How to Calculate Employee Turnover and Find What Is Driving It

For a practical internal turnover measure, divide employee separations during a defined period by average employee headcount during the same period, then multiply by 100. Keep the population, dates, and definition consistent. The percentage tells you how much movement occurred; additional analysis helps you investigate why.

The employee turnover formula

Turnover rate = separations ÷ average headcount × 100. A simple average headcount estimate is beginning headcount plus ending headcount, divided by two. For rapidly changing or seasonal workforces, average monthly or more frequent headcounts can give a more representative denominator.

Define whether separations include resignations, employer-initiated terminations, retirements, and other departures. Decide how internal transfers, contingent workers, and rehires are treated. Record those decisions so later reports do not silently change the measurement.

Worked example for a growing business

These figures are hypothetical. The result is a quarterly measure; do not label it annual turnover. Multiplying one quarter by four is an annualized projection, not the observed full-year result, and can mislead when staffing is seasonal or changing.

Separate voluntary, involuntary, and early turnover

If four of the six departures were resignations, the voluntary turnover rate in the example is 4 ÷ 42 × 100, or 9.5%. The remaining two departures represent 4.8% using the same denominator, subject to rounding. Keep the total visible so a lower resignation rate does not hide other movement.

Early turnover asks a different question. For a completed hiring cohort, divide people who left within the selected window by hires in that cohort. If three of twelve hires left within 90 days, early turnover is 25%. Hires who have not yet reached 90 days need separate treatment; counting them as retained biases the result.

Break the data into meaningful groups

• Location and shift.
• Job family or employee role.
• Tenure at separation.
• Voluntary versus employer-initiated departure.
• Hiring cohort, supervisor, and assignment where appropriate.

Show counts next to percentages. One departure from a four-person team produces 25%, which is important context before comparing that team with a much larger group. Limit access and avoid publishing small groups that reveal individual employees.

Use patterns to investigate, not to declare a cause

A concentration of departures under one supervisor is a signal for review, not proof that the supervisor caused them. Examine workload, shift assignments, pay, hiring expectations, training, employee feedback, and recent changes. Compare similar roles and periods.

Combine separation data with stay interviews, onboarding check-ins, schedule changes, and vacancy coverage. Choose one hypothesis to investigate and one improvement to test, then review both results and implementation.

Compare benchmarks only when definitions match

The Bureau of Labor Statistics publishes JOLTS separation measures using its own employment definitions and calculation methods. An internal annual turnover percentage is not automatically comparable to a published monthly or annual-average separation rate. Check population, period, and denominator before making comparisons.

Holt’s workforce analytics services can help establish definitions, dashboards, and a reporting cadence. For the financial impact of departures, read the separate employee turnover cost guide.

Validate the data before presenting the rate

Reconcile the separation list to the employee roster and confirm dates, duplicate entries, rehires, internal transfers, and changes in employing entity. Decide whether the report covers employees only or a separately defined worker population. Do not use employee headcount as the denominator for departures that also include contractors.

For a workforce that fluctuates during the quarter, compare the simple beginning-and-ending average with the average of monthly headcounts. State the method in the report and apply it consistently. More frequent observations can improve representation, but they do not correct inaccurate source records.

When presenting results, include the period, population, separation count, headcount method, rate, and a short interpretation. Set an investigation owner and review date. If a new onboarding process is introduced, track whether it was actually used before attributing a later change in turnover to that intervention.

Frequently asked questions

Can turnover exceed 100%?

Yes. A business can replace positions more than once during a period, so separations may exceed average headcount.

Is there one good turnover rate for every business?

No. Interpret the rate using role, industry, workforce size, period, separation reasons, and consistent definitions. A percentage alone does not establish whether staffing is healthy.