Enter your headcount and exits for any period. Get your attrition rate, the annualised figure, the voluntary versus involuntary split, retention, early attrition and an estimate of what the turnover cost you, benchmarked against India sector averages.
Attrition rate = exits during the period ÷ average headcount × 100, where average headcount is (start + end) ÷ 2. Multiply a monthly rate by 12 or a quarterly rate by 4 to annualise it. Example: 42 exits on an average of 259 employees is 16.2% for the year.
Spotted something incorrect? Write to [email protected] and we will fix it.
Attrition rate answers one question: out of the people you typically had on roll during a period, what share left? The formula used across HR reporting, and the one this calculator uses, is:
Attrition rate (%) = (exits during the period ÷ average headcount during the period) × 100
The denominator matters more than most people expect. Using the opening headcount alone flatters a shrinking company and punishes a growing one, so the standard approach is an average. This calculator uses the two point average by default:
Average headcount = (headcount at start + headcount at end) ÷ 2
That is fine for a month or a quarter. Over a full year, a company that hired heavily in one quarter and shed people in another can be misrepresented by two data points. If you have monthly headcount numbers, add them up, divide by the number of months, and tick the box in the calculator to use that figure instead. It is the more defensible number in a board pack.
| Step | Working | Result |
|---|---|---|
| Average headcount | (250 + 268) ÷ 2 | 259 |
| Total exits | 34 voluntary + 8 involuntary | 42 |
| Attrition rate (full year) | 42 ÷ 259 × 100 | 16.2% |
| Voluntary attrition | 34 ÷ 259 × 100 | 13.1% |
| Retention | 100 − 16.2 | 83.8% |
| Early attrition | 9 of 60 new hires left within 90 days | 15.0% |
| Estimated turnover cost | 42 × ₹9,00,000 × 75% | ₹2,83,50,000 (about ₹2.83 crore) |
A monthly or quarterly rate cannot be compared to a published annual benchmark without scaling it up. The convention is to multiply by the number of those periods in a year: a monthly rate by 12, a quarterly rate by 4, a half year rate by 2. A 2.7 percent month annualises to roughly 32 percent.
Treat that number as a run rate, not a forecast. It assumes the rest of the year looks exactly like the period you measured, which is rarely true in India where exits cluster around appraisal cycles, bonus payouts and the April to June switching season. If you have twelve months of actual data, use the real annual figure rather than an annualised month.
Splitting exits is the single most useful thing you can do with this metric, because the two point at completely different problems.
Published benchmarks are often voluntary only. India's large IT services firms, for example, report last twelve month voluntary attrition, so comparing your all in number against theirs will make you look worse than you are. Compare like with like.
This calculator reports retention on the same basis as attrition, so retention equals 100 minus your attrition rate. A stricter definition exists: the share of employees present at the start of the period who were still there at the end, which ignores anyone hired and lost within the period. The two only agree when hiring is flat, so state which definition you are using whenever you share the number.
Early attrition, sometimes called infant attrition, is the share of new hires who leave within their first 90 days:
Early attrition (%) = (new hires who left within 90 days ÷ total new hires) × 100
It is measured against new hires, not total headcount, so it sits outside the main formula. It is also the most expensive kind of turnover, because those employees never reached full productivity and the entire hiring and onboarding spend returned nothing. A high figure almost always traces back to one of three things: the role sold in the interview did not match the job, onboarding was thin, or the offer was accepted while the candidate was still shopping.
The cost estimate in this calculator multiplies your exits by the average annual CTC and a replacement cost assumption. Published research puts the cost of replacing one employee at anywhere from roughly half to twice their annual salary, and SHRM has long used a rule of thumb of six to nine months of salary. Senior, specialist and leadership roles sit at the top of that range or above it.
Those percentages bundle together recruiter and agency fees, job advertising, the interview time of everyone involved, notice period gaps, onboarding and training, the ramp up period before the replacement is fully productive, and the productivity hit on the team that covers in the meantime. It is an estimate, not an invoice, but it is usually a large enough number to make the case for fixing the underlying problem.
