What makes a good sales compensation plan?

One a rep can compute in their head, that pays for outcomes they control, and whose structure does not reward timing games. Simplicity is not a nicety — a plan nobody can calculate cannot motivate the behaviour it was designed for.

How many components should a comp plan have?

Few. Every additional accelerator, kicker and multiplier divides attention and adds a place to game. If a rep cannot state what their next deal is worth to them, the plan has too many parts.

Do accelerators work?

They increase effort above target and they also produce pull-forward and push-out around the boundary. Both effects are real; the question is whether the extra output is worth the timing distortion, which you can measure in your own close-date distribution.

How often should comp plans change?

Annually at most. Mid-year changes destroy trust disproportionately to whatever they fix, because a rep who reorganised their pipeline around the old plan is penalised for having done what you asked.

Compensation Plan

The document defining how a seller earns variable pay: what is measured, at what rates, with what accelerators, caps and clawbacks.

What it actually is

Not a payment schedule — an incentive contract. It determines not just how hard people work but which deals they pursue, when they close them, and what they neglect.

Structure moves behaviour independently of size

Accelerators, floors and period boundaries change what people do irrespective of how large the quota is. Research on nonlinear incentive contracts shows measurable seasonality in business activity attributable to contract shape rather than customer demand. Design the structure as deliberately as the number.

Components and what each provokes

Common elements

Element	Intended effect	What it also produces
Flat commission	Simple, predictable	Little pull above target
Accelerator above quota	Reward overperformance	Pull-forward, then push-out once target is safe
Floor before any payout	Guarantee a minimum standard	Abandonment of the period once the floor is unreachable
Multi-product kickers	Steer the mix	Attention split; the simplest product wins
Clawback on churn	Align with retention	Risk aversion, and disputes about attribution

Source: Mechanisms stated here

Testing a plan before you ship it

Model it against last year's actual deals. What would each rep have earned? Surprises here are design errors, not edge cases.
Ask a rep to compute their payout on a hypothetical deal. If they cannot, neither can they optimise for it.
Look for the perverse case: the deal a rep is better off not closing, or better off delaying. Every plan has one; you need to know where it is.
Check it against territory potential. A plan is only fair if the territories underneath it are comparable.

Measuring whether it works

Attainment distribution, not average. A wide spread usually indicates territory or quota design before it indicates talent.
Close dates by day of period. Concentration at the boundary measures the contract, not demand.
Mix against intent. If a kicker was meant to steer product mix, check whether mix actually moved.

RELATED TERMS

Quota Attainment

The proportion of quota a rep achieved, and across a team, the distribution of that proportion.

SPIFF

A short-term incentive layered on top of the compensation plan to direct attention at something specific.

Quota

The revenue target assigned to an individual rep for a period. The most consequential number in a seller's working life, and the mechanism translating a company target into individual accountability.

On-Target Earnings (OTE)

Total expected compensation when a seller hits quota exactly: base salary plus target variable.

COMMON QUESTIONS

What makes a good sales compensation plan?: One a rep can compute in their head, that pays for outcomes they control, and whose structure does not reward timing games. Simplicity is not a nicety — a plan nobody can calculate cannot motivate the behaviour it was designed for.
How many components should a comp plan have?: Few. Every additional accelerator, kicker and multiplier divides attention and adds a place to game. If a rep cannot state what their next deal is worth to them, the plan has too many parts.
Do accelerators work?: They increase effort above target and they also produce pull-forward and push-out around the boundary. Both effects are real; the question is whether the extra output is worth the timing distortion, which you can measure in your own close-date distribution.
How often should comp plans change?: Annually at most. Mid-year changes destroy trust disproportionately to whatever they fix, because a rep who reorganised their pipeline around the old plan is penalised for having done what you asked.