Why 18 months is the worst time to lose a sales rep

The average sales rep leaves their job at 18 months. Sales reps hit their peak performance between two and three years in the role. Xactly’s own research puts the peak-performance window at 24–36 months; the 18-month average tenure figure it cites traces back to HubSpot. Put those two numbers next to each other and the problem is obvious: most reps are walking out the door right as they’re about to become worth what you paid to ramp them.

This isn’t the same argument as promoting from inside your BDR seat — that’s about where your next hire comes from. This is about what happens after someone’s already on the team: the reps you keep for 3–5 years are worth more than the tenure-adjusted math makes obvious, and most comp and hiring conversations don’t account for it at all.

The gap, in two numbers

18 months average tenure. 24–36 months to peak performance. Call the difference the tenure gap — on average, six to eighteen months of runway a company never gets to collect on, because the rep is gone before they get there. Turnover isn’t evenly distributed across roles either: a 939-company survey from Optifai puts annual turnover at 45% for SDRs, 30% for AEs, 28% for sales managers, and 25% for CSMs — the roles earliest in the ramp curve churn the fastest, which is exactly backwards from where you’d want the stability.

“Fully ramped” isn’t the same thing as peak

It’s worth being precise here, because our own ramp-benchmarks post already covers a different milestone. “Fully ramped” means producing at the level of a tenured rep in the same seat — for a mid-market AE, that’s a median of around 6 months. “Peak performance” is a later, higher ceiling: the point where product depth, ICP instincts, and a rep’s own pattern-matching on objections have had time to compound. A rep can clear the ramp milestone at month 6 and still be climbing toward their real ceiling for another year and a half. The ramp benchmark tells you when someone stops being a liability. It doesn’t tell you when they’re done getting better.

What early turnover actually costs

The direct cost is the easy part to see: recruiting, onboarding, an open territory, a new ramp clock starting from zero. The harder cost to see is the one implied by the tenure gap itself — a company that loses reps at 18 months isn’t just paying to replace them, it’s paying full ramp cost on every hire and almost never collecting the higher-output years the ramp was supposed to buy. Xactly cites Forbes on the multiplier effect: a 5% increase in rep attrition raises total selling costs 4–6% — attrition doesn’t just cost what it costs to backfill a seat, it drags on the economics of the whole motion.

The unit economics of staying

Ramp cost is mostly fixed — it’s paid once, whether the rep stays 18 months or 5 years. Output is not fixed — on the numbers above, it keeps climbing for roughly two more years after ramp, then presumably holds near that ceiling for as long as the rep stays engaged and the motion doesn’t change under them. That means the real cost per unit of value a rep produces keeps dropping the longer they stay past their ramp date, not staying flat. A rep who leaves at 18 months delivers barely-past-ramp output against a full ramp cost. A rep who stays 4 years delivers 2–3 years at or near peak against that same sunk cost. Same ramp expense, very different return — and almost no comp plan or hiring model treats tenure as the lever it actually is.

What actually extends tenure

The data above doesn’t tell you how to fix retention, and we’re not going to invent a stat to fill that gap. What’s directionally consistent across the sourcing here: the roles earliest in the ramp curve (SDR, then AE) also churn fastest, and those are the roles with the least visibility into how their own number ties to their own pay — a rep who can’t see their commission move in real time, or can’t see a real path to the next role, has less reason to wait around for month 24. A visible growth path is one lever — promoting from inside instead of hiring cold keeps a rep’s tenure clock running instead of resetting it. Comp transparency is the other, and it’s a lower bar than it sounds: a rep who can see exactly what they’re owed, updated as deals close, isn’t guessing about whether the plan is fair — which is a different problem than the plan actually being fair, but one that’s just as corrosive to tenure when it goes unaddressed.

What to do next

If you’re staffing for the next 12 months, the tenure gap is worth budgeting around explicitly — not just “how many reps do I need to hire” but “how many of this year’s reps will still be here at month 24, and what changes if that number goes up.” For the ramp math specifically, the ramp benchmarks post has the role-by-role timeline; for the comp-plan side of keeping a rep’s trust intact once they’re ramped, how to write your first comp plan is the companion piece.