Modeling an Accelerator Curve Before You Commit to It

An accelerator pays a higher rate above a threshold, rewarding reps who clear quota with more per dollar on the overage. A decelerator does the opposite below a threshold, paying a reduced but nonzero rate. A cliff is more absolute: it pays nothing at all below its threshold. Mixing these terms up during plan design is a common source of confusion once finance, sales leadership, and reps start discussing the same plan using different assumptions about which one applies where.

Take every rep's actual attainment from last year and calculate what they would have earned under the proposed new curve, comparing it to what they actually earned. Say your proposed plan pays an accelerated rate above full quota: if almost none of last year's reps ever cleared that mark, the accelerator is effectively decorative, and if nearly everyone cleared it easily, the threshold is set too low to reward genuine outperformance.

A curve can look fair to an individual rep and still be unaffordable in aggregate if a large share of the team clusters right above the accelerator threshold. Model total company-wide payout under the new curve against last year's full attainment distribution, not just a hypothetical top performer, so finance sees the real cost before the plan launches rather than discovering it the first time a strong quarter triggers it across the whole team.

For example, imagine a team where most reps finish just above quota. An accelerator that looks modest for any single rep will trigger for nearly everyone at once, and total payout can jump well past what finance budgeted. Running last year's real attainment through the curve exposes that clustering before launch. If it appears, you have several options: raise the threshold, reduce the accelerated rate, or cap the accelerated portion. Each changes the cost differently, so model each variant and compare total payout, not just per-rep fairness. Bring the comparison to finance so the choice is made on numbers, not on which version sounds most generous on a whiteboard.

A steep cliff right at the quota line can push reps to hold a deal back into the next period if they are just short, or pull a deal forward if they are just barely over, purely to land on the better side of the threshold. Model whether your proposed thresholds sit close enough to typical deal sizes that this kind of timing gaming becomes likely, and adjust the threshold or smooth the curve if it does.

Once the modeled curve holds up against both individual fairness and aggregate cost, get explicit finance sign-off on the total payout scenario before presenting anything to the sales team. A curve that gets announced and then quietly walked back because finance never actually approved the full cost model does more damage to trust than taking an extra week to model it properly in the first place.

Work through the modeling in this order:

1. Calculate what each rep would have earned last year under the proposed curve and compare it with what they actually earned.
2. Check total company-wide payout against the full attainment distribution, not just a hypothetical top performer.
3. Test whether thresholds sit close enough to typical deal sizes to invite deals being pulled forward or held back.
4. Run the same curve against a weaker year or a soft quarter to see how it behaves.
5. Get explicit finance sign-off on the total payout scenario before the sales team sees anything.

Most curve modeling naturally uses a recent representative year as the reference point, but it is worth also running the same curve against a weaker year, or a hypothetical soft quarter, to see how the plan behaves when the business is not performing at its recent pace. A curve that only ever gets stress-tested against solid years can hide a problem that only surfaces when performance genuinely dips: an accelerator threshold that suddenly looks completely out of reach, or a decelerator that starts biting a much larger share of the team than anyone anticipated.

Understanding how the curve behaves in a down scenario matters for two different audiences. Finance needs to know the plan's cost does not become unpredictable in a bad quarter, and sales leadership needs to know the plan will not demoralize the entire team the first time results genuinely soften. Neither of those questions gets answered by modeling only against a strong or average year, so build the weaker scenario into your standard modeling process rather than treating it as an afterthought only considered once a soft quarter has already arrived.