The check cleared Friday. Four grand, job closed, and it felt like a solid week. You shook the homeowner's hand, loaded the truck, and moved on to the next one.
Here's the thing nobody does. You quoted that job at $4,000 figuring two days and maybe $1,100 in materials. It took three days. The lumber yard bill came in at $1,500. Somebody ran back for parts twice. So did you make money, or did you just stay busy and collect a check that felt bigger than it was?
Most guys never check. The money hits the account, the quote gets filed, and nobody ever lines up what you guessed against what actually happened. That's how you end up bidding the next ten bathroom remodels at a number that loses you money every single time, and never knowing why the year felt busy but the bank account never moved.
Your quote is a guess. The job is the only truth.
When you quote from the driveway, you're guessing. Educated guess, sure, but a guess. Hours, materials, how bad the demo gets once the wall's open. You can't really know until you're in it.
That's fine. Quoting is supposed to be a guess. The mistake is never going back to see how close the guess landed. A quote you never check is just a number you pulled out of the air and got paid on. Sometimes the air is generous. Sometimes it eats you alive.
The fix isn't fancier estimating software. It's a five-minute habit after the job's done, while you still remember what actually happened.
Log two numbers and you're basically done
You don't need accounting for this. You need two real numbers per job: the hours your crew actually put in, and what you actually spent on materials. That's it.
- Actual hours. Not what you quoted. What the crew really burned, including the two runs to the supply house and the hour spent waiting on the inspector.
- Actual material cost. Add up the receipts. The real ones, not the estimate you wrote three weeks ago.
Write them next to the quote. A notebook works. A notes app works. One spreadsheet row works and is easier to search later. Job name, what you quoted, what it actually cost you, done.
If you want to get one number smarter, track your labor at a loaded rate, meaning what an hour of crew actually costs you once you add payroll taxes, fuel, insurance, and the truck. Figure it's a good bit more than the wage on the paycheck. Even a rough loaded number beats pretending labor is free.
Do the math while it's still warm
Line it up the day the job closes, or that weekend at the latest. Wait a month and you won't remember that the "two-day" job ate a third day, or that you swallowed the cost of the wrong fitting.
The math is grade school. Quote minus actual materials minus actual labor. What's left is what you really made. Compare that to what you figured you'd make when you sent the number.
Some jobs you'll be dead on. Good. Others you quoted at $4,000 and cleared $400 after you paid everyone and everything, and that one stings, but it's the one you needed to see.
Find the pattern, not the one ugly job
One job going sideways isn't a problem. Rain, a surprise behind the wall, a sub who flaked. That happens, you move on.
The pattern is the problem. After ten or fifteen logged jobs you start seeing it. Your framing and your straightforward service calls come in right on the money, every time. Your bathroom remodels bleed. Every one of them runs long, every one eats more material than you figured, and every one clears a lot less than it looked like on paper.
Now you know something real. It's not that you're bad at your trade. It's that you're quoting bathrooms with framing-job optimism, and the job type itself is where the leak is. Maybe the demo always surprises you. Maybe tile takes your crew longer than you think. Maybe you forget the little stuff, the valves and the trim and the third trip to the store, that only piles up on the complicated jobs.
You fix that by bumping the number on the next bathroom, or padding the hours you budget, or walking away from the ones that never pencil out. Can't make any of those calls until you've got the pattern in front of you.
Where software helps, and where it doesn't
Logging two numbers is a pen-and-paper habit. You do not need AI to do it, and anyone telling you different is selling you something.
Where a little automation earns its keep is the annoying part: pulling the numbers together so you'll actually look. A simple setup can read your supply-house receipts from your email or a photo, drop the material total onto the right job, and tally it so you're not adding receipts by hand on a Sunday. It can flag the jobs where actual blew past quote by more than, say, fifteen percent, so the losers surface without you hunting.
What it can't do is tell you why a job ran over. That's your read on your own crew and your own trade. Software can hand you the sorted list and circle the bleeders. You still have to know that tile's your slow spot, or that one sub is always late. Keep that line clear and you won't overpay for a dashboard that tells you less than your own eyes would.
Worth doing this week
- Pull your last five closed jobs. Just five.
- For each, write down three things: what you quoted, what you actually spent on materials (dig up the receipts), and roughly how many crew hours it really took.
- Do the subtraction. Quote minus materials minus labor. See what you actually kept on each one.
- Look for the job type that keeps coming up short. That's your leak.
- Bid your next one of those higher, or budget more hours for it. Then log that one too and see if the fix held.
Five jobs and twenty minutes tells you more about where your money goes than a whole year of feeling busy. If you want help wiring up the receipt-reading and the auto-flagging so the habit sticks, that's the kind of thing we set up for Massachusetts trades every week. The logging, though, you can start today with a notebook.