What Does It Really Cost to Build? How Desktop Intelligence Is Changing Development Costings

A quantity surveyor gives you a cost plan. It tells you what a scheme like yours typically costs to build, based on BCIS regional averages and NRM elemental splits. Good as far as it goes.

But it does not interrogate your site. It does not know that the British Geological Survey data shows alluvium beneath your foundations, or that the Environment Agency classifies your drainage catchment as Flood Zone 2, or that Historic England has four listed buildings within your setting. Those are separate consultant instructions, weeks later, at additional cost.

By then you have already committed capital.

A standard desktop cost estimate uses BCIS rates — the Building Cost Information Service published by RICS — to estimate construction cost per square metre. These rates are regional averages, updated quarterly, and they are the industry standard.

The QS then applies NRM (New Rules of Measurement) elemental percentage splits to break the total down into substructure, superstructure, finishes, services, and externals. Again, industry standard.

What this approach cannot do is adjust for your specific site conditions. The BCIS rate for the West Midlands does not know that your particular site in the West Midlands sits on alluvium that requires piled foundations instead of standard strip. It does not know that your site is in a conservation area where the planning officer will require natural stone and slate instead of brick and tile.

These site-specific cost impacts — what the industry calls abnormals — are discovered later in the development process, typically when specialist consultants report back. By then the land deal may already be agreed, the option exercised, or the planning application submitted with a viability assessment based on incomplete cost data.

On a recent 28-dwelling scheme in the West Midlands, the standard BCIS estimate came in at approximately £4.5 million construction cost. Reasonable for the region, the unit mix, and the gross internal area.

But the desktop intelligence identified four site-specific cost impacts that a standard cost plan would not have flagged at feasibility stage:

**Piled foundations** — BGS geology data showed alluvium and superficial deposits requiring piled foundations rather than standard strip. Cost impact: significant uplift to the substructure element, calculated from actual foundation perimeter quantities rather than percentage estimates.

**Heritage specification premium** — Four nearby listed buildings within the setting meant the local authority would require heritage-grade external materials. This applies a 15-18% uplift to external walls, roof, and windows.

**Mandatory BNG costs** — Natural England data and habitat mapping identified biodiversity net gain obligations. At 10% mandatory BNG, the estimated cost was £140,000.

**S106 sustainable transport** — The cross-domain intelligence flagged likely S106 contributions for sustainable transport infrastructure at approximately £84,000.

Total abnormal costs identified: £970,000 — nearly 20% of the base build cost that a standard QS estimate would not have captured at desktop stage.

The difference is data. When you feed 130 authoritative datasets into the cost model — geology from BGS, flood risk from the Environment Agency, heritage constraints from Historic England, ecology triggers from Natural England — you get a cost estimate that reflects your actual site, not just your region.

Foundation type is derived from geology, not assumed. Drainage costs are adjusted for actual flood risk. Specification premiums are applied where heritage or landscape constraints exist. Abnormal costs are quantified from cross-domain intelligence that no single consultant would produce.

The result is a 14-element NRM-aligned cost breakdown with low, central, and high bands — plus a site-specific abnormal cost schedule that flags every cost risk the desktop data can identify.

For a developer evaluating a site, this changes the feasibility calculation. You can make land acquisition decisions with cost intelligence that reflects your actual site conditions. You can negotiate option agreements knowing what the abnormals are likely to be. You can present credible numbers to funders without waiting for specialist reports.

You are not guessing. You are not hoping the ground conditions are straightforward. You have desktop evidence that either confirms your assumptions or flags the risks before you commit capital.

This is a desktop indicative estimate, not an RICS-compliant cost plan. It does not replace a quantity surveyor for RIBA Stage 3+ detailed costing. Ground conditions must be confirmed by site investigation. Heritage requirements must be confirmed with the conservation officer.

But at feasibility stage — when you are deciding whether to commit capital to a site — it gives you cost intelligence informed by your actual site data, not just regional averages. That is the difference between a cost plan and cost intelligence.