A CEO of a 40-person B2B software company sent me his commission plan last quarter and asked one question: "Why are my two worst reps making almost as much as my best one?"
I opened the spreadsheet. Everyone earned a flat 10% on closed revenue, no quota, no tiers, no floor. So a rep who scraped together $180K in a year got paid 10%. The rep who closed $620K also got paid 10%. On a per-dollar basis they were treated identically, which meant the plan quietly told the top rep that pushing harder bought them nothing extra. Within six months she left for a competitor that paid accelerators above quota.
That is what a bad commission structure costs you. Not just the payout math. The person.
Most articles about "sales commission structure" hand you a list of nine models and a template, then wish you luck. That is the wrong altitude. The structure is a set of decisions that have to fit the role, the deal size, and the data you actually have in your CRM. Get the structure right and reps behave the way you want without being told. Get it wrong and no amount of coaching fixes it, because you are paying people to do the thing you keep asking them to stop doing.
Here is how I think about commission structure after building comp plans for teams from their first AE up to 30-rep sales orgs.
What a commission structure actually is
Before models, four terms. If these are fuzzy, the plan will be too.
On-target earnings (OTE) is the total a rep makes at 100% of quota. It is base salary plus target variable. When a job post says "$160K OTE, 50/50," it means $80K base and $80K variable if they hit plan.
Pay mix is the split between base and variable. A 50/50 mix means half the money is guaranteed and half is at risk. A 70/30 mix means most of the pay is guaranteed.
Quota is the number a rep is expected to hit. Commission is almost always measured against it.
Commission rate is the percentage of a deal the rep earns. The median in B2B SaaS is 11.5% of annual contract value at 100% of quota, based on benchmark data across 170-plus companies.
Median commission rate at 100% of quota in B2B SaaS. SMB rates run higher (10 to 15% of ACV), enterprise lower (5 to 8%), because bigger deals carry bigger contract values.
The structure is the way you wire these four things together. Everything below is a decision about one of them.
The pay mix comes first, and it depends on control
The single most useful question when designing a structure: how much control does this rep have over the outcome?
The more a rep personally drives the deal, the more of their pay should be variable. A closing AE who runs the whole cycle can carry a 50/50 mix, because they own the number. An SDR who books meetings but does not close carries more base, usually 65/35, because a booked meeting turning into revenue depends on the AE, not them. A customer success manager who influences renewals but does not run the sales motion sits closer to 70/30 or 80/20.
Pay mix is not a reward. It is a statement about who owns the result. When I see a 50/50 mix on a role that has almost no control over closing, the plan is punishing people for a number they cannot move, and they know it.
Those are 2026 numbers. SMB AEs on sub-$25K deals land around $110K to $150K OTE. Mid-market AEs on $25K to $100K deals run $140K to $200K. Enterprise reps on six-figure deals reach $220K to $320K. The pattern holds: bigger deals, bigger OTE, and a mix that tilts toward variable as the rep's ownership of the number goes up.
The models, and when each one fits
Now the structures themselves. There are a lot of named models floating around, but for a B2B software team you really pick from five.
Flat-rate commission
Every dollar of closed revenue pays the same percentage. No quota, no tiers. This is what the CEO from the intro had.
It is simple to run and easy for reps to understand, which is why founders reach for it first. The problem is the one I opened with: it pays your best and worst reps at the same rate per dollar, so it gives your top performers no reason to stretch and gives weak performers no floor to clear. Fine for your first one or two reps when you have no quota data yet. Replace it the moment you can set a real number.
Quota-based with tiers and accelerators
Commission is measured against a quota, and the rate steps up as the rep crosses attainment thresholds. This is the most common structure on B2B and SaaS teams for a reason: it balances cost predictability for you with real upside for the people who overperform.
A clean version: 10% commission up to 100% of quota, then 15% on every dollar above 100%. That 1.5x accelerator is roughly standard. Some teams add a second step, paying 20% above 150%, to keep the top of the range hungry. About 82% of SaaS startups use accelerators of some kind.
This is the structure I put most closing roles on.
Draw against commission
The rep gets a guaranteed payment (the draw) that commission is measured against. Payout is whichever is higher, the draw or what they actually earned.
A recoverable draw means shortfalls carry forward and the rep pays it back out of future commission. A non-recoverable draw means the shortfall is forgiven. Non-recoverable draws are how you protect a new hire's income while they ramp without gutting their motivation. More on ramp below.
Gross-margin commission
The rep earns a percentage of profit, not revenue. This matters when reps can discount, because a revenue-only plan pays them the same whether they hold the line on price or give away 30% to close faster. Tie commission to margin and the rep suddenly cares about the discount, because it is their money too. If your reps have pricing latitude, this is worth the extra plumbing.
Base plus commission (the hybrid everyone actually runs)
Almost every real B2B plan is a base salary plus one of the commission models above. Pure-commission roles exist in some industries, but in software the base-plus-variable hybrid is the default because it gives reps stability and gives you a variable lever tied to results.
Accelerators and decelerators: the part most teams skip
Accelerators get all the attention. The rate climbs above quota, reps chase the upper tier, everyone is happy. What almost nobody builds is the other half.
A decelerator lowers the rate below a floor. If a rep only hits 40% of quota, do you really want to pay full freight on that revenue? Some plans pay nothing below 50% attainment, or a reduced rate, so the money concentrates on people clearing a real bar. This is uncomfortable to design and reps hate hearing about it, but without it a flat rate quietly funds underperformance for years.
