across the whole process, up 52% since 2021. The rest of that load sits with the function, not with you.; Ashby talent-trends report
1 in 9: Accepted offers that ghost before joining; Indian employers report nearly 4 in 10 offers dropped. We lose 1 in 9.; nasscom community
95%: Offers that close inside your stated band; 20 of the last 21. A flat fee earns nothing from an inflated offer; a percentage of CTC earns more.; Continuity1 tracked engagements

Every brief becomes a success profile before sourcing starts, calibrated with the people who will manage the role. That calibration is the step most hiring skips, and it is why a shortlist either matches the job or matches the job advert.

You review a scored shortlist and make the calls. The filtering never lands on your calendar.

Share a role See the workings and the citations

Questions teams ask

How do you calculate the employee turnover rate?

Separations in the period, divided by the average headcount in the period, multiplied by one hundred. The two things that most often make the result wrong are using closing headcount instead of the average, and mixing probation exits, redundancies and resignations into a single figure that then cannot be acted on.

What is a good employee turnover rate?

There is no single answer that survives contact with a specific company, because the published rates that get quoted blend industries, seniorities and definitions of separation. The comparison worth making is against your own trailing figure and against the same cut a year ago. A number without its definition attached is not a benchmark, it is a decoration.

Should probation exits count in the turnover rate?

Count them and report them separately. Folding them into the headline figure hides a hiring signal inside a retention number, and organisations then respond to a selection problem with a retention programme.

Is high turnover always a management problem?

No, and the tenure split tells you which it is. Concentrated inside the first year it is usually a hiring or onboarding problem. Spread across tenures and concentrated under particular managers it is a management problem. Spread evenly across both it is more often the market or the compensation review cycle.

How do you measure turnover on a small team?

Use counts rather than rates. On a team of a few dozen, a percentage moves so far on a single exit that it invites conclusions the sample cannot support. A trailing twelve-month count of actual separations, split by tenure, tells you the same thing without the false precision.