In this guide
Attrition rate = (exits during the period ÷ average headcount during the period) × 100. Average headcount is the headcount at the start plus the headcount at the end, divided by two.
Nine people left a team that stayed at 64 all year. 9 ÷ 64 × 100 = 14.1%. There is an attrition calculator further down, and then the part nobody puts on these pages: what the number does not tell you.
The formula, in one line
Every version of the attrition rate formula is the same division with different things on the bottom. Exits go on top. Some measure of how many people you had goes underneath. Multiply by a hundred.
The disagreement is entirely about the denominator, and it is a real disagreement rather than a pedantic one, because on a team that grew by half during the year the three versions below will give you three different answers.
Worked through with real numbers
Take a company that started the financial year with 64 people, hired 14, lost 9, and ended with 69.
| Method | Working | Result |
|---|---|---|
| Average headcount | 9 ÷ ((64 + 69) ÷ 2) × 100 | 13.5% |
| Starting headcount | 9 ÷ 64 × 100 | 14.1% |
| Ending headcount | 9 ÷ 69 × 100 | 13.0% |
A spread of just over one percentage point on the same nine people. On a team of 64 that is noise. On a team of 600 in a year when you doubled headcount, the same three methods can be four or five points apart, which is enough to make a board slide say something the business did not do.
Use the average-headcount version. It is the one most HR teams and most reporting tools use, and it is the only one of the three that does not quietly reward you for hiring. Then write down which one you chose, because the person who inherits the spreadsheet will not be able to tell from the number alone.
Calculate yours
Put your own figures in. It uses the average-headcount method, and annualises anything shorter than a year.
Work out your attrition rate
- Attrition rate over 12 months
- 14.1%
- 9 exits against an average headcount of 64.
- Annualised
- 14.1%
- Same figure, since the period is already a year.
- Retention rate over the same period
- 85.9%
- 55 of the 64 people who were here at the start are still here.
What that means for this team. At this rate you replace the equivalent of the whole team roughly every 7.1 years, with a resignation about every 6 weeks. Every one of those is a notice period, a handover and a hiring loop.
Why the denominator is an average
Picture two companies that each lost 10 people last year. One held steady at 100. The other started at 50 and finished at 150 after a funding round.
Measured against starting headcount, the second company looks twice as bad: 20% against 10%. But it did not have 50 people exposed to the risk of leaving for most of the year. It had something closer to 100. Averaging the two ends is a rough fix, and it is rough. If your headcount moved in steps rather than smoothly, average the month-end figures instead. It takes ten minutes once and then it is just a column.
Monthly, annualised and LTM
Three numbers, three jobs, and they are not interchangeable.
- Monthly. Exits in one month over that month’s average headcount. Good for spotting a bad month quickly. Terrible on its own, because one person leaving a team of 20 is 5%, and the month they do not is 0%.
- Annualised. The monthly rate multiplied by 12, or a quarterly rate multiplied by 4. It answers “if this carried on, where would we land?” It is a projection, not a measurement, and it should be labelled as one.
- LTM, or last twelve months. Actual exits over the twelve months just gone. This is what listed companies report, and it is the most honest of the three, because nothing in it is extrapolated. It is also the slowest to move, which is the point: it will not panic you over one resignation.
If you report one number to your board, report LTM. If you watch one number yourself, watch the monthly figure for the shape rather than the level.
The splits worth reporting separately
A single company-wide attrition rate is almost always the least useful way to look at it. Four splits earn their place.
- Voluntary and involuntary. People who resigned, and people you let go. Lumping them together produces a number you cannot act on, because half of it was your own decision.
- Regretted and non-regretted. Of the people who resigned, which ones did you want to keep? This is a judgement call and it should be made by the manager, in writing, within a week of the resignation, not reconstructed at the end of the quarter when everyone remembers it differently.
- By tenure. People leaving inside their first 90 days are telling you about hiring or onboarding. People leaving at eighteen months are telling you about something else entirely.
- By team. A company at 13% with one department at 31% does not have an attrition problem. It has a department problem, and the company-wide number is hiding it.
Retention rate is not 100 minus this
The calculator shows a retention rate next to the attrition rate, and on most teams the two will not add up to a hundred. That is not a bug, and the gap is worth understanding before you put either number in a board pack.