The rule I use: your best rep should earn meaningfully more per dollar than your average rep, and your average rep should earn meaningfully more than someone who missed badly. If the plan does not create that spread, it is not a performance plan, it is payroll with extra steps. Setting the quota those tiers hang off is its own project, and I have written the full method in sales quota setting.
Ramp, draws, and the first 90 days
New reps cannot hit full quota on day one. A structure that expects them to just burns good hires before they produce.
The fix is a ramp. Two levers move together. First, a non-recoverable draw or guaranteed commission for the first three to six months, so the rep has income while their pipeline builds. Second, a tiered quota during that window: start around 25% of full quota in month one and step up to 100% by the end of the ramp. The exact curve depends on your sales cycle. If deals take 90 days to close, a rep cannot show closed revenue in month two no matter how good they are, so the ramp has to respect the cycle length.
I treat the comp ramp and the onboarding plan as one thing. The 30/60/90 sales onboarding plan sets what "ramped" means in terms of activity and skill. The comp ramp just pays for that same curve. When the two disagree, reps optimize for the money, so make them agree.
Clawbacks: the clause that keeps deals clean
Here is the failure mode nobody wants to talk about. A rep is 4% short of quota with two days left. They find a bad-fit prospect, discount hard, promise the moon, and close it. They hit quota, take the accelerated commission, and move on. Ninety days later the customer churns because they were never a fit, and you paid a premium to acquire a logo that cost you money.
A clawback reclaims commission if a customer churns inside a defined window, usually 90 to 120 days. About 53% of SaaS companies include one. It is not about clawing money back for its own sake. It is about killing the incentive to close deals that should not close. When a rep knows a 60-day churn erases their commission, they stop pushing bad deals over the line, and your net revenue retention thanks you for it.
Clawbacks only work if your CRM and billing data agree on what "churn" means and when it happened. If those two systems disagree, the clause becomes a fight every single time, which is why comp design keeps dragging you back to data quality. I wrote about that mess in CRM data quality.
Reps do exactly what the plan pays them to do, not what you tell them in the meeting.
If reps are sandbagging, discounting, or closing junk, read the structure before you blame the people. The behavior is almost always a rational response to how the money is wired.
How to actually build the structure
A commission structure is not a template you download. It is four decisions, made in order, grounded in your own numbers.
Start with OTE and pay mix per role, anchored to the benchmarks above and to what the role actually controls. Then set the quota, because every rate below hangs off it. Then pick the model: flat rate only if you have no quota data yet, otherwise quota-based with accelerators for closers, with a draw layered on for ramping reps. Then add the guardrails, the accelerator, the decelerator or floor, and the clawback window.
The part that breaks in practice is never the model. It is whether your CRM can calculate the plan without a human rebuilding a spreadsheet every month. A structure you cannot pay accurately and on time is worse than a simpler one you can, because a wrong commission check destroys trust faster than almost anything else a RevOps team does. When the spreadsheet starts breaking, that is the signal to look at commission software. And if you want the deeper method behind a plan reps actually believe, I laid it out in sales compensation plans reps trust.
The structure and the plumbing under it are one job. That is the part we build for teams: the CRM and RevOps foundation that lets a comp plan pay out cleanly, and the automation that calculates it without a person babysitting a spreadsheet every cycle.
Not sure your comp plan is paying for the right behavior?
Book a free 30-minute audit. We'll look at your structure, your quota data, and whether your CRM can actually pay it, then show you the two fixes we would make first.
Book an audit →FAQ
What is the most common sales commission structure in B2B SaaS?
A base salary plus quota-based commission with accelerators. The rep earns a set percentage up to 100% of quota, then a higher rate above it. The median rate at 100% quota is about 11.5% of ACV, and roughly 82% of SaaS startups add accelerators above plan. Flat-rate structures show up on very early teams that do not yet have quota data, but they get replaced quickly.
What is a good base-to-variable split for a sales rep?
It depends on how much the role controls the outcome. Closing AEs usually run 50/50, because they own the deal. SDRs who book but do not close sit around 65/35, since revenue depends partly on the AE. Roles that influence but do not drive sales, like customer success, run closer to 70/30 or 80/20. The more control, the more variable.
How do accelerators work in a commission plan?
An accelerator raises the commission rate once a rep passes an attainment threshold. A common setup pays 10% up to 100% of quota and 15% on everything above it, a 1.5x bump. Some plans add a second step above 150%. The point is to make overperformance pay more per dollar so top reps have a reason to keep pushing past plan.
Should I use a draw for new sales reps?
Yes, during ramp. A non-recoverable draw gives a new hire guaranteed income for the first three to six months while their pipeline builds, and you pair it with a tiered quota that starts low and climbs to full by the end of ramp. Recoverable draws, where the rep pays shortfalls back, are riskier for morale and I use them sparingly.
What is a commission clawback and do I need one?
A clawback reclaims commission if a customer churns inside a set window, usually 90 to 120 days. About 53% of SaaS companies use one. It removes the incentive for a rep to close a bad-fit deal just to hit quota, since a fast churn erases the payout. If your reps can discount or push deals over the line at quarter-end, a clawback is worth building, but only if your CRM and billing data agree on what churn means.
Build a structure that pays for the right behavior
A commission structure is the clearest instruction you give your sales team, and most of the time it is telling them something you did not mean to say. Fix the structure and the behavior follows.
If your plan is overpaying laggards, capping your best reps, or living in a spreadsheet nobody trusts, talk to us. We build the comp structure and the CRM plumbing that pays it out clean, so the number on the check matches the plan you designed.