Call center attrition rate is the share of agents who leave during a period, as a percentage of the average number employed: divide leavers by average headcount and multiply by 100. Attrition and turnover usually name the same number in a call center report. In NICE's 2025 survey of 400 contact center leaders, average agent attrition was 39 percent for 2024 and 36 percent in North America, down from 49 percent in 2023.
The rate is most useful once you turn it into hires. At 36 percent a year, a 60-agent team replaces about 22 people, close to two a month, and with recruiting and training lead times of two to three months, each new hire has to start well before the resignation it covers. The table lists each formula in this guide with a worked example.
| Step | Formula | Example |
|---|---|---|
| Monthly attrition rate | Leavers in the month ÷ average headcount in the month × 100 | 3 leavers ÷ 60 agents = 5.0 percent |
| Annual attrition rate | Leavers in 12 months ÷ average headcount over the 12 months × 100 | 18 leavers ÷ 50 agents = 36 percent |
| Hires a year to hold the team level | Seats × annual attrition rate | 60 seats × 36 percent = 21.6, so 22 |
| Hires to start now for a date L months out | Seats needed then − agents on hand × (1 − monthly rate)^L | 60 − 60 × 0.97³ = 5.2, so 6 |
| Cost per departure | 16 percent of annual pay, or three to four months' pay with lost productivity | $7,163, or $11,193 to $14,923 at $44,770 a year |
What is attrition rate in a call center?
Attrition rate is the number of agents who left during a period divided by the average number employed over it, written as a percentage. Report it monthly and annually, and keep frontline agents apart from supervisors and back-office staff; NICE's survey reports agents and back-office employees separately. Contact center reports use attrition and turnover for the same figure, though some HR teams keep attrition for positions that are not backfilled.
Settle what counts as a departure before you compare your figure with anyone's, because definitions vary more than the formula does. The Global Call Center Report, a survey of almost 2,500 call centers in 17 countries, counted promotions, voluntary quits, retirements and dismissals in its total turnover rate (Holman, Batt and Holtgrewe, 2007). NICE's survey reports what it calls unmanaged attrition. A center that counts agents promoted to supervisor as leavers reports a higher rate than one that doesn't, from the same year of exits.
| Departure | In the agent attrition rate? | Why |
|---|---|---|
| Resignation | Yes | The core of the measure |
| Dismissal | Yes, in its own column | It answers a different question from resignations |
| Promotion or transfer to another team | Report it separately | It's a win for the business, but the seat still needs a hire |
| Trainee who leaves before taking calls | Yes, and in an early-tenure rate | You paid for the training and got no calls back |
| Agent on leave | No | Still employed; it belongs in shrinkage |
| Seasonal agent at the planned end of a contract | No, or in a separate seasonal rate | The exit was part of the plan |
How do you calculate call center attrition rate?
Divide the agents who left in the period by the average headcount for the period, then multiply by 100. Average headcount is the mean of the opening and closing counts for a month, or the mean of the monthly figures for a longer period. A team that started October with 61 agents, ended with 59 and lost 3 along the way had a monthly attrition rate of 3 ÷ 60, or 5.0 percent.
In a spreadsheet, with leavers in B2, opening headcount in B3 and closing headcount in B4, the monthly rate is =B2/((B3+B4)/2), formatted as a percentage. For a year, list each month's leavers in C2:C13 and each month's average headcount in D2:D13, and the annual rate is =SUM(C2:C13)/AVERAGE(D2:D13).
Run one more rate beside the headline figure: early-tenure attrition, the share of each hiring class that left within 90 days. Divide the class's leavers by the class size. It separates a hiring and onboarding problem from a retention problem, and the group is large: in the Global Call Center Report, about one agent in five in the liberal-market countries (the US, Canada, the UK and Ireland) had less than a year of tenure.
How do you convert a monthly attrition rate to an annual rate?
Add the twelve monthly rates, or divide the year's leavers by the year's average headcount; both give the annual turnover rate. That figure isn't the share of January's team gone by December. At an even 3 percent a month, annual attrition is 36 percent, yet about 69 percent of the agents employed in January would still be there at year end, because some departures are replacements who left too.
| Monthly attrition | Annual attrition | Starting team still employed after 12 months, at an even rate |
|---|---|---|
| 2 percent | 24 percent | 78.5 percent |
| 3 percent | 36 percent | 69.4 percent |
| 4 percent | 48 percent | 61.3 percent |
| 5 percent | 60 percent | 54.0 percent |
With the monthly rate in B2, =(1-B2)^12 gives the share of the starting team still employed. If your new hires leave faster than tenured agents, which the early-tenure rate will show you, the starting team holds up better than this table and the hiring classes do worse. Build annual comparisons from the trailing 12 months as well: one bad month multiplied by 12 overstates the year.
