How do I calculate employee turnover and retention?

Calculate employee turnover for a defined period and workforce by dividing the included departures by a clearly stated employment denominator, then multiplying by 100. Reconcile hires and departures to the ending headcount. Keep that turnover measure separate from headcount growth and retention of the people employed at the start.

A company can grow while losing people. The recruitment team may have filled vacancies and added new roles at the same time. Looking only at the larger ending headcount can make that movement disappear just when the owner needs to understand it.

Define the workforce and period

State the dates, company or location, employees included and what counts as a departure. Decide how transfers, retirements, temporary roles and other relevant events are handled. Use the actual records and appropriate worker classifications.

Keep employees and outside providers in their correct categories. A contractor engagement ending should not automatically enter a payroll employee-turnover calculation. Likewise, a move between departments can be an internal movement without being an exit from the whole company.

The BLS JOLTS technical note defines separations, distinguishes quits, layoffs and discharges, and other separations, and explains its rate calculations. Those survey definitions are useful references. A company's quarterly calculation should not be presented as identical to a national monthly or annual-average series using different conventions.

Write your own report's definition beside the number. If it changes, recalculate comparable prior periods where the records allow and identify the break where they do not. A cleaner new definition does not make the old figures automatically comparable.

Reconcile the people movement

Consider a wholly fictional company with twenty payroll employees at the start of a quarter. During the period it records ten hires and six departures, all distinct separation events in this example. There are no acquisitions, boundary changes, internal transfers or other headcount adjustments.

Ending headcount is twenty-four: twenty plus ten less six. The net increase is four people, or 20% of the starting twenty. That growth calculation does not establish that only four people joined or that nobody left.

The supplied monthly record shows when the movement occurred. Eight hires join in month one. Six departures occur in month two. Two more hires join in month three. Each person's actual start and departure evidence supports the relevant event inside the fiction.

Scroll horizontally if needed. Headcount movement and a turnover-rate denominator answer different questions.

If the movement does not reconcile, investigate missing or duplicated starts and departures, dates and reporting boundaries. Do not add a vague adjustment simply to reach the ending total.

Use the event date defined by the report, with supporting employment records. A notice received before quarter end may describe a departure taking effect afterward. Likewise, a recruitment offer is not necessarily a completed hire. Record the planned event separately until the chosen definition supports counting it. A delayed payroll-system update should not silently move a real departure into another period.

Preserve a private person and employment-spell reference behind each aggregate event. If someone leaves, returns and leaves again, decide whether the report counts separation events or unique people and label the result. An event rate can exceed a simple share of distinct people without establishing that more than every employee left once. The present fiction has six distinct events and no rehiring; do not import that simplifying condition into actual records that differ.

Name the denominator

For the first calculation, this fictional report uses the mean of opening and closing headcount: (20 + 24) ÷ 2 = 22. Quarterly total turnover under that stated convention is 6 ÷ 22 × 100, or 27.27%.

The twenty-two is a simple endpoint-average proxy. It is not a measured daily average, and the company reached twenty-eight during the quarter. State that limitation rather than describing the denominator as exact employment exposure.

Using the four equally weighted snapshots of twenty at opening and twenty-eight, twenty-two and twenty-four at month ends gives 23.5. Dividing six departures by that different denominator gives 25.53%.

The people have not changed between those two calculations. The averaging convention has. Neither snapshot method establishes the time-weighted average without the relevant dates. Choose a defensible, consistently applied method suited to the report rather than choosing whichever makes the rate look lower.

Headcount and full-time-equivalent measures are also different units. If a separate report supplies eighteen average full-time equivalents, putting six employee departures over eighteen produces 33.33%. That mixes a people numerator with a capacity-weighted denominator; it is not the same headcount rate as 27.27%.

Separate the departure types

Assume the six departures consist of four employee resignations and two employer-initiated terminations, with no retirements or other departure category. Under the endpoint-average denominator, the resignation rate is 4 ÷ 22 × 100, or 18.18%. The employer-initiated rate is 2 ÷ 22 × 100, or 9.09%.

