What Is Job Costing in Construction, and Why Most Contractors Get It Wrong

Job costing is the practice of tracking what a single job actually costs you, line by line, while the job is still running. Not at tender. Not at final account. Every month, in between.

Most small contractors do it twice: once when they price the work, and once when the job is finished and the accountant tells them what happened. The gap between those two moments is where margin quietly disappears.

Why the two-point method fails

A tender is a forecast. A final account is a post-mortem. Neither tells you anything while there is still time to act.

If your labour is running 8% over on the groundworks package, you want to know in month two, when you can still tighten supervision, renegotiate a subcontract, or price the next variation properly. Finding out in month eleven means you absorbed it.

What job costing actually requires

Less than people assume. Four things:

• A cost breakdown you trust. Usually your Bill of Quantities or your priced schedule of works. This is your baseline.
• A time axis. Spread the planned quantities across the programme. If a package is planned over months three to seven, it should show up in months three to seven.
• Actual costs, categorised. Labour, materials, plant, subcontractors, other. Recorded against the same structure as the baseline, or the comparison is meaningless.
• A monthly comparison. Planned versus actual, by category, with the variance visible.

That is it. Job costing is not complicated. It is just tedious enough that most people skip it until a job goes wrong.

The mistake that makes job costing useless

Recording actual costs in a structure that does not match your baseline.

If your tender is broken down by work package and your invoices are filed by supplier, you cannot compare them. You end up with a total that tells you whether you are over or under overall, and nothing about where.

An overall number is not job costing. It is a rear-view mirror. The value is entirely in the breakdown.

Monthly, not weekly

Weekly job costing sounds rigorous and is usually abandoned by week five. Monthly is frequent enough to catch drift while it is still correctable, and infrequent enough that it actually gets done.

Pick a day. The fifth working day of the month works well, because supplier invoices for the previous month have usually landed. Block two hours. Do it every month without exception.

What to look at first

When you open your monthly figures, look at three things in this order:

1. Margin against tender. Is the job still delivering the margin you priced? A one-point drop in month three is a trend, not a rounding error.
2. The worst-performing category. One category usually drives most of the variance. Find it before you look at anything else.
3. Cash position. Profitable jobs still fail on timing. If you are paying subcontractors at thirty days and getting paid at sixty, margin will not save you.

Spreadsheet or software?

For a contractor running one to five jobs at a time, a well-built spreadsheet is usually the right answer. Construction ERP systems are powerful and expensive, and most of their value sits in features a small contractor will never use.

The argument for a spreadsheet is not cost. It is that you will actually use it. A system that requires everyone on site to change how they work gets abandoned. A spreadsheet that one person updates monthly gets maintained.

The argument against is fragility: spreadsheets break when they are built badly, and a broken cost model is worse than none. Protect your formula cells, keep your input cells separate, and never let two versions circulate.

The point

Job costing does not make a bad job good. It makes a bad job visible, early enough that you can do something about it.

That is the whole value. Not reporting. Not compliance. Just seeing the drift while it is still small.