Job Costing for Small Contractors: The Method That Fits a Trades Crew

Job costing has a reputation problem in small shops. The phrase sounds like software demos and CSI cost-code manuals, so owners file it under "when we get bigger." Meanwhile the question job costing answers is one every owner already asks in the truck on the way home: did we make money on that job?

Here's the honest secret: for a trades crew, job costing is mostly one habit. Labor is your biggest and most volatile cost, and it's also the only major cost that evaporates if it's not recorded at the moment it happens. Materials leave receipts. Subs leave invoices. Hours leave nothing but memory. Capture the hours to the right job and the rest of job costing is arithmetic.

01

The four numbers

The minimum viable job cost system tracks four numbers per job:

1. Estimated labor hours. What you bid. If you bid in dollars, divide by your average rate to get hours; you need this number in hours, because hours are what you can watch during the job.
2. Actual labor hours. What the crew really spent, tagged to the job at clock-in, not reconstructed on Friday.
3. The burdened labor rate. What an hour of a worker really costs you: wage plus payroll taxes, workers' comp, and the rest. It typically runs 20 to 40 percent more than the wage, and using the raw wage instead is the classic way contractors convince themselves a losing job made money. The full math is in How to Calculate Labor Burden.
4. Labor cost per job. Actual hours times burdened rates, person by person. The step-by-step with a worked example is in How to Calculate Labor Cost per Job.

With those four you can answer the truck question with a number: on an $8,000 job you bid 120 hours, burned 141, and at a $34 burdened average that overrun cost $714, which took the margin from a planned 38 percent to an actual 29. No accounting department required.

Where each of the four numbers comes from.

02

Where the data has to come from

The method stands or falls on number 2, actual hours by job, and this is where most attempts die. If hours are written down at the end of the week, they get allocated by vibe: the big job absorbs the mystery hours, the little jobs look free, and the numbers say nothing. A worker who split Tuesday across two sites won't remember the split on Friday. Nobody does.

The repair is to record the job at the punch. In Clox, every clock-in is tagged to a job and task, and switching jobs mid-day is changing the tag when you roll to the next address. The hours land allocated the moment they happen, which is the only moment anyone actually knows the truth. Add a travel task and even windshield time gets costed instead of vanishing into the biggest job on the board, which connects to whether and how you pay for drive time.

If your work involves distinct phases you bid separately, rough and finish, demo and install, add tasks for them. But keep the list short. A task list longer than about nine entries stops being used honestly, which is a lesson we spell out in Cost Codes for Small Crews.

03

The Friday 15 minutes

Data does nothing by itself. The whole return on job costing comes from a short weekly ritual, and 15 minutes is genuinely enough at this crew size:

1. Open the week's hours grouped by job. In Clox this is the Reports view you already use for payroll approval, so the ritual piggybacks on a Friday habit you already have.
2. For each active job, compare hours burned to date against the estimate. A job at 60 percent of budgeted hours and 40 percent complete is yelling at you, on Friday of week two instead of in the accountant's report next quarter.
3. Ask one question about any overrun: scope creep, a bad estimate, or a slow day? Scope creep becomes a change order conversation while the client still remembers asking. A bad estimate becomes a note in the bid file. A slow day becomes a crew conversation.
4. When a job closes, write two numbers in the bid file: estimated hours and actual. That file is where your next bid gets sharper, which is the whole feedback loop described in Bid Jobs Accurately With Labor Data.

04

What about materials and subs

Add them when you're ready, but don't let them block the labor habit. Materials and subs already generate paper, so a shoebox-and-spreadsheet version works: receipts coded to the job weekly, sub invoices likewise. Most small contractors find labor explains the majority of the variance between good jobs and bad ones anyway, because materials get quoted and subs get contracts, while labor gets estimated and then drifts. When you want a benchmark for labor's share of a job, build it from your own closed jobs, as laid out in What Should Labor Cost Be as a Percentage of a Job?

If you run QuickBooks, your tagged hours can ride along: Clox exports carry the hours out cleanly, and the QuickBooks integration guide shows the paths.

05

Start smaller than feels serious

The failure mode for job costing is launching the perfect system and abandoning it in three weeks. The success mode is embarrassingly small: this Monday, have the crew tag hours to jobs, with one travel task. Friday, spend 15 minutes comparing two numbers per job. That's the whole system for a month, and it will already have paid for itself the first time a job goes sideways on week one instead of at final invoice.

Clox is free for 14 days with no credit card, and includes job and task tagging, the approvals flow, and payroll exports on its one plan, $29 a month with the first 3 users included, then $6 per user. Start a free trial and find out what last week actually cost, job by job.

Reviewed August 2026.