9 min read
Designing a Sales Compensation Plan That Drives the Right Behaviour
A compensation plan is the clearest statement of strategy a company makes. Reps read it more carefully than any deck. Design it around two or three measures, make quota attainable by most of the team, and govern changes deliberately.
Key takeaways
- If a plan needs a spreadsheet to explain, it will not change behaviour.
- Two or three measures maximum; more dilutes every one of them.
- Target 60–70% of the team hitting quota — much lower means the plan is broken.
- Never change the plan mid-year without written governance and communication.
What a compensation plan is actually for
Compensation is not a retention tool or a fairness exercise. It is the mechanism by which strategy reaches daily behaviour. Whatever the plan pays for is what the team will do, regardless of what the strategy deck says.
So the design question is not 'what is competitive?' but 'what behaviour do we need next year that we are not getting today, and does the plan pay for it?'
The structural choices
Four decisions define most plans. Get these right and the details matter far less.
- Base/variable split — typically 50/50 for new-business closers, 60/40 or 70/30 for account management and longer, more technical sales
- Measures — new ARR is the default primary; a second measure such as retention, margin or a strategic segment is often justified; a third rarely is
- Quota level — set so that a clear majority of the team can realistically reach it
- Accelerators — pay more per unit above 100% so top performers keep selling instead of sandbagging into next quarter
Setting quota honestly
Quota should be built bottom-up from territory potential and historical productivity, then reconciled against the company plan — not derived by dividing the board number by headcount.
The diagnostic is attainment distribution. If under a third of the team hits quota, the plan is a tax on morale and your hiring costs will rise as good sellers leave. If nearly everyone clears it easily, the plan is underpricing your growth.
Common failure modes
Most broken plans fail in one of these predictable ways.
- Too many components, so no single one changes a rep's day
- MBO or discretionary elements that reps cannot forecast their own earnings from
- Caps on upside, which quietly tell your best seller to stop in November
- Paying on bookings while the business needs cash or retention
- Changing the plan mid-year in response to one quarter's results
Governance: how to change a plan
Plan changes are among the highest-risk actions a commercial leader takes, and they are often made informally. Introduce a simple governance standard: changes are annual by default, modelled against the previous year's actual deal data before approval, communicated in a single session with worked examples, and documented with an effective date.
Mid-year changes should require an explicit exception decision, not a manager's discretion. The cost of an ungoverned change is not the payout — it is the loss of trust in every future plan.
Test the plan before you publish it
Run last year's closed deals through the proposed plan, rep by rep. You will immediately see who wins, who loses, and where the plan pays for behaviour you do not want. Then ask three reps to explain the plan back to you from a one-page summary. If they cannot, simplify it.
Frequently asked questions
What is a typical base-to-variable split for B2B sales?
50/50 is the common benchmark for quota-carrying new-business roles. Longer, more consultative or technical sales often move to 60/40 or 70/30, and account management roles typically carry a smaller variable component.
Should sales compensation include team or company measures?
Sparingly. A small company-performance modifier can support collaboration, but anything a rep cannot personally influence weakens the plan's power to change behaviour.
How often should a sales compensation plan change?
Annually, aligned with the fiscal planning cycle. Mid-year changes should be exceptional, documented, and modelled against real deal data before approval.