# How to Track Job Costing in the Field: A Simple Guide
A crew of three works a nine-hour day on a commercial HVAC retrofit. Two guys remember to write their hours on a paper timesheet. One forgets. The foreman estimates. The office manager rounds. By the time payroll runs, the job has been credited with 22 labor hours instead of 27. That five-hour gap, at a fully loaded labor rate of $65 an hour, is $325 in cost that just vanished — on one job, one day.
That's field job costing failing in real time, and it's the norm rather than the exception at most small trades businesses. The fix isn't complicated, but it does require a system that captures labor hours, material costs, and overhead at the job site — not reconstructed from memory at the end of the week.
Disclaimer: This article contains general business guidance. Consult a licensed accountant or financial advisor before making significant changes to your accounting or cost-tracking systems.
Job costing means tracking every dollar — labor, materials, subcontractors, equipment, overhead — against a specific job instead of a general expense pool. Most small trades businesses don't do this. They do "bank account accounting": look at what came in versus what went out each month, and guess at profitability by job.
That approach tells you whether the business is alive. It doesn't tell you which jobs are killing you.
Industry estimates from construction cost-overrun research put rework and inefficiency losses at somewhere between 5% and 15% of total project value on trades and construction work. For a business doing $800,000 a year, that's a wide range — $40,000 to $120,000 — but even the low end is real money walking out the door on problems that job-level data would have caught in week one, not month twelve.
Field job costing exists to answer three questions:
Without answers, you reprice the same bad jobs next year and wonder why margins don't move.
Labor typically runs 30% to 50% of total job revenue in a trades business, and it's the cost most often tracked sloppily — through memory, rounding, and end-of-week reconstruction.
The fix is a mobile app where each job gets a unique code. Techs clock in and out against that code, and the app captures start, stop, travel time, and breaks — no estimating required.
Worth evaluating:
This isn't about surveilling your crew. It's about removing the guesswork that makes your labor data unreliable. In practice, most experienced techs prefer it — they stop absorbing blame for cost overruns that were actually bad estimating, not bad work.
Tracking hours is pointless if you don't know what an hour actually costs you. Fully loaded rate includes:
A $32/hour electrician may actually cost $52–$58 fully loaded. If you're estimating jobs at $35/hour labor, you're underwater before you buy a single fitting.
For a business with 4 to 15 field employees, the stack doesn't need to be sophisticated — it needs to be fast enough that people actually use it.
| Function | Tool options | Monthly cost range |
|---|---|---|
| Time tracking + job codes | ClockShark, QuickBooks Time, Jobber | $40–$150 |
| Material capture in the field | Jobber, ServiceTitan, or a photo-log workflow | $49–$300+ |
| Job cost reporting | QuickBooks Online, Knowify, CoConstruct | $30–$199 |
If you're not ready to invest in field management software, the photo-receipt workflow costs nothing beyond what you already pay for: each tech gets a folder in a shared Drive or Dropbox labeled by job number, photographs every receipt same-day, and an admin reconciles weekly. It's not elegant, but it closes the gap.
ServiceTitan is the dominant platform for larger HVAC, plumbing, and electrical operations, with job costing built into dispatch and invoicing — but it starts around $398/month and scales up from there. Under $500,000 in annual revenue, that's a hard number to justify. Above $2 million, the reporting tends to pay for itself through better bid accuracy alone.
The two failure points: receipts that never make it back to the office, and materials pulled from truck stock that never get logged to a job.
Job-specific purchase orders solve the first problem. Every job gets a PO number before work starts, and major suppliers — Ferguson, Wesco, Fastenal, Home Depot Pro — can attach purchases to that PO directly, so your monthly statement arrives pre-sorted by job.
Truck stock needs a different fix. A tech's van might carry $3,000 to $8,000 in fittings, valves, and supplies at any given time. Use a standard replenishment sheet: after each job, the tech logs what they pulled from the van, ties it to the job, and the shop restocks at week's end based on that log. Skip this step and that inventory just evaporates into cost of goods sold with no job attached — which means your material tracking, however careful otherwise, has a hole in it the size of your truck stock.
Buying software and seeing no change is its own common failure mode. Usually it's one of these:
You track costs but never compare them to the estimate. If you bid 40 labor hours and spent 54, that's a fact worth acting on. Left in a spreadsheet unreviewed, it's just a number.
You wait until the job closes to look at costs. Too late by then. Build a checkpoint at 50% completion on anything running more than two days — if labor's already over budget, you can address scope creep before the margin is gone rather than after.
You allocate overhead arbitrarily, or not at all. Every job should carry a slice of rent, insurance, vehicle depreciation, and software costs. Take total monthly overhead, divide by monthly billable hours, and you get a burden rate per hour. At $12,000/month overhead and 400 billable hours, that's $30/hour — and it belongs on every job cost, not just the year-end P&L.
You skip training the field. Owners set up the software and tell the crew to use it. Two weeks later, nobody does. It takes a twenty-minute session, a one-page reference sheet, and a foreman who actually enforces the habit. The software was never the hard part.
Once you have 60–90 days of clean numbers, a few things become visible fast:
Which job types are actually profitable. It's common for a trades business to discover that the work they've always chased — new construction, say — is their lowest-margin category, while routine service calls quietly outperform it with far less overhead and risk.
Where your estimates are consistently wrong. If you run 15% over on labor for bathroom remodels but hit target on kitchens, your remodel labor factor needs revision. This is the mechanism by which estimating gets better instead of staying permanently optimistic.
Which crews need a different conversation. Job costing by crew, not just by job, surfaces real productivity gaps. Handle it carefully — but a crew running 20% over on hours consistently either needs training, better pre-staged materials, or a different mix of job types, and the data tells you which.
Why your bids can get more confident. Once you know your real delivery cost, you stop padding bids out of anxiety and stop taking margin-negative work just to keep the crew busy.
These vary meaningfully by trade, and it's worth knowing roughly where you should land before you panic over a single month's numbers. Service-oriented HVAC, plumbing, and electrical work typically targets 55–65% gross margin on labor and materials combined. New construction and low-bid commercial work often runs much thinner — 25–40% is common. If your job costing shows you consistently below 40% across the board, that's a signal to check labor burden, overhead allocation, and pricing — usually in that order, since burden rate errors are the most common root cause.
Pull your last five closed jobs and compare actual labor hours to what you estimated. You don't need new software for this — payroll records and your original bids are enough. That single comparison will tell you more about where your job costing needs to start than any tool evaluation will.
This content is for general informational purposes and does not constitute accounting or financial advice.