Turnover Cost Calculator

The Six Bills Inside One Departure

A departure is not one invoice. It is six costs that land on different budgets: two on the hiring budget, one on the department's output, one on training, one on the new hire's slow start, and one on payroll at the exit. Most managers see only the first and the last.

Cost per departure = recruiting spend + HR time + vacancy loss + training + ramp loss + exit cost

Vacancy loss = daily salary x vacancy days x coverage gap

Ramp loss = monthly salary x ramp months x productivity gap

Two of those lines are cash you can find in an expense report. The other two are output that never got produced, and on most jobs they are the bigger half. The calculator above prices all six from your own numbers, then multiplies by the departures you had in the last 12 months.

Two rules keep the total honest. Count only what changes because the person left. And put a year on every rate you borrow from a study, because a 2019 recruiting cost applied to a 2026 hire is a guess in a suit.

Lines One and Two: Recruiting Spend and HR Time

Start with the money you can trace. SHRM's 2022 benchmarking report, built from data collected between April and November 2021, put the average cost per hire at $4,683 across 472 organizations. The median in the same table was $1,244. That spread is the whole story: a few executive searches average $28,329 each and drag the mean up, so a warehouse or front desk hire almost never costs the average.

Build the line from your own receipts instead. A job board post at $399, a referral bonus at $500, and a background check at $65 comes to $964 (399 + 500 + 65 = 964). Add an agency fee only if you used one, and remember that a 20 percent fee on a $60,000 role is $12,000 before the person has answered a single email.

HR time is the line that hides inside salary. A requisition that eats 22 hours of recruiter and coordinator time at a loaded $34 an hour costs $748 (22 x 34 = 748). Count the interview panel too: four people at one hour each, at a loaded cost of $46 an hour, is another $184 (4 x 46 = 184).

The Berkeley survey also reported that direct replacement costs averaged about 9 percent of the annual wage. That figure counts recruiting and training only. Add the vacancy and ramp lines and the same job lands near 28 percent, which is how two honest analysts end up quoting numbers that differ by a factor of three.

Line Three: What an Empty Desk Costs

An empty desk costs money only where the work stops. Divide salary by 260 workdays to get the daily rate, then apply a coverage gap: the share of the job's output that nobody else picks up.

At $52,000 the daily rate is $200 (52,000 / 260 = 200). A 32 day search with a 60 percent gap costs $3,840 (200 x 32 x 0.6 = 3,840).

The 60 percent is a judgment, and it should be. A reception desk in a staffed office may run near zero, because coworkers answer the phone and the mail still goes out. A single payroll specialist with no backup runs near 100 percent in the week a pay run is due. Write down which one you are pricing, because that one number swings the total more than the salary does.

Overtime paid to cover the gap is a separate cash line. Two coworkers picking up 20 hours each at $28 an hour adds $1,120 (2 x 20 x 28 = 1,120). Leave it out and the vacancy looks cheaper than the payroll clerk will tell you it was.

Long searches are where this line gets out of hand. Median time to fill runs close to six weeks in most surveys, and specialized roles run past 60 days. Each extra week at $200 a day and a 60 percent gap costs $840 on the rep in our example (7 x 200 x 0.6 = 840).

Every row holds the daily rate fixed at $200, so the only moving parts are how long the search runs and how much of the work goes uncovered. The last column is the seat nobody covers, which lifts the line by two thirds. That column is the one to argue about, because a role with no backup stays uncovered for the whole search.

Lines Four and Five: Training and the Slow Ramp

Training is the easy half: a course, a certificate, a badge, a week of a supervisor's time. Put a number on it and move on.

The ramp is where the money hides. A 2010 working paper from UC Berkeley's Institute for Research on Labor and Employment, based on a panel of California employers surveyed in 2003 and 2008, found that professional and managerial hires took about 3.3 months to reach full group productivity and non-professional hires about 2 months. The authors model output as a straight climb from zero, which puts the lost output at roughly half the salary paid during that window.

So a $52,000 rep with a three month ramp at a 50 percent gap loses about $6,500 (52,000 / 12 = 4,333.33, and 4,333.33 x 3 x 0.5 = 6,500). A $95,000 developer with a four month ramp at a 60 percent gap loses $19,000 (95,000 / 12 = 7,916.67, and 7,916.67 x 4 x 0.6 = 19,000).

This is the line a written 30-60-90 plan attacks. Every week cut from the ramp saves a week of partial salary and costs nothing but attention. A manager who checks in daily for the first two weeks and then weekly for two months usually beats a formal program that nobody runs.

