Sales commission structures, with the formulas

A commission structure is the rule that turns a rep's results into money. That makes it the strongest instruction a company gives its sales team — stronger than the pitch deck, the training or the weekly meeting, because it is the only one attached to their mortgage. Reps read the plan and optimise it. If the plan pays the same on a renewal and a new logo, you will get renewals. If it pays on revenue and ignores discount, you will get discounts.

So the question is never "what is the standard structure". It is: which behaviour do we need more of next year, and which shape pays for exactly that. The seven below cover almost everything in use. Each one is given with its arithmetic and the specific way it goes wrong, because every structure has one.

Scoring works the same way, which is why the scoring model your gamification programme runs on has to be designed with the same care as the comp plan — a board that ranks the wrong number does quiet damage all year.

1. Straight commission

No base salary. The rep earns a percentage of what they sell and nothing otherwise.

Formula: commission = revenue × rate

At a 15% rate on $40,000 closed in a month, the rep earns $6,000. Close nothing, earn nothing.

Where it fits: short cycles, high volume, independent contractors, and businesses where a rep can genuinely produce inside their first month.

How it fails: it selects for people who can survive a bad quarter financially, which is a filter on personal wealth rather than on selling ability. It also makes long cycles unstaffable — nobody works a nine-month enterprise deal with no income.

2. Base plus commission

A fixed salary plus a percentage of sales. The most common structure in B2B.

Formula: pay = base + (revenue × rate)

The split is usually expressed as a ratio of on-target earnings. A 60/40 split on $120,000 on-target earnings means $72,000 base and $48,000 of variable pay at 100% attainment. If quota is $600,000, the implied rate is 8%.

Where it fits: almost everywhere with a sales cycle longer than a few weeks.

How it fails: the split gets copied from another company without checking the cycle length. A rep on a nine-month cycle needs a heavier base than one on a three-week cycle, or they leave in month four.

3. Tiered commission

The rate rises as cumulative attainment rises. Distinct from an accelerator in that tiers can sit below 100% too.

Where it fits: teams with a wide performance spread, where you want the top quartile to have somewhere to go.

How it fails: cliffs. If a rep can see they will not reach the next band, the current band's rate is all they are working for, and the last two weeks of the period go quiet. Marginal rates fix this better than step changes — pay the higher rate only on the portion above each threshold, never retroactively on the whole amount, or the cost curve becomes unmanageable.

4. Accelerators above quota

A raised rate that applies only above 100% of quota. Covered in full in the commission accelerator definition, summarised here because it composes with every other structure on this page.

Formula: commission = (quota × base rate) + (excess × accelerated rate)

A rep at 118% of a $500,000 quota, on 8% to quota and 14% above: ($500,000 × 8%) + ($90,000 × 14%) = $40,000 + $12,600 = $52,600.

Where it fits: whenever you would rather have a few reps massively overperform than everyone land at 101%.

How it fails: sandbagging. A rep who has cleared quota in month two of the quarter has an incentive to hold deals until the next period starts if the accelerator resets. Rolling annual attainment removes the incentive entirely.

5. Gross-margin commission

Pays on margin rather than revenue, so discounting costs the rep directly.

Formula: commission = (revenue − cost of goods) × rate

On a $60,000 deal at 55% margin with a 20% rate: $33,000 × 20% = $6,600. Discount that deal by 10% and the margin falls to roughly $27,000, so the commission falls to $5,400 — a 10% discount costs the rep 18% of their commission. That asymmetry is the whole point.

Where it fits: businesses where reps control pricing, and any business bleeding margin through discretionary discounts.

How it fails: it needs margin data the rep can see before they quote. If they cannot calculate their own commission at the point of negotiation, the structure teaches nothing.

6. Draw against commission

An advance against future commission. A recoverable draw is repaid from later earnings; a non-recoverable draw is effectively a guaranteed floor.

A rep on a $4,000 monthly recoverable draw who earns $2,500 in commission carries $1,500 forward as a balance. Earn $6,000 the following month and $1,500 is recovered, leaving $4,500.

Where it fits: ramping new hires, and seasonal businesses with predictable troughs.

