Lukas Blazek via Pexels
Every new VP Sales joining a UK scale-up gets asked the same question in their first week, usually by the CFO: can we put a cap on commission? And every recruiter reciting the company line on Glassdoor has the same answer ready: we're uncapped, of course. Both sides are usually wrong for the same reason — they're treating the cap-vs-uncapped decision as ideology when it should be plan design.
This piece is the contrarian take. Caps are usually the wrong tool. But there are three specific scenarios where a ceiling is the right answer, and pretending otherwise costs UK sales orgs real money. We'll name them, show the £ figures, and explain how to design a cap that doesn't kill motivation.
TL;DR: Should UK sales commission be capped?
For a normally-distributed quota-carrying rep selling your core product, no — caps are a tax on your best people and a signal that your plan was modelled wrong. But caps are defensible in three narrow situations: (1) channel or partner-sourced deals where the rep didn't do the selling, (2) genuine one-off "whale" deals far outside the plan's design assumptions, and (3) inbound, marketing-sourced gimmes where the company, not the rep, generated the pipeline. The right move is almost never a hard cap on total earnings — it's a per-deal mechanism (decelerators, split rates, deal-size tiers) that targets the specific economics you're worried about.
Why the "always uncapped" orthodoxy is half-right
The SaaS playbook says: never cap commission. The argument runs that capping signals distrust, demotivates A-players, and pushes top performers to your competitors. There is real evidence behind that — if a rep can see exactly how much more revenue they could close and exactly how little extra they'd be paid for it, they'll either coast or leave. Behavioural research on sales compensation has been consistent on this point for decades.
But "never cap" is shorthand for a more nuanced reality. It applies cleanly to a rep selling a defined product, into a defined territory, against a quota the company modelled honestly. When those conditions hold, the rep generating 200% of plan is doing roughly twice as much of the work the company was paying for. Paying them twice the variable comp is fair, and it's what the plan promised.
The orthodoxy goes wrong when any of those conditions break. A deal lands that the rep barely touched. A whale closes that was ten times the size the plan was designed around. An inbound lead converts on the second call because marketing did six months of nurturing first. In those cases the rep's payout no longer tracks their effort or skill — and paying full commission on it isn't "rewarding A-players", it's bad plan design hiding behind a recruiting slogan.
The finance reality: why CFOs ask about caps after the first big quarter
Finance teams almost never ask for caps in Q1. They ask in Q3, after one rep blew through 250% of quota on a deal nobody saw coming, and the commission accrual line item ate the variance buffer. The CFO's instinct — cap everything — is wrong. But the underlying concern is right: the plan didn't model for this, and we're now paying out a number we never budgeted.
There's a UK-specific layer here that US-imported plan templates ignore. Commission is treated as earnings for both Income Tax and National Insurance purposes — the HMRC Employment Income Manual at EIM00520 lists commissions explicitly alongside salaries, fees, wages and bonuses as earnings subject to PAYE and NICs. That means every £1 of commission paid also triggers employer Class 1 secondary NICs on top.
For the 2026/27 tax year, HMRC's rates and thresholds for employers confirm the secondary (employer) NIC rate sits at 15%, with the Secondary Threshold at £96 a week (£5,000 annually) — meaning essentially every £ of commission paid to a quota-carrying rep is fully exposed to that 15% on-cost. So a £40,000 surprise commission payout isn't a £40k variance to the model; it's £46k by the time you've paid HMRC. That's the number CFOs are reacting to.
What does a commission cap actually mean?
A commission cap is any plan mechanic that limits how much variable pay a rep can earn, either in total or above a defined performance level. In practice it shows up in three forms, and they are not equivalent:
| Cap type | How it works | What it signals |
|---|---|---|
| Hard annual cap | Rep earns nothing on revenue closed above a fixed £ ceiling | "We don't trust our own plan." Pushes top reps to coast or leave. |
| Performance cap | Variable pay flattens above, say, 200% of quota | Slightly softer; still kills the back half of the year for over-performers |
| Per-deal cap or decelerator | Commission rate drops on deal value above a threshold, or on specific deal types | Targets the actual economic problem without punishing volume |
When people argue "caps demotivate", they're almost always arguing against the first row. The third row is a different animal — it's not really a cap, it's selective rate-shaping. And it's where most UK sales orgs should be playing. We've written about the mechanics in more detail in our piece on designing commission decelerators.