Using unrounded figures, those components add to the 27.27% total. Keep the denominator, period and category definitions the same. Do not compare a monthly resignation rate with a quarterly all-departures rate as if the difference measures management performance.

The categories describe events. They do not establish why employees resigned, whether a termination was appropriate or whether someone was satisfied with the job. Those questions need their own evidence and fair process.

For small teams, show the actual count alongside the percentage. One departure from a small denominator can move the rate sharply. Avoid publishing identifiable personal details or turning an aggregate report into a label attached to one person.

Calculate starting-cohort retention separately

Of the six fictional leavers, five belonged to the starting twenty. The sixth joined in month one and left in month two. Fifteen original employees therefore remain at quarter end.

Starting-cohort retention is 15 ÷ 20 × 100 = 75%. It is not 100% less the 27.27% turnover measure. The cohort uses the original twenty as its group; total departures include a newer employee, and the turnover denominator uses a different rule.

The closing twenty-four consists of fifteen original employees and nine of the ten new hires. That reconciles the cohort view with the total workforce. A larger company can have 20% headcount growth and 75% starting-cohort retention in the same period.

The customer-retention guide illustrates why customer count and recurring revenue also need separate definitions. Here the subject is employees and employment events, not customer subscriptions or employee value measured by revenue.

Do not turn one quarter into a forecast

Multiplying 27.27% by four creates an annualized arithmetic figure under an assumption that the same pattern repeats. It is not the next year's observed turnover or a prediction that particular employees will leave.

Use the full actual year when reporting actual annual movement. If you compare with a benchmark, examine its workforce, event definitions, denominator and rate convention. BLS's monthly and annual-average series have specified methods; do not copy a national percentage into a company dashboard without understanding the difference.

A seasonal operation may have expected departures that differ from an ongoing specialist team. That context can matter without making every departure harmless or every low rate desirable. A single target cannot replace understanding the work and the people involved.

Cost the response you actually propose

Suppose the fictional owner identifies four specific replacement roles to recruit and train. This is a new proposed action, not a multiplication of every departure by an industry average cost. Other hiring and continuing operations have separate funded plans.

For each of those four roles, the complete supplied recruitment-and-start scope includes $1,500 sourcing charges, six paid interview-team hours at an all-in $50, twelve paid new-worker training hours at $20, four paid trainer hours at $40, and $200 of required setup.

The per-role total is $2,400: $1,500 + $300 + $240 + $160 + $200. Four roles need $9,600. A separately available $10,000 allocation covers that defined work and leaves $400.

The interview team has the required twenty-four hours available, the trainers have sixteen, and the new workers receive forty-eight paid training hours in total. Those are distinct resources; the same hour is not allocated to two simultaneous tasks.

This is not the full economic cost of turnover. Vacancy effects, separation obligations, continuing wages, tax and other business costs are outside the defined recruitment-and-start comparison. Establish actual applicable requirements separately. The existing hiring-affordability guide follows the broader funded employment commitment.

Investigate before choosing a fix

Use appropriate records and respectful conversations to understand patterns in roles, tenure, scheduling, supervision and working conditions. Keep an employee's account distinct from a conclusion about cause. Similar departure dates do not prove the same reason.

If most movement concerns people leaving soon after joining, inspect the actual job description, hiring expectations, onboarding and support. If a particular schedule is difficult to sustain, examine the schedule rather than assuming that a generic bonus addresses it.

Choose a bounded change with a clear purpose and review point. Preserve what else changes during the period so a later reduction is not automatically attributed to the intervention. A rate can improve because hiring slows or the denominator changes while a genuine work problem remains.

The growth-measure guide keeps the operating question beside the count. Apply the same care here: which work needs attention, and what evidence would show that the arrangement became better?

Keep the turnover record

1. Define the workforce, reporting boundary and dates.
2. Reconcile opening employees, hires, departures and closing employees.
3. State departure categories and the employment denominator.
4. Show counts alongside rates and keep calculation precision until display.
5. Calculate starting-cohort retention as its own measure.
6. Check comparability before benchmarking or annualizing.
7. Investigate the actual work and fund any proposed response.