Accelerators and Decelerators: How to Structure Them Without Inviting Gaming
Accelerators and decelerators shape rep behavior more than any other lever in a comp plan. The mechanics that work, and the common ways they get gamed.
, 3 min read, Compensation
Key takeaways
- An accelerator that starts too low invites deal-splitting; one that starts too high never gets used and stops motivating anyone.
- Decelerators are riskier than accelerators because they feel punitive even when the underlying logic is sound.
- Gaming isn't a rep integrity problem first, it's a plan design problem: reps optimize for whatever the plan actually rewards.
Accelerators and decelerators do more to shape a rep's actual behavior, quarter by quarter, than the base commission rate ever will. Get the thresholds and multipliers right and a plan quietly nudges reps toward the deals that matter. Get them wrong and reps will find the gap in the math faster than finance will.
What each one is for
An accelerator raises the commission rate once a rep crosses a threshold, usually 100% of quota. A plan might pay a flat 8% up to quota and 12% on every dollar past it. The logic: the marginal dollar above quota is more valuable to the business than the average dollar below it, since fixed costs are already covered, so pay the rep more for it.
A decelerator lowers the rate below a threshold, for example paying a reduced rate on deals closed while a rep sits under 50% attainment. The logic here is murkier: it's meant to discourage low-quality deals rushed out purely to hit a number, but it also punishes reps for a slow quarter regardless of cause.
The mechanics that hold up
Start accelerators at or near 100%. Below 80%, an accelerator loses its meaning: it stops being a reward for exceeding target and starts looking like a standard rate the company disguised as a bonus. Above 120%, it stops motivating anyone who won't realistically get there, which on most teams is the majority of reps.
Use a small number of tiers, two or three at most. A plan with five accelerator bands past 100% turns commission into a math problem reps solve with a spreadsheet instead of a sales strategy they execute in the field. Two tiers, say 100% to 150% and 150%-plus, cover the incentive without the complexity.
Tie the accelerator to the same period as the quota. An accelerator that resets monthly on a quarterly quota, or vice versa, creates artificial cliffs where reps rush deals to beat a reset date that has nothing to do with the actual sales cycle.
If you use a decelerator, tie it to something the rep controls. A decelerator on discount depth (deals closed below a margin floor pay a lower rate) targets a real behavior. A decelerator tied purely to low attainment penalizes reps for market conditions, territory quality or a shortened ramp, none of which they chose.
The failure modes to design against
Deal-splitting. When crossing a threshold pays disproportionately more than the deal itself is worth, reps ask customers to sign two smaller contracts instead of one, timed to cross the line twice. This shows up almost immediately in deal-size distributions once a plan launches: a sudden cluster of deals just above a threshold, and a gap just below it.
Deal-shifting across periods. A rep close to quota in the final week of a quarter has every incentive to push a deal that's ready to close into next period instead, banking it against a fresh, easier-to-hit number. The fix isn't punitive, it's structural: make sure the accelerator reward for closing now exceeds the reward for waiting.
Sandbagging. The mirror image: a rep who has already cleared the accelerator threshold has no incentive to close anything else this period, since the marginal deal earns the same rate either way until a new tier. Multiple accelerator tiers reduce this by keeping a reason to keep closing all the way through the period.
Discount races near a decelerator cliff. If a decelerator triggers below a hard attainment line, reps sitting just above it near period end will discount aggressively to protect their position, giving away margin the accelerator was never meant to cost the business.
Modeling before you launch
The only reliable way to catch these failure modes before they cost real money is to run last year's actual deal-level data through the proposed thresholds and watch where the incentives point. If deals cluster suspiciously around a threshold in the simulation, they'll cluster in the field too, just with real revenue attached.
The honest tradeoff
Accelerators are cheap insurance: they cost the company money only when a rep is already outperforming, which is exactly when the company wants to pay more. Decelerators are a harder sell, because they save the company money precisely when a rep is already struggling, and reps read that correctly as punishment layered on top of a bad quarter. Use accelerators generously. Use decelerators sparingly, and only where the trigger is something a rep genuinely controls.
Frequently asked questions
- At what attainment level should an accelerator start?
- Most plans start accelerating at 100% of quota, sometimes earlier at 80% to reward strong quarters even when a rep narrowly misses target. Starting above 120% means most reps never reach it and the accelerator stops functioning as an incentive.
- Do decelerators actually improve performance?
- Rarely on their own. A decelerator can discourage bad-faith discounting or a rush of low-quality deals near quarter end, but it also punishes reps for factors outside their control, like a shortened quarter or a slow market, and tends to damage trust faster than it changes behavior.
- What is deal-splitting and how do accelerators cause it?
- Deal-splitting is when a rep asks a customer to sign two smaller contracts instead of one large one, timed to cross an accelerator threshold twice instead of once, or to push a deal into the next period where a fresh threshold resets. It's a direct response to a threshold that rewards crossing a line more than the size of the deal itself.