A price for a job means little until you check it against something. Most builders know the feeling: the total is in, the tender is due, and something about the number feels off. Too low, maybe, or oddly high on one trade. Construction cost benchmarking is how you turn that feeling into a check you can actually act on.
The most useful comparison is not a published cost guide. It is your own past jobs: what you actually spent on a similar scope, on similar ground, with a similar programme. That comparison catches the item that is priced too low long before the client's QS does.
Why a number on its own tells you nothing
Say, hypothetically, your tender for a two-storey extension comes out at a certain total. Is it right? You cannot tell from the total. You can tell a lot more if you know what the last three similar extensions cost you, broken down the same way. If the new job's substructure is noticeably cheaper than all three, either the ground is better, the design is simpler, or something has been missed. Any of those is worth knowing before submission.
That is all benchmarking really is. Not a science, just a disciplined comparison with something you trust, done early enough to act on.
Keep your own numbers, not just the published ones
Published cost guides and data services like BCIS have their place, especially at early cost plan stage. But they move slowly, they average across many firms and regions, and they rarely match your own experience of what things cost to build in a particular region or on a particular type of site.
If you keep a record of what your last few jobs actually cost, broken down the same way each time, you build something far more useful than any external guide: a set of figures you know are accurate for how you actually work. Your labour outputs, your subcontractors, your overheads. Nobody else has that data.
It does not need a special system. A spreadsheet with one row per job and one column per element is enough to start with. What matters is that someone fills it in at the end of every job, while the final account figures are fresh, rather than trying to piece it together a year later.
Construction cost benchmarking only works like for like
A comparison only works if the categories match. If your own cost records are not broken down the same way as the new price, you are comparing two different things and the check is worthless. Consistency in how jobs are recorded matters more than how detailed any single record is.
A simple way to set it up:
- Pick one structure and stick to it. Group costs by the same elements on every job, whether you follow NRM1-style elements or your own trade headings.
- Record the basics of each job. Gross internal floor area, number of storeys, type of construction, location, start date and duration.
- Record actual cost, not tender price. The figure that matters is what the job cost at final account, element by element.
- Note what was unusual. Poor ground, restricted access, a client who changed their mind twice. Without these notes, an odd figure looks like a benchmark when it is really an exception.
- Divide by something sensible. Cost per square metre of floor area for the whole job, and cost per unit for elements where it makes sense (per door, per metre of wall), so jobs of different sizes can be compared.
Your own job history, recorded the same way each time, is the best cost data you will ever have. Most builders throw it away the day the final account is paid.
What moves the number
Material prices, labour availability and ground conditions all shift a price away from what a similar job cost last year. Before deciding a rate looks wrong, check whether the market has actually moved or whether the new job simply has a genuine complication the last one did not. Both are common, and mixing them up leads to the wrong conclusion.
Time is the other trap. A job you built a couple of years ago was priced in a different market. If you compare against it without allowing for price movement since, every new price will look expensive. It is better to note the date against each benchmark and treat older ones with more caution than to adjust them with an index you are not sure of. There is more on the factors that push prices around in what actually moves your price between tender and final account.
Using this on the next tender
The point of keeping your own job history is not to look backwards for its own sake. It is so that when the next enquiry lands, someone can sense check the price against real experience before it goes out, rather than finding out it was wrong once the job is under way.
Use benchmarks as a check, never as the price. A measured estimate built from the drawings is still the number you submit. The benchmark is there to ask questions of it. If you want to see what construction estimating services produce when every element is measured and priced, compare it against your own records and see where they differ. The differences are where you learn.
The quick sense check
Before any tender goes out, compare its cost per square metre and its three biggest elements with your two or three most similar past jobs. If anything is well out of line and you cannot explain why in a sentence, stop and find out. That one habit is the whole of construction cost benchmarking for most builders, and it catches more errors than any published guide.