The three scenarios where a UK commission cap is defensible
1. Channel and partner-sourced deals
If a deal originates with a channel partner, a reseller, or a referral programme, the AE's role is fundamentally different. They're typically managing the relationship and closing paperwork rather than prospecting, qualifying and running discovery from cold. Paying the full direct-sale commission rate on partner-sourced revenue overpays the rep relative to the work they did — and it usually means you're paying twice, because the partner is taking a margin too.
Worked example: take a £45k base / £15k OTE AE on a 70/30 split, where the £15k variable is earned at 100% of a £600k annual quota. That's a 2.5% effective commission rate on direct-closed ARR. If a channel partner brings a £200k deal that the rep helps close in three meetings, the default plan pays the rep £5,000 on it. A capped channel rate of, say, 1% — £2,000 on the same deal — recognises the rep contributed but didn't source it. The £3,000 difference, grossed up for the 15% employer NIC on-cost, is about £3,450 of margin recovered per deal. Across a year of partner-sourced revenue, that's real money.
The motivation question matters here. A 1% channel rate is still positive economics for the rep — they should want the deal. The cap isn't "you get nothing"; it's "you get paid in proportion to what you actually did".
2. Genuine one-off whales
Most commission plans are designed around an assumed deal-size distribution. A team selling mid-market SaaS with ACVs of £20k–£80k builds quota and rates around that range. When somebody lands a £750k three-year deal — ten times the band the plan was modelled for — the rep's commission can balloon into territory that has nothing to do with skill multiples.
This is where a per-deal decelerator earns its keep. Above a defined deal-value threshold (say 3x median ACV), the commission rate drops or steps. The rep still gets paid handsomely — they should — but the payout tracks reality rather than a linear extrapolation of a plan built for different physics.
A whale closing isn't a 10x performance event. It's a 10x deal-size event. Paying a 10x payout to it treats those as the same thing — they're not.
Worked example with the same £45k/£15k AE. On a £750k closed deal at the flat 2.5% rate, gross commission is £18,750 — more than the entire OTE variable on a single deal. A stepped structure that pays the standard 2.5% on the first £150k of deal value, 1.5% on £150k–£400k, and 1% above £400k pays the rep: (150k × 2.5%) + (250k × 1.5%) + (350k × 1%) = £3,750 + £3,750 + £3,500 = £11,000. The rep still gets a life-changing single-deal payout. The company keeps ~£7,750 of variance, plus another ~£1,160 of employer NIC on-cost it didn't have to pay. Both sides are fine.
The key plan-design point: the threshold has to be set off the plan's original modelling assumptions, not retroactively after a deal you didn't like. Apply it consistently from the start of the year and reps will accept it. Apply it after the fact and you'll lose them.
3. Marketing-sourced gimmes
The most contentious of the three. If your inbound engine produces leads that close on the second call — because marketing did the awareness work, the demand-gen team booked the meeting, and the product-led signup already showed intent — paying full new-business commission on them rewards the rep for being available rather than for selling.
This is awkward territory because it can read as anti-rep. Reps need to convert inbound, and good inbound conversion is a real skill. But the gap between "converted a warm inbound" and "sourced and closed a cold outbound" is enormous, and a single flat rate ignores it.
A defensible approach is a split rate by lead source: outbound-sourced deals at the full rate, marketing-sourced deals at, say, 60–75% of the rate. Some UK orgs run this as a quota carve-out instead — a portion of the rep's number can only be hit with outbound-sourced revenue. Both achieve the same thing: paying for the work, not just for the close.
The critical implementation detail is having a clean, agreed source-of-truth for lead origin. If marketing and sales argue about whether a deal was inbound or outbound for the purpose of commission, you've created exactly the kind of dispute that erodes trust in the plan. This is one of the strongest arguments for moving off spreadsheets — see our piece on why commission spreadsheets get expensive fast.
