Job Costing for Contractors

What job costing actually is, and what it isn't

Job costing is the practice of allocating every dollar of cost, labor, materials, equipment, subcontractors, overhead, to the specific job that consumed it. The output is one number per job: gross margin. The input is a process most contractors never finish setting up.

Most contractors think they have job costing because their FSM tracks job revenue. They don't. Revenue is half the equation. Without cost allocation tied to the same job, you have a revenue ledger and a payroll ledger, but no margin.

The job-costing problem: revenue is only half the equation

Across 1,747,089 completed jobs from 1,791 companies, Level measured a 44.3% median job-level gross margin. That calculation is only possible when labor, materials, and subcontractor cost are tied to the same job as revenue.

This is the silent killer of contractor profitability. Profitable jobs subsidize unprofitable ones, the owner sees a healthy company-level P&L, and individual losers stay invisible until cash gets tight.

The fix starts with operational discipline, every timesheet codes to a job, every PO references a job, every receipt gets allocated, but it requires accounting to mirror the field. Most contractors stop at one of those two halves.

What good looks like: percentile benchmarks

Level's completed-job distribution runs from 10.8% at P10 to 75.1% at P90, with a 44.3% median. The spread is why a blended company P&L cannot tell you which job types need a pricing or scope correction.

The operational standard is simple: every timesheet, PO, receipt, and subcontractor bill maps to a job before the monthly close. That lets you separate a real pricing problem from incomplete job data.

The small-jobs trap: where margin blindness compounds

767,554 service calls under 4 hours make up 51% of the job count in our dataset and about 9% of revenue. Jobs under 24 hours are 93% of the count and about 27% of revenue. Job count and revenue share run in opposite directions, and the small end consumes more than half the operational complexity.

Most cost tracking discipline gets applied to the big jobs. The small ones get bucketed, billed, and forgotten. If even 20% of those small jobs run at a loss, that's $140M in revenue subsidizing losses across the dataset, and nobody can see it.

The fix isn't to chase a perfect time sheet on every $200 service call. It's to use job-type averages from the jobs you DO track to flag the ones that look out of pattern. A $300 job that took 6 hours and used $150 in parts isn't profitable, but it'll never get flagged unless you're building the average and comparing.

Fixing the process: the 90-day playbook

• Audit your last 90 days of completed jobs. Pull every job with revenue and check whether labor cost, material cost, and subcontractor cost are coded to that job. Compute the percentage. If it's under 80%, the problem is operational, not accounting.
• Set a rule: every timesheet codes to a job, every PO references a job, every credit-card receipt gets allocated within 7 days. No exceptions. The exceptions are where margin disappears.
• Pick a job-type taxonomy you can actually maintain. Five categories beat fifty. PM, callback, install, T&M project, change order is enough for most service contractors.
• Build the weekly report: every job completed last week with revenue, cost, and margin. Sort by margin descending. The bottom 10% is your re-bid list, your scope-creep list, or your customer-fire list.
• Reconcile monthly. The job-level P&L should sum to the company P&L within rounding. If it doesn't, something is unallocated and your numbers are wrong.

When DIY stops working

Job costing is the kind of problem where the cost of getting it wrong scales with revenue. A $1M contractor with broken job costing has $50-$100K of margin hidden somewhere. A $20M contractor with the same broken process has $1-$2M in compounded blind spots and probably can't sell the business at a fair multiple.

If you've tried to fix this internally twice and it didn't stick, the issue isn't your team, it's that the gap between operations and accounting needs something whose job is to bridge them, week after week. That's what Level's operating layer is for.