Industry salary and attrition surveys have reported India's overall attrition falling for three years: roughly 18.7 percent in 2023, 17.7 percent in 2024 and around 16.2 percent in 2025 across sectors. These and the sector figures below are indicative and unverified: they come from private surveys and company disclosures with different methods, and we could not link a single public primary source for each. The sector spread is wide, so the average is only a starting point.
| Sector | Indicative annual attrition | Notes |
|---|---|---|
| All India, across sectors | About 16 percent | 2025 figure, down from 17.7 percent in 2024 |
| IT services, large listed firms | About 13 to 15 percent | Last twelve month voluntary attrition at the top five firms, far below the 2022 peak of about 23 percent |
| IT and ITeS, broader industry | Around 25 percent | Smaller firms, staffing heavy and support roles run well above the listed majors |
| Banking, financial services, insurance | Around 25 percent | Frontline sales and collections roles can run far higher, in some segments above 100 percent |
| E-commerce | Around 29 percent | Driven by operations, delivery and warehouse roles |
| Retail and hospitality | 40 percent and above | Frequently exceeds 60 percent once seasonal and part time roles are counted |
| Manufacturing | About 10 to 15 percent | Fell to roughly 10.6 percent in 2024; skilled trades run hotter |
| Healthcare and life sciences | About 13 percent | Reduced to roughly 13.3 percent in 2024 from about 18 percent in 2023 |
There is no universal safe threshold. A useful way to read your own figure:
The most actionable cuts are by tenure (are you losing people in year one or year five?), by manager, by function, and by performance rating. Losing your lowest performers at a high rate is a different story from losing your highest.
In everyday HR reporting in India the two terms are used interchangeably and calculated with the same formula. Where a distinction is drawn, attrition refers to positions that are not refilled, while turnover refers to employees who leave and are replaced. Because most HR systems do not separate the two cleanly, treat any published figure as the same metric unless the source says otherwise.
Average headcount. Opening headcount distorts the rate whenever the workforce changes size during the period. The average of the opening and closing figures is the accepted default, and averaging month by month headcount across the full period is more accurate still.
You can, and it produces a run rate: what the annual figure would be if every month looked like the one measured. It ignores seasonality, which is significant in India where exits spike after appraisals and bonus payouts. Use the annualised figure to compare a short period against an annual benchmark, and use twelve months of real data whenever available.
That is a policy decision to make once and apply consistently. Most companies calculate attrition on employees on their own payroll and exclude interns, third party contract staff and consultants. Whatever you decide, use the same population in the numerator and the denominator and note the choice alongside the figure.
Yes. Any separation from the payroll counts in the total attrition figure. Retirements and deaths are normally grouped under voluntary attrition in survey methodology, though many companies report them separately as unavoidable attrition so the voluntary number reflects only preventable exits.
Startups usually run above the commonly reported all India average of roughly 16 percent, and early stage companies often sit in the 20 to 30 percent range without anything being structurally wrong. The pattern matters more than the headline: high attrition in the first 90 days points at hiring and onboarding, while exits between 18 and 36 months point at career path and pay bands.
It multiplies the number of exits by the average annual CTC and a replacement cost percentage. Published research puts replacement cost at roughly half to twice annual salary depending on seniority, and SHRM has used six to nine months of salary as a working figure. The estimate covers recruitment spend, interview time, onboarding, ramp up and cover load.
Most likely because the metrics differ. Large Indian IT services firms report last twelve month voluntary attrition, excluding terminations and layoffs, across a long tenure workforce. If your figure includes involuntary exits or contract staff it is not comparable. Compare your voluntary rate instead.
Yes, whenever the two are calculated on different bases. Reported as 100 minus attrition against average headcount, they always sum to 100. Under the stricter definition, where retention counts only employees present at both the start and end of the period, the two will not sum to 100 in any period with hiring.
A large share of turnover is decided before day one, in how roles are screened and how honestly the job is described. GetHirePlus automates resume screening, assessments and first round interviews so your team spends its time on the candidates who will actually stay.
Start hiring with GetHirePlus, freeSee how candidate assessments help you check role fit before the offer.
AI hiring software for resume screening, AI phone interviews, assessments, live interviews and applicant tracking.
Get started