It's the quickest way to price a job, and it's usually the way it goes wrong. Take the floor area, multiply by a rate you used on the last similar job, and you've got a number in five minutes instead of five hours. Cost per square metre estimating feels efficient right up until the job's on site and the number doesn't hold.
That doesn't make it useless. It makes it a tool for one stage of a job, and a dangerous one at every other stage. The trouble starts when a figure built for a first conversation ends up on a tender form.
What a square metre rate actually contains
A rate per square metre isn't a measurement. It's a summary of a job that's already finished. When you divide the final cost of a past job by its floor area, everything about that job gets folded into one figure: the ground it sat on, the specification the client chose, how easy the site was to get a lorry onto, how much external work went round it, the roof shape, the number of openings, even how the trades performed that month.
Then that figure gets applied to a new job that might share none of those things. The floor area matches. Very little else has to.
Why cost per square metre estimating fails at tender
Two extensions of the same floor area can carry completely different costs depending on foundations, drainage runs, roof form or how easy the site is to get materials onto. A square metre rate can't tell the difference. A measured takeoff can, because it prices what's on this drawing rather than what was on the last one.
Say you're pricing two single-storey rear extensions, both the same internal area. The first sits on good ground with a short connection to the existing drainage and a simple lean-to roof. The second needs deeper foundations because of a tree nearby, a much longer drainage run round the side of the house, a flat roof with a lantern, and access through a narrow side passage where everything goes in by barrow. Same area. Same square metre rate if you price it that way. Very different jobs.
On the first, the rate might even leave you a bit of fat. On the second, you've committed to a price that was never going to cover the work, and you'll only find out when the ground's open.
A square metre rate prices the average job. You never actually build the average job.
What measuring the actual job gets you instead
When you measure off the drawings, item by item, you're pricing what's actually there rather than what was there on a different job with a similar floor area. A longer run of drainage costs more. Difficult access costs more. A simpler roof costs less. None of that shows up in a per square metre rate, and all of it shows up in a measured takeoff.
The things a rate per square metre most often hides are the ones that don't scale with floor area at all:
- Substructure. Foundation depth and type depend on ground conditions and nearby trees, not on how big the room is.
- Drainage. The length of the run to a connection point can be three metres or thirty on jobs of the same size.
- Roof form. A pitched roof with hips, a flat roof with a lantern and a simple lean-to all cost differently per square metre of floor below them.
- Openings. A wall of glazing and a wall of blockwork cover the same area and carry very different prices.
- Access and logistics. Scaffold, skips, barrowing and restricted delivery times sit outside the building footprint entirely.
- Tie-ins and making good. Work to the existing building is often where extensions lose money, and it barely relates to the new floor area.
That's the difference between a number that survives contact with the site and one that quietly falls apart by week three, once the ground's opened up or the access turns out to be worse than it looked on the drawing.
Where the shortcut still has a place
A per square metre figure isn't useless. It's fine for a very rough early conversation before there's anything to measure properly: a client asking whether their idea is even in the right ballpark, or a feasibility check before anyone pays for full drawings. At that stage, a range is honest and a detailed price would be pretending to know things nobody knows yet.
It just isn't a number to bid a tender on or manage a budget against, and too many jobs get priced that way because there wasn't time to do it properly. If you want to see where each type of figure belongs, our guide to the different types of construction estimate sets out what each one is for.
Moving from a rate to a measured price
The step up from cost per square metre estimating to a proper price is measurement. You don't need a better rate; you need the quantities. Once you know how many metres of foundation, how many square metres of blockwork, how many linear metres of drainage and how many openings the drawing actually shows, the rates you apply become a lot less of a gamble.
Send us the drawings you've got and we measure the job to NRM2, element by element, and price it on current UK rates, local to the job, for a fixed fee agreed before we start. The number reflects what's actually on the drawing rather than what a rough rate assumes is typical.
A simple rule for when to stop using it
Here's the test. If the client is going to sign something, or you're going to commit labour and materials, based on the number, it needs to come from measured quantities. If it's only there to decide whether a conversation is worth having, a square metre figure will do.
Use a rate to start the conversation. Use a takeoff to finish it. Mixing those two up is how a five-minute estimate turns into a five-figure problem.