How it fails: an unbounded recoverable draw. A rep who ramps slowly accumulates a debt they cannot clear, which is demoralising long before it is repaid, and it is usually written off anyway. Cap the recovery period at two quarters and make the cap explicit in the plan. The same problem is better solved at the target: see ramp quota.

7. Team or pooled commission

A pool is funded on collective performance and split by a stated rule — evenly, by role weighting, or by individual contribution.

Where it fits: genuine multi-threaded selling where attribution is contested, and pods where an SDR, an AE and a solutions engineer all touch the same deal.

How it fails: free-riding, and it is the failure everyone predicts. Two things contain it: keep pools small enough that individual effort is visible to peers, and pair the pool with individual activity recognition so contribution stays legible even when pay does not track it. A live leaderboard on the activities each person controls does that job without splitting the pool differently.

Choosing between them

Four questions settle most of it.

How long is the cycle? Longer cycles need a heavier base. Below about a month, straight or lightly-based structures work.

Does the rep control price? If yes, pay on margin. If no, paying on margin punishes them for a decision finance made.

How wide is the performance spread? Wide spreads justify tiers and accelerators. Narrow ones do not need the complexity.

What is the one behaviour you need more of? Whatever it is, some line in the plan should pay more for it than for the alternative. If no line does, the plan is not asking for it.

The test every plan has to pass

Can a rep calculate their own commission on a live deal, in their head, at the negotiating table?

If not, the plan cannot influence behaviour, because behaviour changes at the moment of the decision and not when the statement arrives six weeks later. Plans fail this test by stacking too many modifiers: a tier, a margin multiplier, a product kicker, a quarterly bonus and an annual accelerator, all interacting. Every additional term halves the number of reps who can predict their own pay.

Two or three moving parts is usually the ceiling. If the plan needs more, that is a sign the strategy is asking for too many things at once — which is a strategy problem the compensation plan cannot solve.

Where gamification sits alongside the plan

Commission pays for outcomes, which arrive late and are partly outside a rep's control. A scoring programme gives feedback on the work that produces those outcomes, which arrives the same day and is almost entirely inside their control. They are complementary instruments and they fail in opposite directions — commission is slow to change behaviour, points programmes have nothing at stake.

The practical combination is a stable comp plan doing the heavy lifting on outcomes, and a scored programme on leading activity giving the weekly signal. Which KPIs are worth putting on a board is the prior question, and it has the same discipline behind it: if a rep can move the number without doing more real selling, you are paying for the workaround.

FAQ

What is the most common sales commission structure?

Base salary plus commission, usually expressed as a split of on-target earnings such as 60/40 or 50/50. It is the default in B2B because it funds the rep through a sales cycle longer than a payroll cycle while still tying a substantial share of pay to results. Straight commission remains common in short-cycle, high-volume selling and among contractors.

What is a good commission rate?

There is no cross-industry number worth quoting, because the rate is a function of quota, on-target earnings and gross margin rather than a standalone figure. Work it backwards: decide on-target earnings for the role, decide the base-to-variable split, then divide the variable portion by the quota. That quotient is your rate. A rate copied from another company with a different quota means something different.

What is the difference between a tiered commission and an accelerator?

An accelerator is a raised rate that applies specifically above 100% of quota. Tiers can sit anywhere on the attainment curve, including below quota, and there may be several. In practice most plans use tiers to shape the approach to quota and a single accelerator to reward overperformance beyond it.

Should commission be paid on revenue or on gross margin?

Pay on margin when reps control pricing, because it makes discounting cost them directly and is the most reliable way to protect margin. Pay on revenue when pricing is fixed centrally, since margin-based pay would otherwise penalise reps for decisions made elsewhere. The deciding question is whether the rep can see and influence the margin at the point of quoting.

How do you avoid sandbagging at the end of a period?

Sandbagging is caused by attainment resetting, which makes a deal worth more next period than this one. Measuring attainment on a rolling annual basis removes the incentive outright. Where the period has to reset, avoid retroactive rate changes at thresholds and keep accelerators marginal, so no single deal is worth dramatically more on one side of a date than the other.