How do you design a commission cap without killing motivation?
The single biggest design mistake is the hard total-earnings cap. It tells your top rep, in writing, that there is a deal at which they should stop selling. Don't do this. Instead:
- Cap the input, not the outcome. Use per-deal rate adjustments (decelerators, lower rates by deal source, stepped tiers) rather than a hard ceiling on total commission. The rep can still earn unbounded amounts by closing many deals — they just can't earn unbounded amounts on a single anomaly.
- Set thresholds off plan design assumptions, not last year's outliers. If your median ACV is £50k, a deal-size decelerator at 3x median (£150k) is defensible. A decelerator at "the size of the deal Sarah closed last September that broke the budget" is not.
- Publish the rules at plan rollout, never mid-year. Reps will accept almost any reasonable structure if they knew about it in January. Retro caps are how you lose your top performers in February.
- Document the rationale for each carve-out in writing. "Channel deals pay 1% because the partner takes 8% margin and we're paying twice otherwise." Reps can argue with maths but not with logic.
- Run worked examples in the rollout deck. Show the rep what they'd earn on a £200k channel deal, a £750k whale, and a £40k marketing-sourced inbound under both the old plan and the new structure. Numbers settle most disputes.
For more on the full rollout mechanics, see our guide on commission plan rollout and change management.
When is uncapped commission genuinely the right choice in the UK?
Uncapped commission is the right answer when the conditions the orthodoxy assumes actually hold:
- The rep is selling a defined product into a defined territory.
- The quota was modelled honestly off realistic deal-size and conversion assumptions.
- Pipeline generation is meaningfully driven by the rep, not by inbound or partner sources.
- Your plan has accelerators above 100% so the marginal £ of effort still pays disproportionately, but not absurdly so.
- Finance has budgeted for the employer NIC on-cost (15% in 2026/27 per HMRC) on the full expected commission spend, not just the on-plan number.
If all of that is true, leave the cap off. The recruiting line is genuine, the maths works, and a 220% rep is earning every penny. We've covered the accelerator side of this in detail in our piece on designing commission accelerators.
Frequently Asked Questions
Is it legal to cap sales commission in the UK?
Yes. There is no UK statutory restriction on capping commission, provided the cap is clearly written into the employment contract or commission plan document and applied consistently. The risk is contractual and reputational, not regulatory. The bigger legal exposure for UK employers is around changing or removing earned commission mid-cycle, or applying caps retroactively without consent.
Do caps reduce employer National Insurance liability?
Indirectly, yes. Commission is earnings for NIC purposes per HMRC's EIM00520, and employer Class 1 secondary NICs apply at 15% for 2026/27. Lower commission payouts mean a lower employer NIC bill. But "reduce NIC" is a poor headline reason to cap — the right reason is plan-design integrity. The NIC saving is a secondary effect.
What's the difference between a commission cap and a decelerator?
A cap is a hard ceiling: above a defined point, the rep earns zero additional commission. A decelerator reduces the commission rate above a threshold but the rep continues to earn — just at a lower rate per £. Decelerators preserve motivation because the rep is still incentivised to close the next deal; caps do not.
Should sales commission caps apply to SDRs and BDRs the same way as AEs?
Usually no. SDR/BDR variable comp is typically smaller, more activity-driven, and less exposed to the deal-size anomalies that justify AE caps. A cap on SDR commission tends to read as petty rather than protective. If you have a runaway-payout problem at the SDR level, the issue is almost always quota-setting, not the absence of a cap.
How do I introduce a cap on a previously uncapped plan without losing reps?
Never do it mid-year. Announce structural plan changes at least one cycle before they take effect — ideally bundled with the annual plan rollout — and pair any new cap with a sweetener elsewhere (a better accelerator, a higher OTE base, a quota reduction). Document the rationale, publish worked examples, and offer 1:1 walkthroughs to top performers. The change-management work matters as much as the maths.
Ready to fix your commission process?
Join the early access list and be first to try Commit when we launch.
Get Early Access