The job's three weeks in and the numbers stop adding up. Materials are running ahead of what was allowed, a trade's cost more than the rate you priced, and there's an item on site that somehow never made it into the original figure at all. Now you're managing a problem that started weeks before the first spade went in the ground.
Most budget overruns aren't a site problem. They're a tender problem that hasn't shown up yet. If you want to understand why construction budgets overrun, look at the price before you look at the programme.
Why construction budgets overrun: it starts at tender
If the quantities were guessed rather than measured, the job needed more material than you allowed for, and you won't know until it's ordered. If the rates were carried from an old job, the margin was thinner than you thought from day one. If something got missed entirely, whether it's an attendance, a preliminary, or a fiddly item that didn't make it into a rushed quantity takeoff, that cost doesn't disappear. It just turns up later, at the worst possible time, as a shortfall you're now absorbing.
Site problems do happen. Ground conditions, weather, a supplier who lets you down. But those are the overruns everyone expects and plans for. The ones that hurt are the ones that were baked in before the contract was signed, because nobody on site can fix a number that was wrong on paper.
The usual suspects behind a runaway budget
On most jobs, the overruns that trace back to the tender fall into a small number of familiar groups:
- Short quantities. The blockwork, the concrete, the plasterboard. Measured roughly, ordered from the rough measure, and short by the time the job's half done.
- Stale rates. Prices lifted from the last similar job, which was priced in a different market.
- Missing items. Builder's work, making good, attendances, temporary works. Real costs with no line in the price.
- Preliminaries by percentage. Site costs added as a figure from habit, then the programme runs longer and every week costs more than was allowed.
- Scaled estimates. A cost per square metre from another job applied to this one, with none of its differences accounted for.
Every one of these is a decision made at the desk, usually under time pressure, usually for good reasons on the day. None of them looks like a problem until the job's running.
That's the uncomfortable answer to why construction budgets overrun on so many small and medium jobs. It isn't bad luck and it isn't usually bad site management. It's a set of shortcuts that each saved an hour at the desk and each cost far more than an hour on site.
A measured number closes the gap before it opens
The fix isn't better site management. It's a number that was right in the first place. When quantities are measured off the drawings element by element, to NRM2, and priced against current UK rates, the gap between what you allowed for and what the job actually needs is as small as the drawings allow. The number you tendered on is the number the job holds you to, and it was built on something real from the start.
That doesn't mean nothing will ever change. Clients change their minds. Drawings get revised. Ground turns out worse than the survey said. But those are variations, and you can price them against a measured baseline. A number built on a hopeful multiplier gives you no baseline at all, so even a genuine variation becomes an argument about what was in the price.
Site problems are the overruns you plan for. Tender problems are the ones that were baked in before anyone picked up a shovel.
What that buys you on site
This is the real value of a proper estimate: it moves the risk back to where it belongs, before the contract's signed, not partway through it. You go into the job knowing the figure was built on real quantities, so when something on site does move, you're managing an exception rather than firefighting a number that was never right to begin with.
It also makes tracking the job possible. With a measured allowance for each element, you can compare what's being spent against what was allowed, week by week, and see a package drifting while there's still time to act. There's a practical way to do that in checking a running job against its allowance. Without a measured baseline, all you can see is the total going up, and by the time it's obvious, it's usually too late.
And it changes your conversations with clients. An overrun you can explain, line by line, as a change from the measured scope is a variation. An overrun you can't explain is your loss.
How to stop the next one before it starts
You don't need a different way of running sites. You need a different habit at tender stage:
- Measure before you price. Every element off the drawings, not a rate per square metre from memory.
- Price on today's rates. Check the big-spend materials and the main subcontract packages for this job.
- Price the gaps. Preliminaries against the programme, attendances and builder's work with their own lines.
- Write down what you couldn't measure. An assumption in the tender protects you. A silent gap doesn't.
We measure the job properly before you commit to a price, to a fixed fee agreed before we start, delivered to your tender deadline.
The test for any price you're about to submit
Look at the tender and ask: if this job overruns, will I be able to show exactly which line it overran against? If you can, you've got a price you can manage. If the honest answer is that it'll just be "over", the overrun has already started. It just hasn't reached site yet.