Pay on revenue and you have bought volume and simplicity: the rep can calculate their own check at the kitchen table, and the margin decision stays with whoever prices the work. Pay on profit and the rep has a reason to care what happens after the signature, but your job costing becomes a payroll system. If you cannot close a job's costs cleanly inside your pay cycle, you do not have a profit plan.
Written by JD Davenport, who still sells for a living. Updated 4 November 2026.
Roofing in the question because that is where the examples come from. The arithmetic is the same for any business that quotes the work, does the work and sends an invoice: landscaping, fencing, decking, power washing, painting, HVAC, plumbing, electrical, remodeling.
Most answers to this start with which one is fairer. That is the wrong opening, because both are fair and they are fair to different things. The useful question is what each plan pays people to do when nobody is watching, and what each one requires from the rest of your business in order to work at all.
Here is what I have learned running both.
Pay a percentage of contract value and you have bought volume. The rep is paid to sell, and every hour spent anywhere other than in front of a homeowner costs them money. That is not a flaw. For a company that needs throughput, it is the correct instrument.
What you have not bought is any interest in what the job costs. A rep on revenue is indifferent between a job that nets twenty points and one that nets four. So the margin decision moves entirely to whoever prices the work, and it stays there. If your estimating is tight and your production is predictable, that is fine. If either is loose, a revenue plan will grow you into a smaller profit.
The second thing a revenue plan buys is simplicity, and simplicity is worth more than it sounds. The rep can calculate their own check at the kitchen table. Nobody argues about an invoice they have never seen. Payroll closes on time.
Pay a percentage of job-level net profit and the rep now has a reason to care about the things that happen after the signature. Change orders get documented. Material returns get chased. A crew that burns two extra days is a problem the rep can feel.
That is the promise. Here is the part the promise leaves out.
A profit plan turns your job costing into a payroll system. The moment commission is calculated from job cost, every uncoded invoice is a pay dispute, every miscoded one is a pay error, and every late supplier bill is a check you cannot cut. Accuracy stops being a reporting nicety and becomes a legal and relational obligation. If you cannot close a job's costs cleanly inside your pay cycle, you do not have a profit plan. You have a plan that pays people an estimate and reconciles it later, which is worse than either option.
Before you adopt profit-based pay, answer one question honestly: can you produce, at line level, every cost on a job within the period you pay in? Not a total. The lines. If the answer is no, fix that first. The comp plan is downstream.
There is a belief that a margin floor solves this. Set a minimum, refuse to sell below it, and profit takes care of itself.
A margin floor protects the price. It does not protect the outcome.
I have audited jobs that sold at or above the floor and still lost money, and they were not rare or exotic. They were priced correctly and executed expensively. The overrun happened after the contract was signed, in labor and in material, in ways the pricing rule at the top of the funnel has no ability to reach.
This matters for comp design because both plans get blamed for it. On a revenue plan, the owner concludes reps do not care about cost and moves to profit. On a profit plan, the rep concludes they are being charged for a production failure they did not cause, which is often exactly true. Neither conclusion is about the comp plan. The cost is escaping between the estimate and the invoice, and no percentage split fixes that.
Diagnose that before you change how you pay people. If your losses are concentrated in a handful of jobs rather than spread across all of them, you have an execution problem with a tail, and changing the comp plan will move the money around without reducing it.
Three questions, in order.
Can you close job costs at line level inside your pay cycle? If no, pay on revenue until you can. This is not a compromise. Paying on numbers you cannot defend costs you more trust than the extra margin is worth.
Where do your losses concentrate? If a small number of jobs carry most of the shortfall, that is execution, and it belongs to production. If the shortfall is even across every job, that is pricing, and a profit plan will genuinely help because it recruits the rep into the pricing conversation.
What do you need more of right now? Revenue plans buy volume. Profit plans buy care. A company that needs to grow and a company that needs to stop leaking are not solving the same problem and should not copy each other's comp plan.
Two rules that cost nothing and prevent most of the damage.
Never reprice completed work. Jobs sold under the old plan stay on the old plan until they close. Retroactive changes are the difference between a change people accept and a change people resent, and the resentment lasts much longer than the money.
Write down what happens when the number is wrong. Not if. When. Who recalculates, in what window, and what the rep is owed in the meantime. A comp plan without a correction procedure has one anyway, and it is whatever mood the room is in the day it happens.
If you take one thing from this, take this. The comp plan is not the control. Job cost accuracy is the control, and the comp plan only converts it into behavior.
Measure three things before you touch the split:
The third one determines which plans are available to you. The first two determine which one you should want.
The written version of the plan, the agreement language, a calculator and a margin floor are The Commission Standard. Closing job costs at line level is The Job Cost Standard. More answers on the questions page.