Most tender prices that fall apart don't fall apart because the contractor's bad at his job. They fall apart because the number was built on shortcuts nobody had time to fix before the deadline. Strip away the noise and cost estimating best practices come down to a few things that separate a number that holds up from one that doesn't.
None of them are secret. Most builders could list them. The difference is whether they actually get done when there are two days left and three other tenders on the desk.
Measured quantities, not a feel for the job
The first is simple to say and hard to do under deadline pressure: measure it properly. A quantity pulled from memory, or scaled up from a similar job, drifts from what the drawings actually show. Measure every element off the drawings to NRM2 and the quantity in front of you is the quantity the job needs, not an approximation of it.
Scaling from a similar job is the most tempting shortcut because it feels like evidence. You did a four-bed last year, this one's a bit bigger, so you scale up. But the new one has a different roof shape, more glazing, a steeper site and a different floor build-up. The floor area went up by a bit. The cost went up by a lot more, and in different places.
Rates that reflect now, not last year
The second is pricing on current rates. Materials and labour move, sometimes month to month, and a rate you carried from a job priced a year ago is already out of date before you've submitted the tender. A rate built on what things cost today gives you a margin you can actually rely on, rather than one that's been quietly eaten by inflation you didn't account for.
Location matters just as much as date. A labour rate that works on your own patch may not hold for a job an hour and a half away, where the local market is tighter or where the gang needs travel time paid. Current rates means current for this job, in this place.
Nothing missed, nothing folded into a guess
The third is completeness. Preliminaries, attendances, the small items that never make it into a rushed take-off, these are exactly the things that erode a margin over the life of a contract. A properly measured bill catches them because it's working element by element off the drawings, not working from what an estimator remembers to include.
The related habit is not folding awkward items into a round sum. When something's fiddly to measure (the making good, the small bits of steel, the fire stopping), it's tempting to write an allowance and move on. Sometimes that's the right call. But if it's done to save time rather than because the information genuinely isn't there, you've swapped a measured item for a guess.
The contractors who price properly aren't the ones who win every bid. They're the ones still standing after the jobs they do win.
Cost estimating best practices you can check on every tender
Here's the short version, as checks you can run before any price goes out:
- Quantities traced to drawings. Every main quantity can be pointed to on a specific drawing and revision.
- Rates dated. You know when each major rate was last checked, and none of them are from a job you can barely remember.
- Preliminaries priced to the programme. Site costs worked out for this job's duration, not added as a percentage by habit.
- Subcontract quotes on the same scope. Every sub priced from the same measured quantities, so their quotes can be compared.
- Attendances and builder's work included. The work between trades has a line of its own.
- Assumptions written down. Anything that couldn't be measured is stated in the tender as an assumption or exclusion.
- A second pair of eyes. Someone other than the person who priced it looks at the biggest lines before it goes.
That last one gets skipped most. Everyone reads their own work as they meant it, not as they wrote it. A fresh look at the five biggest lines catches more than an hour of re-reading your own sheet.
Why this beats instinct, every time
None of this is complicated. It's just slower than guessing, which is exactly why it gets skipped when a deadline's close. Instinct isn't useless. It's good at telling you a number looks wrong. It's bad at telling you why, and it can't tell you what you've missed, because you can't have an instinct about an item you never saw.
The best estimators use both. They measure first, price on current rates, and then step back and ask whether the total feels right for a job of this kind. When it doesn't, they go looking for the reason in the measured lines. That order is what cost estimating best practices really mean in day-to-day pricing: evidence first, gut feel as a check, never the other way round.
That's where a measured estimate earns its fee. We measure the job to NRM2, priced against current UK rates, with nothing folded into a guess. The fee is fixed and agreed before we start, and you get the full breakdown, so you can put your own judgement on top of real numbers rather than using it to make the numbers up. For a closer look at what that gives you, read a tender price you can actually stand behind.
The rule of thumb for your next price
Before the next tender goes out, ask one question about the total: if this price turns out wrong, where will it be wrong? If you can name the lines you're least sure of, you can check them now. If you can't, the price hasn't been built in a way you can check at all, and that's the bigger problem.