Comp Plans Should Shape Behavior, Not Create It

Every sales comp plan is a little manifesto.

It tells the team what matters. It tells the team what to ignore. It tells the team what the company actually values once the Slack messages and all-hands speeches are over.

That is why bad comp plans create weird behavior so fast.

If you pay for meetings, people chase meetings. If you pay for revenue at all costs, people drag bad-fit deals across the line. If you make the rules too complex, nobody trusts the math. If you reward individual heroics only, collaboration quietly dies.

I think a lot of teams treat compensation like an HR or finance document.

It is not.

It is a GTM design tool.

A compensation plan has one real job:

Align behavior with outcomes.

That sounds obvious, but most plans miss it.

They either:

That last one is more expensive than people think.

In Salesforce’s guide to sales incentive plans, the recommendation is blunt: make it abundantly clear how sellers are being evaluated and how they are performing. If comp lives in a spreadsheet nobody trusts, reps start doubting the logic and the earnings.

That is not a spreadsheet problem.

That is a trust problem.

Salesforce makes the same case in its broader overview of why incentive compensation matters: plans need clear structure, clear metrics, fair and consistent payouts, and transparent access for employees. Otherwise morale and trust in leadership suffer.

I agree with that completely.

Because once reps stop trusting comp, they stop trusting the system beneath it.

Sales is getting more complex, not less.

According to Salesforce’s 2026 sales statistics roundup, reps now interact with prospects across an average of 10 channels. That means influence over revenue is more distributed than it used to be. Marketing affects deal quality. Presales affects confidence. Customer success affects expansion. Multiple people often touch the same outcome.

If your comp plan still assumes one lone wolf dragged the whole deal across the line, you are rewarding a fantasy.

Modern GTM is much more collaborative than the comp plans of a lot of teams.

That mismatch creates politics fast.

I have seen comp plans break teams in very predictable ways.

A company says it wants clean handoffs and high-fit pipeline.

Then it pays SDRs only on booked meetings and AEs only on closed-won revenue.

Now everyone is “working hard,” but nobody is truly aligned.

The SDR wants volume. The AE wants closable deals. Marketing wants attribution. Customer success wants realistic expectations. Finance wants clean rules.

And leadership wonders why the room feels tense.

The answer is simple: the comp plan is teaching conflicting behaviors.

That is why I think good compensation is less about motivation than translation.

It translates company goals into everyday behavior.

If I were designing a comp plan for a GTM team right now, I would use these five rules.

If someone cannot explain how they get paid in under 60 seconds, the plan is too complicated.

This sounds basic. It is not basic enough.

In Salesforce’s guide to sales compensation plans, clarity and simplicity are framed as foundational. I think that is exactly right. People do not execute clearly inside confusing systems.

Do not pay people for outcomes they barely influence.

If an SDR has no control over late-stage pricing decisions, do not load their entire upside onto closed revenue.

If a customer success manager is supposed to protect retention, do not pretend that only new logo revenue matters.

A good plan should not just reward outcomes. It should protect the business from fake wins.

Examples:

If you want multithreading, handoffs, and collaboration, reward them somehow.

Split credit is not always elegant, but pretending one person did everything is worse.

A comp plan should not be “set and forget.”

Markets change. Motions change. AI changes how work gets done. Your plan should evolve with the motion.

This is the part I think more GTM teams need to address in the next two years.

AI is changing which tasks deserve incentive weight.

If AI can:

then the value of the human shifts.

You should not over-reward activity that software can fake or inflate cheaply.

You should lean more toward outcomes, judgment, collaboration, and high-quality execution.

That does not mean “never reward leading indicators.”

It means choose leading indicators that still signal real value.

For example:

Those are a lot harder to game than raw activity counts.

Let’s say you run a GTM team with:

And your current pain is this:

Here is how I would simplify it.

Pay for:

Do not pay only for raw volume.

Pay for:

Pay for:

Pay for:

Now the motion starts to feel aligned.

Not perfect. Aligned.

Before you finalize anything, I would run this exercise.

Example: “We need more high-fit revenue without increasing churn.”

Example:

If not, fix the plan.

Be honest here. Every plan gets gamed a little.

Almost every plan can be simpler.

That is usually a good sign, not a dangerous one.

A comp plan should make the right behavior feel obvious.

Not academically correct. Not technically impressive. Obvious.

When the plan is good:

When the plan is bad:

That is why I do not see compensation as an admin task.

It is one of the clearest windows into whether your GTM system is actually designed on purpose.

Your people will always pay attention to what pays.

So make sure the plan teaches the behavior you actually want to scale.