What is the average call center turnover rate?
The latest large survey puts agent attrition in the high 30s. NICE's 2025 survey of 400 contact center leaders found average agent attrition of 39 percent in 2024, 36 percent in North America and 41 percent in EMEA, against 49 percent in 2023. Older research found in-house centers turning over less than outsourcers. Compare your own trend with these figures; none of them is a target.
| Source | Figure | What it covers |
|---|---|---|
| NICE, 2025 Workforce Management Trends survey | 39 percent in 2024, 49 percent in 2023 | Average agent attrition; 400 leaders split between North America and EMEA |
| Same survey | 36 percent North America, 41 percent EMEA | Mean agent attrition, 2024 |
| Global Call Center Report (2007) | 20 percent typical; 25 percent median in liberal-market countries | Total turnover, promotions and dismissals included |
| Same report | 19 percent in-house, 25 percent at subcontractors | Average annual turnover |
| Gans, Koole and Mandelbaum (2003) | Exceeds 50 percent a year "in many centers" | An observation in a review of call center research, not a survey |
Sources read November 10, 2026.
Two findings matter most for a center you run yourself. NICE's respondents mostly felt attrition was rising (58 percent said agent attrition had increased) in the same year their reported averages fell ten points, so check the trend in your own numbers before you act on the feeling. And the 2007 report found in-house centers turning over about a quarter less than subcontractors. Keep that in mind when an outsourcing quote prices agents by the hour and leaves turnover on the provider's side of the contract; the outsourcing comparison covers the rest of that math.
Occupation-wide figures from O*NET point the same direction. It projects 289,500 openings a year for customer service representatives from 2025 to 2035, against 2,666,000 people employed in 2025, while employment declines (O*NET 43-4051.00). That's about one opening for every nine jobs each year. Those openings count people leaving the occupation, not agents moving to another employer, so turnover at any one employer runs higher.
How many agents do you need to hire to cover attrition?
Multiply the seats you have to keep filled by your annual attrition rate. At 36 percent, a 60-agent team needs about 22 hires a year to stay level. Then move each hire earlier by your recruiting and training lead time, which Gans, Koole and Mandelbaum put at two or three months or more, and size it for the attrition that happens while the new class is in training.
Their paper, Telephone Call Centers: Tutorial, Review, and Research Prospects, describes a simple method its authors had seen in use: hire now enough people that, once trained, they cover the gap between seats needed at that date and the agents you'll still have after turnover. With a monthly rate m and a lead time of L months:
Hires to start now = seats needed in L months − agents on hand × (1 − m)^L
Say you need 60 seats in three months, you have 60 agents and nobody in training, and attrition runs 3 percent a month. Sixty times 0.97 cubed is 54.8, so you start 6 people now, though nobody has resigned yet. In a spreadsheet, with seats needed in B2, agents on hand in B3, the monthly rate in B4 and the lead time in B5:
=ROUNDUP(B2-B3*(1-B4)^B5,0)
Add people already in training to B3, discounted for the months until they take calls. Like the paper's version, the formula assumes nobody drops out of training, so divide by one minus your in-training drop-out rate if you track one: 6 hires with one trainee in six lost means a class of 8. The seats figure itself comes from your staffing model for the busiest intervals, divided by one minus shrinkage, which the workforce management guide walks through.
How much does call center attrition cost?
The Global Call Center Report put the simple cost of replacing one agent at 16 percent of a call center worker's gross annual earnings, about two months' pay, and three to four months' pay once lost productivity is counted. At O*NET's 2025 median wage for customer service representatives, $44,770 a year or $21.53 an hour, that's about $7,163 per departure, or $11,193 to $14,923.