Retention rate = (people from the starting headcount who are still here ÷ starting headcount) × 100.
The difference is who gets counted. Attrition divides exits by an average headcount, which includes everyone you hired during the period. Retention looks only at the people who were there on day one and asks how many survived. Someone you hired in March and lost in August appears in your attrition rate and is invisible to your retention rate.
Take the same company from earlier: 64 at the start, 14 hires, 9 exits, 69 at the end. Attrition is 13.5%. If all nine leavers came from the original 64, retention is 55 ÷ 64, or 85.9%. Those do not sum to 100, and they should not.
On a team that held steady all year the two track each other closely. On one that grew by half, they can sit several points apart, and the gap widens the faster you hire. If you report both, say which is which.
Attrition rate or turnover rate?
Most Indian HR teams use the two words for the same calculation, and you will not confuse anyone by doing the same. Where a distinction is drawn, attrition means the seat stays empty and turnover means you backfill it.
It is a distinction worth keeping in one place only: your hiring plan. Nine exits that you replace costs you nine hiring loops. Nine exits you choose not to replace is a restructure. The attrition rate is identical. What happens next is not. There is more on the vocabulary in what attrition means in HR.
What the number cannot tell you
Here is the uncomfortable part, and the reason this page exists on a wellbeing company’s website rather than a payroll vendor’s.
An attrition rate is a count of people who have already gone. By the time it moves, the decision was made months ago, the notice period has been served, and the handover document is written. You are reading a receipt.
Gallup put numbers on how early that decision happens, in research published in March 2019. 52% of people who left voluntarily said their manager or organisation could have done something to keep them. 51% said that in the three months before they left, nobody had talked to them about how they were doing or where they were going.
Three months of a conversation that did not happen. That is the window the attrition rate cannot see into, and it is the only window in which anything could have changed.
So calculate the rate. You need it, and you should split it four ways and watch it quarterly. Then go and find a measure that moves before people resign rather than after. We wrote about how that works in burnout as the leading indicator, and about what to do in the week you spot something in the signs someone is about to quit.
Questions people ask
What is the attrition rate formula?
Attrition rate = (number of people who left during the period ÷ average headcount during the period) × 100. Average headcount is usually taken as the headcount at the start plus the headcount at the end, divided by two. A team that began the year with 64 people, ended with 64, and saw 9 people leave has an average headcount of 64 and an attrition rate of 14.1%.
How do you calculate attrition rate for a month?
Use the same formula over a one-month window: exits in the month divided by the average headcount for that month, times 100. To compare it with an annual figure, multiply by 12. Be careful doing that with a small team. One person leaving out of 20 in a single month annualises to 60%, which says more about the arithmetic than about the company.
What is the attrition formula used in BPO?
The same one. BPO and contact-centre teams usually report it monthly and then annualise, because hiring and exits both move quickly, and they often split it by tenure so that people leaving in the first 90 days are counted separately. Some report against starting headcount rather than average headcount, which gives a slightly higher number on a team that is growing; whichever you pick, use it consistently.
What is the employee retention rate formula?
Retention rate = (employees present at the end who were also present at the start ÷ employees at the start) × 100. A team that began the year with 64 people and still has 55 of those same people has a retention rate of 85.9%. Note that it deliberately ignores anyone hired during the period, which is exactly why it is not the same as 100 minus your attrition rate.
Is retention rate just 100 minus attrition rate?
No, and treating it as one is the most common mistake in retention reporting. Attrition divides exits by an average headcount that includes people hired during the period. Retention divides survivors by the starting headcount only. On a team that did not change size the two are close; on a team that grew 50% during the year they can be several points apart, and the gap widens the faster you hire.
Is attrition rate the same as turnover rate?
In practice, most Indian HR teams use them interchangeably and calculate them the same way. Where people do draw a line, attrition means the role is not being backfilled and turnover means it is. The arithmetic does not change. What matters far more is whether you are separating voluntary from involuntary exits, because mixing them hides the number you can actually do something about.
What is a good attrition rate?
There is no honest universal answer, and anyone quoting one without naming the sector, the company size and the year is guessing. A contact centre and a product engineering team are not comparable. The useful question is not whether your number beats an average but whether it is moving, which exits were regretted, and how many of them you could see coming.