You win the job, you allow the costs, you get on site. Eight weeks in, someone asks how it's tracking and you genuinely don't know. You've got invoices piling up, labour on the timesheet, materials on order, but nothing lined up against what you actually allowed at tender. That's the gap project cost tracking is supposed to fill, and on a lot of jobs it simply isn't done until the end.
By then it's too late to do anything about it. The overrun has already happened, and the answer is whatever it is.
Costs drift quietly, then all at once
A job rarely blows its budget in one big hit. It's a few extra days of labour here, a materials price that crept up there, a package that cost more than allowed because the scope on site wasn't quite what was priced. Each one is small. None of them trigger an alarm on their own. It's only when you add them up against the original allowance that you see the job's in trouble, and by then the work is already done.
The reason it stays hidden is usually the way costs get recorded. Invoices go to the accounts. Timesheets go to payroll. Nobody puts them next to the tender. So each cost is correct in its own place, and nobody can see the pattern.
Project cost tracking against what was allowed
The way to catch this while there's still time is to keep the job's actual costs running against the allowance from the original bill, trade by trade, as the job progresses. Labour against what was allowed for labour. Materials against what was allowed for materials. Each package against its own line, not lumped into one number that hides where the problem actually is.
It doesn't need a big system. A spreadsheet works if it's kept up. What matters is that the columns match the way the job was priced.
- Set up the baseline. Take the tender allowances for each package or element and put them in one sheet: labour, materials, plant and subcontract, separately where you can.
- Code every cost. Each invoice and each week of labour gets tagged to the package it belongs to. If it doesn't fit anywhere, that's a sign it wasn't priced.
- Record progress, not just spend. Money spent means nothing on its own. You need to know how much of the work is done for that money.
- Project the finish. For each package, work out what it'll cost to complete at the current rate of spend, and compare that to the allowance.
- Review on a fixed rhythm. Weekly or fortnightly, same day, same sheet. The value is in spotting a trend early, which you can't do if you only look once a month.
A worked example of catching it early
Say the brickwork on a small job was allowed 20 gang days of labour. Four weeks in, the timesheets show 12 days used and the site manager reckons about 40% of the brickwork is laid. At that rate, the full package takes 12 divided by 0.4, which is 30 days. That's 10 days over the allowance.
Nothing has gone wrong yet in the accounts. Nobody's been overpaid. But you now know, in week four, that brickwork is heading for a 50% labour overrun. You can ask why. Maybe the gang's short-handed. Maybe the scaffold's slowing them down. Maybe there's more brickwork on site than was measured, which is a different conversation. Whatever the reason, you've got time to act on it.
A job that's drifting off its allowance is a problem you can manage if you see it in week four. It's just a loss if you see it at the end.
Tracking only works if the baseline was real
This only works if the original allowance was measured properly in the first place. If the brickwork allowance was a rough figure, the comparison tells you nothing. You can't tell whether the gang is slow or the tender was light.
We measure every job to NRM2, so what you're checking site costs against is a real quantity and a real rate, not a rough guess that was never going to hold up as a comparison. When each element in the bill has its own measured quantity, you can compare work done against work allowed, not just money spent against money allowed. That's what makes the tracking useful. Project cost tracking against a guessed baseline just produces confident-looking numbers that mean very little, and it can send you chasing a gang that was never the problem.
It also helps with the money coming in, not just the money going out. The same package breakdown that drives your cost tracking is the basis for your valuations, and lining the two up gives you a clear view of cash flow stage by stage.
What to do when a package starts to slip
Seeing the overrun is half the job. The other half is doing something while it still matters. Check whether the scope on site matches what was measured. If it doesn't, you may have a variation to raise, and you want to raise it now while the evidence is fresh. If the scope's the same and the cost is higher, look at productivity, sequence and supply before the next package starts repeating the same pattern.
Leave it and the answer turns up in the final account, when all you can do is explain it.
The one number to check every week
If you only track one thing, make it this: projected cost to complete for each package, against its allowance. Not money spent to date. Money spent tells you what's happened. Projected cost tells you what's about to happen, and that's the only number you can still change.