Worked Example: A $52,000 Customer Service Rep

A rep earning $52,000 leaves in March. The company fills the seat in 32 days, and the new hire is up to speed in about three months.

1. Recruiting spend: $1,200 for a job board post and a referral bonus.
2. HR time: 22 hours at $34, which is $748 (22 x 34 = 748).
3. Vacancy loss: $200 a day for 32 days at a 60 percent coverage gap, which is $3,840 (200 x 32 x 0.6 = 3,840).
4. Training: $1,500 for a two week onboarding course and a buddy's time.
5. Ramp loss: $4,333.33 a month for 3 months at a 50 percent gap, which is $6,500 (52,000 / 12 = 4,333.33, and 4,333.33 x 3 x 0.5 = 6,500).
6. Exit cost: $900 for the close-out paperwork, a benefits notice, and the unused vacation the company chose to pay out.
7. Total: 1,200 + 748 + 3,840 + 1,500 + 6,500 + 900 = 14,688.
8. As a share of salary: 14,688 / 52,000 = 28.2%.

The two output lines, vacancy and ramp, come to $10,340, or about 70 percent of the total. That is the normal shape of this math, and it is why a manager who budgets only the job board post thinks a departure costs about a thousand dollars.

Compare that with the SHRM average of $4,683. The gap is not an error in either number. SHRM measures what a company spends to fill a seat. This model also counts what the seat produced while it was empty.

Worked Example: A $95,000 Developer With a Recruiter Fee

Now a developer at $95,000, hired through an agency at an 18 percent fee, in a search that ran 52 days.

1. Agency fee: 18 percent of $95,000, which is $17,100 (95,000 x 0.18 = 17,100).
2. HR time: 30 hours at $42, which is $1,260 (30 x 42 = 1,260).
3. Vacancy loss: $365.38 a day for 52 days at a 60 percent gap, which is $11,400 (365.38 x 52 x 0.6 = 11,400).
4. Training: $2,500 for a conference and internal tooling time.
5. Ramp loss: $7,916.67 a month for 4 months at a 60 percent gap, which is $19,000 (7,916.67 x 4 x 0.6 = 19,000).
6. Exit cost: $2,000 for three weeks of severance and the administrative close-out.
7. Total: 17,100 + 1,260 + 11,400 + 2,500 + 19,000 + 2,000 = 53,260.
8. As a share of salary: 53,260 / 95,000 = 56.1%.

Same six lines, more than three times the bill, and the agency fee alone is larger than the entire customer service total. That is the pattern worth remembering: senior roles cost more to lose because the search runs longer and the ramp is steeper, not because the paperwork is harder.

From One Departure to a Yearly Number

One departure is a story. A year of them is a budget line.

Turnover rate = departures in 12 months / average headcount

A 120 person company that lost 22 people runs an 18.3 percent turnover rate (22 / 120 = 18.3%). At $14,688 a departure the year costs $323,136 (14,688 x 22 = 323,136). Payroll for 120 people at $52,000 is $6,240,000 (120 x 52,000 = 6,240,000), so turnover equals 5.2 percent of payroll (323,136 / 6,240,000 = 5.2%).

For context, BLS counted 38.0 million quits in 2025, 60.6 percent of 62.8 million total separations, and the annual average quits rate was 2.0 percent a month. A 120 person company running at exactly the national pace would see 28.8 quits a year (120 x 0.02 x 12 = 28.8). At $14,688 each, that is $422,054 (28.8 x 14,688 = 422,054).

The national rate is a mirror, not a target. Call centers and restaurants run far above it, law firms and utilities far below. Compare your own rate with your own history first, then with your industry, and treat the national figure as the last check rather than the first.

What This Number Leaves Out

This model prices what a company can count. It leaves out several costs that are real:

• Customers who followed the person who left, which matters most in sales and account management.
• Knowledge that walked out and has to be rebuilt, which can take a year on technical work.
• Morale. Two more people often follow a manager out the door within six months.
• The hiring manager's own hours, which never appear in a recruiter's time log.
• The replacement's higher salary. If the new hire signs 8 percent above the old rate, that increase is permanent, and on a $52,000 role it adds $4,160 a year for as long as the person stays (52,000 x 0.08 = 4,160).

None of those belong in a calculator that has to defend every line. What belongs there is a range. Run the model three times: a fast fill with a short ramp, your actual history, and a slow fill where the seat sits empty for a full quarter. The spread between the low and high runs is the honest uncertainty, and it is usually wider than the difference between two studies you might argue about.

One more caution. This is an estimate for planning and budgeting, not an accounting entry and not tax advice. If you need a figure for a financial statement, ask your accountant which of these lines can be booked and how.