Know your cost. Price off it. Pay your people from it. This isn't the math — it's the why. Five pieces, in order, and what each one is actually protecting.
Most shops price off a guess and pay off a guess, and then wonder why a busy year ends broke. This system replaces both guesses with one true number — what an hour of your labor actually costs — and then builds everything on top of it. Your prices come from that number. Your techs' quota comes from that number. Their commission comes from what they produce above that number. Nothing gets paid out until the company is already covered. That's the whole design.
There are two ways a service business quietly bleeds, and almost every shop does at least one of them.
You price off what the other guy charges, or a number that "feels right," or a flat per-head rate. It has nothing to do with what the job costs you. Some jobs make money. Some lose it. You can't tell which.
You pay hourly (so there's no reason to produce) or a percent of the ticket (so there's every reason to chase big material). Either way, pay isn't connected to whether the company actually made money.
Both guesses come from the same root: nobody ever calculated the real cost of an hour. Fix that one thing and both problems collapse — because now there's a true number to price from and pay from.
Everything sits on this. Your burden rate is what it truly costs to put one tech in front of one customer for one hour — not their wage, the loaded number: payroll taxes, workers' comp, liability, benefits, the truck and fuel, tools, phone, software. Then divided across the hours you can actually bill, because you pay for 40 and you don't bill 40.
It's almost always double the wage or more. And it isn't your price — it's your floor. It's the line every price has to clear before you've made a nickel.
You cannot set a price, a quota, or a commission without it. Every other step in this system is built on this number. Get it wrong and you've built a house on sand — everything downstream is confidently, precisely wrong.
Once you know your cost per hour, you stop pricing jobs one at a time and start pricing packages built off that cost — small job, medium, large, each with a set number of techs and estimated man-hours.
This is the piece that gets you out of the truck. Any tech can quote a job at a price that already protects your margin, without calling you. You're no longer the bottleneck on every estimate.
It also kills the per-head problem. The job where you trench around a tree and the wide-open backyard are not the same cost — but per-head pricing charges them the same. Packages price the time, which is the thing that actually varies and the only thing you control.
Here's the mental shift that makes it work in the field: your tech isn't estimating hours — he's selling a slot on his own calendar. The slot covers the whole commitment: the estimate trip, the load-out, the drive, the work, the pack-out, the phone calls. A "one hour job" was never one hour, and the slot already knows that. One rule holds it all together: techs always round up into the next slot. A 40-minute job bids as a 1-hour slot. Work that eats the day is a day. Over-bid a small job and the tech shrinks his own board — so honest bidding is the best-paying move, and nobody has to police it.
And building the packages takes one question per package, not a spreadsheet: "How many of these does a good tech sell in a normal day?" A tech-day has a cost — burden × 8. If six 1-hour jobs fill a day, each one has to bring back a sixth of that day or you're paying to work. That's the floor, and every package price gets set above it. The engine does this arithmetic for you the moment you answer the question.
Materials are always their own line item, marked up on their own. Never bury them inside a labor price. The moment material hides inside your labor number, you've lost the ability to tell whether the labor made money.
Each tech gets a monthly quota, and it isn't a number you pull out of the air. It's built from their burden — their real cost — plus a fair contribution to overhead and the minimum net profit the company needs.
That quota is the line they have to cross before a single dollar of commission exists.
This is the most misunderstood piece of the whole system, so read it twice: the quota is what makes commission safe. Because it's built on real cost plus profit, the company is fully covered before any commission is paid. Every commission dollar comes out of profit that wouldn't exist if the tech hadn't produced above the line.
A good pay plan has exactly two halves, and it needs both:
Straight hourly pays a tech the same whether they crush it or coast — so you get time, not results. Straight commission strips out the floor and pushes people to cut corners or sell what the customer doesn't need. Base plus a cut above quota is the balance: a floor for safety, a ceiling they own.
And critically — the commission is calculated on production above cost, not on revenue. Pay off raw revenue and you can end up paying commission on a job that lost money. That never happens here.
The split itself is fixed: everything above quota splits 60/40 — 60% to the crew, 40% to the company. A solo lead takes the full 60. A two-man crew splits it 30/30 — and here's the part that builds culture on its own: two techs sharing one pool only come out ahead of a solo guy if they more than double what the crew sells. Two vans, splitting up to cover estimates, helping each other finish — the math rewards exactly the behavior you want, without a single meeting about it.
The first floor covers the company — every expense plus your profit target. The second sits above it: what the same package has to sell for a good tech's commission to actually reach a career number — the $80–100K that keeps him from ever opening a job board. You set that number once, and the system shows both floors under every price. Price between the two and something quiet happens: the company's fine, the margins look fine... and eighteen months later your best tech gives notice. The gap between the floors is his paycheck. This system makes sure your prices were built to pay for it.
Both come off the top before the 60/40 split — and both are a percentage of the sale, which means they can't be buried in an hourly burden rate. The system handles it the only mathematically honest way: your floors get grossed up so the customer funds the royalty and the PM, not your tech's check. A 6% franchise with a 10% PM means only about 85¢ of every customer dollar reaches the pool — the engine shows you exactly what that costs per tech-day, in writing, probably for the first time.
The last piece is the one people skip, and it's the one that changes behavior: every pay period, each tech sees their production, their quota, and their commission. In plain numbers.
An incentive nobody can see isn't an incentive. A tech who watches that number manages themselves toward it. A tech who finds out at year-end doesn't. Visibility is the mechanism — it's what turns a pay plan into a scoreboard.
Here's the part that isn't about money, and it's the reason I built the whole thing.
When your pricing is dialed and your pay system runs itself, you stop spending every day fighting fires, chasing quotes, and refereeing payroll. The math buys you time. And time is what lets you do the thing you got into business for — lead your crew. Coach instead of bark. Sit down with a tech and talk about where he wants to go.
And your techs feel it. When the upside is theirs and they can see it, they take ownership. They stop watching the clock because the clock isn't where the money is anymore. They stop leaving — because nobody walks away from a real income with benefits and a clear path to earn more.
It all traces back to the boring stuff: know your cost, price off it, pay your people from it. Nail that, and you earn the time and the margin to actually take care of your crew — and you end up with a business that runs without you. Which, if you ever want to sell it, is the only kind worth buying.
These five steps are a sequence, not a menu. Burden rate → packages → quota → split → reporting. Each one is built on the one before it. Skip a step and the ones after it are running on a guess again.