A compensation plan states what the company will pay for. Sellers, who are paid on it, read it more carefully than the people who wrote it and optimise against what it actually says rather than what it meant. That gap is where the surprises live.
The useful way to review a plan is to stop asking whether it is fair and start asking what a rational person would do if they read only this document and wanted to maximise the number at the bottom.
The test to run before shipping a plan
Take the plan and answer, in writing, what it tells a seller to do in each of these situations. If two people give different answers, the plan is ambiguous and the ambiguity will be resolved in whichever direction pays more.
- A deal can close this quarter at a discount, or next quarter at full price. Almost every quarterly plan says take the discount, and then leadership is surprised by discounting.
- A small deal is easy and a large one is hard. A flat commission rate says do the small ones. An accelerator says otherwise. Whichever you chose, you chose.
- A renewal is at risk and a new logo is available. If new business pays and retention does not, the plan has told them which to abandon.
- The customer is a poor fit but will sign. If nothing claws back on early churn, the plan pays for the signature and is silent on the outcome.
None of these is a hypothetical. They are the four situations a seller meets most weeks, and the plan answers all four whether or not anyone intended it to.
The components, and what each one actually buys
| Component | What it buys | What it costs you |
|---|---|---|
| Flat commission | Simplicity. Everyone can calculate their own pay. | No steering. Every deal looks equally worth having. |
| Accelerators | Effort past quota, which is where the margin is. | Sandbagging into the next period once quota is safely hit. |
| Multipliers by segment | Direction. Sellers go where you point them. | Complexity, and sellers who cannot predict their own pay stop trusting the plan. |
| Clawbacks | Alignment with what happens after signature. | Real friction, and a genuine grievance if applied to churn the seller could not influence. |
The pattern is that every mechanism that adds steering also adds complexity, and complexity has a hard limit: a plan a seller cannot calculate in their head stops functioning as an incentive and becomes a monthly surprise.
The complexity ceiling
A useful rule is that a seller should be able to look at a deal and know roughly what it pays them, without opening a spreadsheet. Past that point each additional rule buys less steering than it costs in trust.
Teams arrive at three or more the same way every time: each rule was added to fix a real behaviour, one at a time, and nobody removed anything. The plan accretes. An annual review that only adds is how a two-line plan becomes a document requiring a calculator.
Quota is part of the plan, and it is usually the broken part
A well-designed plan on an unreachable quota is an unreachable plan, and sellers work that out within weeks. The behaviour that follows is rational and expensive: if the number cannot be hit, the incentive shifts to protecting a base salary and staying long enough to find something else.
The diagnostic is the distribution rather than the average. If nearly everyone lands just over quota, the number is set where it was meant to be. If most of the team is far under and two people are far over, the quota is not calibrated, it is a lottery, and the two winners will be read as proof the plan works.
When not to change the plan
Mid-year, unless it is broken in a way that is costing real money. Changing the specification partway through breaks every calculation a seller has made about their year, and the trust cost outlasts whatever the change fixed.
To fix a performance problem with one person. Plans are a blunt instrument aimed at everybody. A plan rewritten around one underperformer changes the incentives for the whole team to solve a management conversation.
Before you know what the current plan is actually rewarding. Run the four situations above against the existing plan first. Teams frequently discover the plan is working exactly as written and the problem is that what was written is not what was wanted.
That last one is the same failure as everywhere else in a go-to-market system: the rule is doing precisely what it says, and nobody has read it back since it was written. The same question decides whether a sales motion survives its author: is the intent written down anywhere somebody else can check it.
Questions people ask
How complex should a comp plan be?
Simple enough that a seller can estimate a deal's payout without a spreadsheet. Past that, extra rules buy less steering than they cost in trust.
Should commission be clawed back on churn?
It aligns the plan with what happens after signature, and it is a real grievance when applied to churn the seller could not influence. Scope it to the window they can affect.
What does sandbagging indicate?
Usually accelerators combined with a hard period boundary. Once quota is safe, the plan pays more for a deal that lands next period, so it lands next period.
How do I know the quota is right?
Look at the distribution, not the average. Most of the team landing just over target means it is calibrated. A handful far over and the rest far under means it is not.
Can a plan fix a discounting problem?
Sometimes, and check first whether the plan is causing it. A quarterly target with no price protection instructs sellers to discount to close in period, which is a specification problem rather than a discipline one.
Or skip the search
Want the person who owns it after it ships?
If the work is a standing obligation rather than a bounded project, tell me which system you want built first. You get example profiles and a call with your pick before anything starts.
Placement runs through AYTalent. Engineers are employed and trained on the bench before they start with you.