A broader review lands in the same place. The Center for American Progress pooled 30 case studies from 11 research papers and found a typical turnover cost of 20 percent of annual salary for jobs paying under $50,000 a year. Here is what the call center figures add up to for a 60-seat team at three attrition rates.
| Annual attrition | Departures a year | Hires a month | Simple replacement cost a year | With lost productivity |
|---|---|---|---|---|
| 20 percent | 12 | 1.0 | $86,000 | $134,000 to $179,000 |
| 36 percent | 21.6 | 1.8 | $155,000 | $242,000 to $322,000 |
| 50 percent | 30 | 2.5 | $215,000 | $336,000 to $448,000 |
Both studies price a departure against pay rather than loaded cost, and $44,770 is a national median, so read the table as a floor. To build your own figure, add up these costs per departure:
| Cost | Where the number comes from |
|---|---|
| Recruiting | Ad spend, agency fees and recruiter time, divided by hires |
| Interviews and onboarding | Hours per hire times the hourly cost of everyone involved |
| Paid training before the first live call | Training weeks times trainee pay, plus trainer time |
| Ramp | New agents' handle time against tenured agents' in their first months (average handle time) |
| Cover for the empty seat | Overtime hours, or the service level lost in the intervals that seat covered |
What causes high call center attrition?
The reasons managers give differ by region. NICE's North American respondents cited market competition, dissatisfaction with pay and budget cuts more often than EMEA respondents, who more often reported high workload. Your own exit list says more than either, so sort leavers by tenure, team, shift and stated reason before you pick a fix.
| What your exit list shows | Where to look first |
|---|---|
| Most leavers had under 90 days of tenure | The hiring profile, how the job was described in interviews, and the first weeks on the phones |
| Leavers at every tenure cite pay or another offer | Your pay against local employers hiring for the same skills |
| Leavers concentrate under one supervisor or team | How that team uses metrics and coaching (call center metrics) |
| Leavers cite workload or burnout | Occupancy in the busiest intervals (call center occupancy) |
| Leavers cluster after a schedule change, or on night and weekend shifts | Schedule flexibility and how shifts are assigned (schedule adherence) |
The survey and the report back up the last two rows. Of NICE's respondents, 48 percent said flexible scheduling had reduced attrition, yet only 44 percent offered it. In the Global Call Center Report, typical turnover was 9 percent in centers where jobs combined high agent discretion with low monitoring, and 36 percent where agents had little discretion and close monitoring.
Where Telvana fits in the hiring plan
Telvana's AI agents answer calls from a phone-menu option, as overflow or after hours, and run outbound confirmation and reminder calls from a list within the hours you set. A call that needs a person transfers to the number you choose. The calls an agent resolves come out of the volume your staffing model covers, so fewer seats get multiplied by your attrition rate.
Nights and weekends are a sensible place to start. In calls to our customers' business lines, 37.6 percent arrived outside weekday 8am to 6pm (what 181,555 AI agent calls show), and an after-hours and overflow rule can send those hours to an agent. Our team builds each agent from your scripts and transfer rules and tests it with you, so it's live in days rather than after a two-to-three-month hiring cycle. Pricing starts at $399 a month based on call volume, plus a one-time setup fee (pricing).
Frequently asked questions
What is a good attrition rate for a call center?
No standard target exists. The most recent survey average is 36 percent a year in North America (NICE, 2024 figures), and the 2007 Global Call Center Report found 19 percent at in-house centers. A rate that falls year over year, with early-tenure attrition falling alongside it, is a better test than any single benchmark.
What is the difference between attrition and turnover?
In most contact center reports there is none: both mean leavers divided by average headcount. Some HR teams use turnover for departures that get replaced and attrition for positions left unfilled. Whichever word you use, say which departures you count, promotions and dismissals above all, whenever you compare your figure with someone else's.
How do you calculate monthly attrition rate?
Divide the agents who left during the month by the month's average headcount, the mean of the opening and closing counts, and multiply by 100. Three leavers from a team that averaged 60 agents is 5.0 percent. Track it monthly, but build annual comparisons from the trailing 12 months, because one bad month multiplied by 12 overstates the year.
Why is call center turnover so high?
Survey respondents point to pay and competition for workers in North America and to workload in EMEA (NICE, 2025). Job design matters as well. The Global Call Center Report found typical turnover of 9 percent where agents had high discretion and low monitoring, against 36 percent where they had little discretion and close monitoring.
How far ahead should a call center hire for attrition?
By your full recruiting and training lead time, which Gans, Koole and Mandelbaum put at two or three months or more. Hire for the seats you'll need when the class goes live, minus the agents you expect to still have by then after attrition, not for the resignations you've already received.
Put last year's exits through the worksheet
Pull last year's leavers by month, tenure, team and reason, and run them through the formulas above before you compare with any benchmark. Our staffing calculator gives the agents your busiest half hour needs, which is the seat count your attrition rate multiplies. To hear an AI agent take one of your call types, book a demo.