Sales compensation and territory

A compensation plan is the clearest statement a company makes about what it actually wants. Reps will optimize it precisely, including in the ways you did not intend, and that is not a character flaw — it is the plan working.

Size territories on potential, not on the current book

A territory built from existing accounts rewards whoever inherited the good one and guarantees the argument that follows. Build from addressable potential — accounts that fit, weighted by how many of them could plausibly buy — then check that each territory can support the quota you intend to put on it.

A quota a territory mathematically cannot carry is not a stretch goal, it is a resignation on a delay. Do that arithmetic before assigning, and expect to find one or two territories that fail it.

Rebalance on a known cadence, so it is a process rather than a punishment, and account for in-flight pipeline when accounts move.

Set quotas people hit often enough to believe

If most of the team misses, the plan has stopped motivating and started demoralizing; the number is noise and everyone knows it. If nearly everyone clears easily, you are paying above market for average performance.

A distribution where a solid majority reach target, with real upside above it, keeps the plan credible. Build the aggregate from the territories rather than dividing the company number by headcount — the second is how you end up with quotas nobody accepted.

Account for ramp explicitly for new hires rather than pretending a first quarter is a normal one.

Pay on the outcome you actually want, and keep it simple

• What you pay on should be the thing you want more of. Paying purely on new bookings gets you new bookings, including bad ones that churn.
• Consider what protects quality: margin or discount-adjusted credit, a clawback window on early churn, or paying on collected revenue rather than signed.
• Three components is usually the limit. A plan a rep cannot compute in their head does not change behavior, because they cannot see which action pays. Complexity in a comp plan is a design failure, not sophistication.

Accelerators above target are worth the money — that is where discretionary effort lives. Decelerators below target usually just accelerate the departure of someone already struggling.

Draws, caps and clawbacks

A recoverable draw supports ramp but creates debt; make the recovery schedule explicit or it becomes a surprise. Caps on upside are almost always a mistake — you are punishing the outcome you paid to create, and the rep will simply move the deal into the next period, which is worse for forecasting than the extra commission.

Clawbacks for early churn or non-payment are defensible when the window is short and stated in advance. Applied retroactively they cost more trust than they recover in money.

Write down how disputes get settled

Split credit, who owns an account that moved territories, what happens when a deal slips across a period boundary, how a commission on a modified contract is computed. Decide these before they occur and publish them — every one of these arguments is worse when the answer is invented under pressure by someone with an interest in the outcome.

Change plans with notice, and honor what was already earned

Mid-period changes are occasionally necessary and always expensive in trust. Give notice, explain the reason, and never restate a commission already earned under the prior plan. A team that believes the plan can be changed retroactively stops treating it as an incentive.

Sources

references/sources.md in this skill lists the outside authorities that settle the questions here — what each one is authoritative for, and what you may do with it. Check them before answering on anything they cover, and cite what you used. Most are free to read and not free to reproduce; the use note on each is binding.

Tooling

Commission calculation: CaptivateIQ, Spiff, Everstage, Xactly, and similar. The threshold for buying is not headcount, it is the point at which a spreadsheet error would go unnoticed — which arrives earlier than most teams expect.

Territory and quota planning: Fullcast, Salesforce Maps, Varicent, and similar; a well-built model covers a single-segment team.

Whatever calculates it, reps need a statement they can check themselves. Disputes are expensive in trust long before they are expensive in money.

Never

• Assign a quota to a territory that cannot mathematically support it.
• Add a component to a plan that a rep cannot compute without a spreadsheet.
• Cap upside on the performance you designed the plan to produce.
• Change terms retroactively for a period already worked.