Employee Turnover

How to calculate turnover rate

The math is simple. The arguments are about what goes into the numerator and the denominator, so write your definitions down before you publish the first number.

Count a separation when someone leaves payroll. Include resignations, involuntary terminations, layoffs, retirements, deaths and job abandonment. Leave out internal moves, people on leave of absence and contractors you pay through a staffing firm, unless you track them in a separate contingent-workforce rate.

For the denominator, use average headcount over the period, not the year-end number. A site that grows from 300 to 500 people will look far better or worse than reality if you divide by only one of those figures. When hiring is seasonal, average the month-end counts instead of using start plus end divided by two.

Then split the total. At minimum, report:

• Voluntary turnover: the employee chose to leave. See voluntary turnover for how to classify edge cases like job abandonment.
• Involuntary turnover: you ended the employment, for performance, conduct or a reduction in force.
• First-90-day turnover: exits within three months of hire, which usually point to a hiring or onboarding problem rather than a management one. Many teams track this as new hire turnover.

A worked example from a multi-site retailer

A regional grocery chain with 14 stores averages 1,250 hourly employees across the year and records 812 separations. Overall turnover is 812 ÷ 1,250 = 65%. On its own that number starts an argument and ends nowhere.

The HR manager, Dana, breaks it down. 590 exits were voluntary (47%), 222 involuntary (18%). Of the voluntary leavers, 310 left within their first 90 days. Two stores account for 40% of all early exits, and both changed store managers in the spring. Night stockers turn over at roughly twice the rate of cashiers.

Now there are three specific problems to work on: early attrition in two stores, a manager transition that went badly, and a night-shift role that may need a pay differential or a more honest job preview. A single 65% figure would have produced a generic engagement survey instead.

Typical turnover ranges

Benchmarks vary so much by industry, pay level and labor market that a single national average is close to useless. Practitioners see roughly these annual ranges:

• Restaurants, hospitality and retail hourly roles: often 60% to well over 100%.
• Warehouse, logistics and light industrial: commonly 40% to 80%, higher at sites that rely on seasonal peaks.
• Contact centers: typically 30% to 60% for in-house centers, often higher in outsourced and offshore operations. See call center attrition for how that industry measures it.
• Salaried professional and corporate roles: usually 10% to 20%.

In the US, the Bureau of Labor Statistics publishes monthly separation rates by industry in its JOLTS survey, which is a reasonable public reference. Your own trend line by role and site is still more useful than any external figure.

What turnover really costs

Every exit sets off a chain: a vacancy, overtime or missed output while the seat is empty, recruiting spend, training time, and a new person who is less productive for weeks or months. For hourly roles the direct replacement cost is often a few thousand dollars per exit. For skilled and salaried roles it commonly runs from half to more than a full year of pay once lost productivity is counted. The cost of turnover page walks through how to build that estimate for your own roles.

The hidden cost in high-volume operations is hiring capacity. A 600-person site with 70% turnover has to hire about 420 people a year just to stand still. Cut turnover to 50% and that drops to 300 hires, which is 120 fewer requisitions, interviews, background checks and training seats.

Turnover vs attrition

People use these words interchangeably, and in many contact centers and BPOs attrition simply means agents who left. In stricter HR usage, turnover counts every departure, whether or not the role is refilled, while attrition refers to departures that reduce headcount because the seat is not backfilled, such as retirements in a hiring freeze. The attrition rate page covers that distinction in detail. Whichever convention you use, state it in the footnote of every report, because finance and operations often assume different definitions.

Common mistakes

• Reporting one blended number. Voluntary, involuntary and early turnover have different causes and different owners. Blend them and nobody owns the fix.
• Counting transfers as exits. A promotion into another department is a retention win, not a separation.
• Dividing by year-end headcount at a growing site. The rate will look lower than it really is.
• Ignoring who left. Losing the bottom 10% of performers is not the same as losing your best trainers. Track regrettable and non-regrettable exits separately.
• Trusting exit reasons at face value. Better pay is the most common stated reason because it is the most polite one. Compare it with what stay interviews and manager-level data tell you.

How to reduce employee turnover

Start where the data points. If most exits happen in the first 90 days, fix the front end: realistic job previews, clear schedules before the offer, a structured first two weeks and a check-in at day 7 and day 30. If turnover clusters under certain supervisors, that is a management selection and coaching issue. If it follows a pay cutoff, benchmark your rate against the employers people leave for.

Schedules matter more than most leaders expect for hourly staff. Predictable shifts, published two weeks out, and an easy way to swap shifts often move turnover more than a small raise.

Finally, ask people who stay why they stay. Use our stay interview template with tenured employees each quarter, and track monthly exits in an attrition tracker so you see a problem in month two instead of at the annual review.