Sales rep turnover in 2026: attrition rates, cost of one lost rep, and the famous numbers

What is the average sales rep turnover rate?

Around 25%, if you want one number with a publisher behind it. Xactly’s sales statistics page lists "Sales organizations reported 25% average turnover over the last 12 months" and attributes it to Salesforce’s State of Sales. Salesforce’s own 2026 statistics page carries no turnover figure today, so the number is secondhand, though it is the most defensible one in circulation. Every other figure is a change rather than a level, a claim about a subset, or a number whose original cannot be found.

On the by-role question, no public breakdown with a named, reachable publisher exists. The by-role table on the most detailed 2026 vendor page is labeled "Industry averages, Xactly 2025", which we could not find on Xactly’s site. The by-role number that matters is your own, computed as described below.

What is a good or acceptable sales attrition rate?

No publisher defines "good", so use two tests instead of one number. Against the market, at or under the 25% benchmark is normal and 35% is the high end of what circulates. Against your own plan, the number that matters is regretted turnover among ramped reps, because a rep who leaves after ramp takes the investment with them. The bands below are our rule of thumb, not an industry standard.

Tenure at exit is the second test. Xactly’s own data puts a rep’s peak performance "between two and three years in their role". A team whose leavers average 18 months keeps paying for ramp and never collects the peak. Pull tenure at exit for your last ten departures; if most are under two years, the rate is the smaller problem. Pay is a third check. Xactly Insights reports 50% less sales turnover at companies paying at the 75th percentile or higher (its own data, method not stated), a reason to check pay against the market before you look at the managers.

How do you calculate sales turnover, and why does regretted vs non-regretted matter?

Turnover rate = departures in the last 12 months ÷ average headcount over the same 12 months × 100. Count everyone who left the roles you are measuring, and divide by the average headcount rather than the year-end number, or a growing team flatters itself. The regretted split is what turns the rate into a decision, because the two kinds of exit cost different money and are fixed by different people.

1. Count departures over a trailing 12 months. Fix the roles you include and keep them fixed.
2. Compute average headcount. Add each month-end headcount and divide by 12; (start + end) ÷ 2 is an acceptable shortcut.
3. Divide and multiply by 100. Three departures on an average of ten reps is 30%.
4. Split voluntary from involuntary. A rep you exited is a hiring or management decision and belongs in a different conversation.
5. Split regretted from non-regretted. Regretted means you would have kept them at their last rating.
6. Record tenure at exit and the manager for every leaver. This is the line a VP reads by manager.
7. Note when the manager first knew. The weeks between the first signal and the notice are what the cost model turns into money.

Where do "35% vs 13%", "18 months", "$115,000" and "1.5–2× salary" come from?

Three of the five figures below trace to one 2022 Xactly post that cites HubSpot; the HubSpot page behind the 35% no longer exists, and the one behind the 18 months rests on a 2018 Bridge Group report. The two replacement-cost multiples have no public original we could reach. That does not make them wrong. It means nobody can say what they measured, so they cannot anchor a budget.

When a deck shows a turnover statistic, ask for the publisher, the year and the definition. If the answer is "HubSpot" or "a study", it is one of the rows above.

What does replacing one sales rep cost, line by line?

A departure costs you in two currencies. Cash goes out on recruiting, onboarding and the manager’s hours. Quota capacity goes unproduced while the seat is empty and again while the new rep ramps. Most published models add the two into one total; we keep them apart, because a dollar of quota is not a dollar of cost and only you know your margin. Every assumption below is illustrative; replace it with your own.

Now the team. Ten reps carry $9.6 million of quota a year on these assumptions. At the 25% benchmark you lose two or three reps a year (2.5 on average); at the circulating 35%, three or four.

For comparison, Zyverno’s vendor model (May 2026) prices a $70,000-base, $120,000-OTE rep at "$215,500-$292,000" per replacement event, with a 5.7-month ramp gap of $140,000–$175,000 as the largest line, and calls the $115,000 figure "conservative". It adds cash and pipeline into one number. Both models say the same thing. The ramp gap is the bill; the recruiting fee is only the receipt.

Why do sales reps quit?

Career advancement, on the only current large-sample source. Salesforce’s 2026 State of Sales statistics (February 2026, from a survey of over 4,000 sales professionals) state that "Lack of career advancement is the #1 reason sales reps want to change jobs" and call it "the single biggest threat to retention".

The Gartner sales survey of August 2022, as reported by destinationCRM because Gartner’s own release is not publicly accessible, found that "Nearly 90 percent of sellers report feeling burned out from work and more than half (54 percent) are actively looking for new jobs as a result". Take the wording as the survey’s own self-report; this page claims nothing beyond it. The useful part for a VP is Gartner’s four sources of drag, quoted as "Lack of development opportunities; Feeling like just another cog in a machine; Lack of clear manager feedback; and High burden of non-value-added administrative tasks". Two of the four sit with the first-line manager.

• A career path they cannot see. Salesforce 2026, reason number one. Ask each manager which reps asked about a next role this year and what they were told.
• No development, no clear feedback. Two of Gartner’s four drag sources. Both show in whether 1:1s happen and what they contain.
• Administrative load. Gartner’s fourth source; Salesforce 2026 puts non-selling work at "60% of their time".
• Pay below market. Xactly Insights reports 50% less sales turnover at companies paying at the 75th percentile or higher (its own data).
• Nothing a survey shows. Some leavers were simply recruited away. That is the unpreventable share; the next section separates it out.

Which turnover costs can a first-line manager prevent, and how early would they have to notice?

The cash lines are spent the moment a resignation lands. The capacity lines are the ones a manager can shrink, and each depends on how early the manager knew. A stay conversation before the rep starts interviewing avoids the whole bill. A cover plan drafted at the first credible signal cuts the vacancy. A ramp watched weekly shortens the gap. The timings below are illustrative.

A manager who learns of a resignation on notice day pays the full bill. A manager who saw the drift two months earlier gets to choose which bill to pay. That is why a VP should ask for the lead time as well as the rate. None of the signals below is proof; any two in the same month are a reason for a conversation this week. Rules on recording, monitoring and employment differ by jurisdiction and this page is not legal advice; check with HR or an employment lawyer before it becomes policy.

For the full list, see signs a sales rep is about to quit and how to spot sales rep burnout early.

What should a VP track about turnover, by manager?

Four numbers per manager, once a quarter, each with the evidence next to it. Ask for them in the monthly review and the managers will start collecting them.

• Regretted turnover of ramped reps, by manager. One manager at 30% next to three at 8% is a manager question. All four at 25% is a pay or career-path question.
• Tenure at exit, against the two-to-three-year peak. Leavers under 24 months mean you funded the ramp and skipped the return.
• Lead time, the weeks between the first signal the manager can point to and the notice. Zero, quarter after quarter, means the manager is not looking or cannot see.
• Stay conversations held, as a count per manager against the reps the manager was concerned about this quarter, without names in the VP review. A concern without a conversation is a cost you chose.

The rate says whether you have a problem. These four say where it lives and how early anyone could have acted. Bring the benchmarks to the board; bring these to the managers.

Frequently asked questions