A consumption deal closes at one number and bills at another. The commitment says $30,000, the first quarter bills $52,000, and by month nine the account is running at $120,000 annualised. Every one of those is a defensible basis for commission, and they pay the rep very differently. Most consumption comp plans never say which one they mean.
Commission needs a number, and consumption revenue does not hold still long enough to give you one. So the plan has to name the moment revenue becomes real.
At close. Pay on the committed or estimated amount the day the deal is won. Simple, fast, and familiar to anyone who has run a seat-based plan. It also pays the same commission whether the customer goes on to consume nothing or triple the commitment.
At first full billing period. Wait until one complete cycle has billed, then pay on that. More accurate, delays the rep's cash by a month or a quarter, and still misses everything that happens afterwards.
Continuously. Recalculate as usage lands, so the commission tracks the account. Most accurate, and the only option that pays growth. It also means a paid commission can change.
Take one customer, closed in January on a $30,000 commitment and consuming steadily above it all year. Hold the rate constant at 8 percent and change only the basis the rate is applied to, across the first twelve months. The three answers are all defensible and none of them are close.
| Basis | Revenue used | Commission | What the rep learns |
|---|---|---|---|
| At close | $30,000 committed | $2,400 | Land the logo, move on |
| First billing period | $52,000 annualised | $4,160 | Land it and get it live |
| Continuous | $120,000 actual | $9,600 | Land it, get it live, grow it |
The spread is not a rounding difference. Paying at close values this customer at a quarter of what continuous valuation does, and it pays exactly the same whether the account grows or stalls.
That is the real argument for continuous recalculation, and it is a behavioural one rather than an accounting one. If expansion is what the business needs and the plan pays at close, the plan is buying logos.
Once you are paying on actual consumption, the next question is what selects the rate. The instinct is to use quota attainment, because that is how most plans work. For consumption it is usually the wrong driver.
Attainment tiers make the same customer worth different amounts to different reps. A $200,000 account lands in a high band for a rep at 130 percent and a low band for a rep at 60 percent, so the payout says more about the rep's quarter than about the account. In a motion where the goal is growing accounts, that noise works against you.
Value tiers read the consumption number on the account itself. A band is a band regardless of who owns it, so growing an account from $75,000 to $150,000 pays the same to anyone. Combined with marginal bands, each slice of consumption pays its own rate and the incentive rises smoothly with account size instead of stepping at a quota line.
Continuous recalculation is the right answer for most consumption businesses and it introduces three problems that per-deal plans never have.
Usage falls, and so does a commission you already paid. A customer that drops from $120,000 to $70,000 takes the rep's earned commission down with it. This is arithmetically correct and it is the single most disputed thing in consumption comp. Decide up front whether adjustments are capped, whether a floor protects previously paid amounts, and how far back a correction can reach, then write it in the plan rather than settling it per case.
The rep's number changes without anyone touching it. In a seat-based plan a commission only moves when a human edits something. Here it moves because a customer ran more queries. Statements have to show which usage figure produced which payout and when it was last recalculated, or every conversation starts from zero trust.
Month-end stops being a close and becomes a snapshot. There is no moment when consumption is final, so the period close is a decision about which usage reading you are paying on. Pick a cutoff, apply it consistently, and expect the following period to carry adjustments from the previous one.
One rule does this. It measures the consumption revenue field rather than a closed-won amount, tiers by the value of that field instead of by attainment, and uses marginal bands so each slice of consumption pays its own rate. Monthly attainment keeps the recalculation window short enough that corrections stay small.
Because the rule reads a field rather than a snapshot, a change to consumption re-runs the same calculation and updates the commission in place, with the previous value kept in history. Run it against a period you have already paid before switching anything over, so you can see how large the corrections would have been on real accounts.
This plan has one rule.
Pays a share of consumption revenue, with the rate chosen by the deal's consumption revenue. Runs on every closed deal, as it closes.
| Deal value | Pays |
|---|---|
| $0 to $25,000 | 6% of consumption revenue |
| $25,000 to $75,000 | 9% of consumption revenue |
| $75,000 and above | 13% of consumption revenue |
Each band pays its own rate on the portion that falls inside it.
This is the whole plan. Send it to POST /v1/plans and it is live; change a rate and re-send it to ship a revision.
curl -X POST https://api.compcode.ai/v1/plans \
-H "Authorization: Bearer $COMPCODE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"name": "Consumption AE 2026",
"effectiveStart": "2026-01-01",
"config": {
"rules": [
{
"name": "Usage Revenue",
"measure": "consumption_revenue",
"executionPhase": "per_deal",
"attainmentPeriod": "monthly",
"tierBy": "value",
"tierValueField": "consumption_revenue",
"tierMode": "marginal",
"tiers": [
{
"minThreshold": 0,
"rate": 0.06
},
{
"minThreshold": 25000,
"rate": 0.09
},
{
"minThreshold": 75000,
"rate": 0.13
}
]
}
]
}
}'Connect your CRM, create the plan with one call, and simulate before anyone gets paid.
Start free →