A cost plan isn't something you write once at tender stage and file away. It's a working document that follows the job from the first drawings through to the final account, and it only earns its keep if it's kept up to date as the job changes.
Plenty of contractors know that already. The problem is knowing how to build a construction cost plan that is actually usable at each stage, rather than a spreadsheet that made sense on the day it was sent and nobody trusts a month later. So this is the practical version: what goes in at each stage, how the detail builds up, and what keeps it alive once you're on site.
A cost plan isn't a one-off number
The single biggest mistake is treating the cost plan as an answer. It's closer to a record of what you know, what you've assumed and what you've priced, at a given point in time. When the drawings change, the plan changes. When the client adds a rooflight, the plan changes.
That sounds obvious. But look at how most plans are laid out and you'll see a column of element totals with nothing underneath them. There's no trail from the figure back to a drawing, a quantity or a rate. Which means the first time something moves, nobody can say which figure it moves.
How to build a construction cost plan in stages
You rarely have full information on day one, and pretending you do is how thin drawings end up priced like finished ones. Build the plan in layers instead, and be honest about which layer you're in.
- Set the scope boundary. Before a single figure goes in, write down what's in, what's out and who supplies what. If the scope was never properly nailed down at this point, every number that follows is built on sand.
- Break the job into elements. Substructure, frame, upper floors, roof, external walls, windows and doors, internal finishes, services, externals. Use the same element structure all the way through so later detail slots into the same boxes.
- Put a broad figure against each element. Early on, when drawings are thin, that's a figure per element based on what you know so far and on similar work you've priced. Mark each one clearly as an allowance, not a measured price.
- Replace allowances with measured quantities. As drawings develop, take off the quantities for each element to NRM2 and price them line by line rather than as a lump. Do the elements with the most money in them first.
- Separate labour, plant, materials and preliminaries. Keep them as distinct lines so you can see later where cost is actually moving.
- Record your assumptions and exclusions. Every gap in the information gets a note. Those notes are what you'll point to when the scope shifts.
By the time the job's on site, the plan should be detailed enough that a variation or a specification change can be slotted straight into the right element without reworking the whole thing.
What changes between the early plan and the detailed one
Say you're pricing a two-storey extension with a flat roof. At first sketch stage, you might carry the roof as a single element figure based on its rough area and a similar job you did recently. That's fine. It's a placeholder, and it's labelled as one.
Once the architect issues the roof plan and sections, the placeholder goes. In its place you get deck, insulation, membrane, upstands, outlets, trims and any rooflights, each measured and priced. The element total may barely change. Or it may jump because the sections show a parapet detail nobody mentioned at sketch stage. Either way you now know why the number is what it is, and that's the whole point of doing it properly.
That is really how to build a construction cost plan: it gets less wrong as the information gets better, and the structure never needs tearing up.
A cost plan is only as useful as its weakest element. One lump sum hiding in a measured plan is where the surprise will come from.
Keeping it live once work starts
Once the job's running, the cost plan has to be checked against what's actually happening on site, not what was assumed at tender. Actual labour output, actual material costs, actual subcontractor prices. Where the numbers drift, you want to know in week three, not at the final account, when there's nothing left to do about it.
In practice that means a short, regular comparison. Each valuation period, look at each element: what did we plan to spend, what have we spent, and what's left to spend to finish it. The third question is the one people skip. A job can be on budget to date and still heading for a loss if the remaining work has grown.
Kept this way, the plan becomes an early warning system rather than a nasty surprise at the end. It tells you where the job is heading while there's still time to do something about it. If drawings are still being revised during construction, there's more on keeping a cost plan together when the drawings keep changing.
Where cost plans usually go wrong
The most common failure isn't bad arithmetic. It's the scope problem above: nobody agreed exactly what was included, so the plan is chasing a moving target however carefully the sums are done.
The second is poor communication between whoever built the plan and whoever's running the job. The estimator knows which figures are soft. The site manager doesn't. So the two drift apart without either side noticing until it's too late to fix cheaply.
A few other patterns turn up again and again:
- Allowances never replaced. A placeholder from sketch stage survives all the way to site because nobody went back and measured it.
- Preliminaries as a flat percentage. Site set-up, welfare and supervision get added as a round figure instead of being priced against the programme.
- Old drawing revisions. Quantities taken off revision A while the job is being built to revision D.
- Variations priced in isolation. Extras get agreed on the fly and never fed back into the element they belong to.
None of these is complicated. They just need someone to own the plan and keep it honest.
One test for your next cost plan
Pick any element at random and ask: can I trace this figure back to a drawing, a quantity and a rate in under a minute? If the answer is yes, the plan will stand up on site. If the answer is "it's an allowance" and the drawings for that element were issued weeks ago, that's the line to measure next.
If you'd rather hand over the measuring, our cost planning services build the plan from your drawings, measured to NRM2 and priced on current rates local to the job.