TL;DR
- A good compensation model is simple enough for a rep to compute their commission in their head: complexity kills the motivating effect.
- The accepted ratio in phone sales: 60% to 75% base, the rest variable. Too little base creates pressure that damages both calls and retention.
- Every model distorts something: commission on closes alone encourages pushing and neglects follow-up and retention. You manage the distortion, not ignore it.
- Call data unlocks a new layer: a quality bonus on process adherence, aligning the rep's interest with the customer's good.
You can write a vision, phrase values and hang scripts on the wall, but reps will do what the compensation pays for. A commission model is not an HR appendix, it is the floor's most powerful management tool, and therefore its most dangerous: a warped model produces warped behavior with exactly the same efficiency. In this guide: the building principles, the classic traps, and the new layer data makes possible.
Principle 1: simplicity beats sophistication
The test: can an average rep compute, mid-month, in their head, what they are going to earn? If not, the model does not motivate. Three components are a healthy maximum: base, sales commission, and one focused bonus. Models with seven coefficients, multipliers and nested thresholds look sophisticated in the deck and vanish from the rep's mind by day three.
Principle 2: the base is not a cost, it is stability
The temptation to cut base and raise commission is understandable, and its price is well documented: reps under existential pressure push instead of sell, abandon slow but valuable leads, and burn out fast. The ratio that works on most phone floors: 60% to 75% base. The variable should be large enough to move behavior, and small enough that one weak month does not break the rep.
The classic traps, and the fixes
- Commission on closes only: neglects follow-up, retention and documentation. Fix: part of the commission conditioned on the deal surviving the cancellation window.
- Sharp tiers: a commission jump at target produces end-of-month pushing and start-of-month sandbagging. Fix: gentle tiers or a continuous rate.
- Competition at the cost of sharing: when only first place pays, nobody shares techniques. Fix: a small team bonus alongside the personal one, restoring the interest in the shared call library.
- Frequent changes: a model that shifts every quarter teaches reps not to trust it. Fix: one major change a year, well explained, with a transition month.
The new layer: a data-based quality bonus
The old problem with quality pay was measurement: you cannot reward "good calls" when 5% of calls are reviewed, subjectively. When every call gets a score on transparent criteria, a new option opens: a quarterly bonus on process adherence, stage coverage, full discovery, objection handling, documentation. The beauty of this layer: it aligns long-term interests. A sales commission pays for this month; a quality bonus pays for the habits that bring all the following months. And it is fair to whoever drew weak leads this month: the rep fully controls the process, only partly the outcome.
Transparency is half the compensation
An excellent model with an opaque calculation feels like a lottery. The rep should see, at any moment, where they stand: sales made, commission accrued, distance to target, and their quality score, on the same dashboard where they see their calls. When the numbers live in front of their eyes, the model works between paychecks; when they surface on the payslip, it works two days a month.
Frequently asked questions
Commission on revenue or on profit?
For most floors, revenue: simple, transparent, computable in the head. Profit commission makes sense only when the rep genuinely controls discounts, and then real-time numbers are essential. A profit model without full transparency reads as a trick.
Should meetings and leads be compensated, or only closes?
In roles that produce meetings (SDR, appointment setting), you must: the close is far from them. The rule is rewarding the output the role controls, with quality control against inflation: a meeting that happened, not one that was booked, and a process score verifying it was booked properly.
What about a veteran whose income the new model reduces?
Bridge personally: an adjustment period with a safety net, and a clear path to match and exceed the old income through the new components. A strong veteran leaving over a model transition costs more than all the bridging months. And if their income was built on behavior the new model is meant to fix, that is a different honest conversation, also best had early.