Why Your Development Appraisal Is Probably Wrong

We've reviewed hundreds of development appraisals. More than half contain the same structural error.

Net Developable Area

Everything flows from NDA. Get this wrong and everything that follows is wrong.

The first measurement on any site is square feet per net developable acre. The benchmark for two-storey housing is approximately 15,500 sqft/acre. If the number is significantly off, something is wrong with the layout or the NDA assumption.

Most appraisals never check this.

Gross site area minus deductions — highways, public open space, SuDS, strategic landscaping, community land — equals Net Developable Area. That's the foundation. Every revenue calculation, every levy calculation, every density assumption flows from it. If your NDA is overstated by 10%, your entire appraisal is optimistic by 10%. If it's understated, you're leaving density and revenue on the table.

The square feet per acre metric tells you immediately whether a site is under-plotted, over-plotted, or in the right range. Sometimes a little more, sometimes a little less, but 15,500 is the ballpark for standard two-storey housing. Three-storey plots higher but sells worse. Bungalows plot at half the density but command £100+/sqft premium.

Before you run a single revenue or cost calculation, verify the NDA. Walk the site. Check the highways drawings. Confirm what land is actually developable. This single number is the gateway to everything else.

The Mix

An old colleague once told me: the value is in the mix. He was right.

But most appraisals accept the architect's mix as given. They never test whether fewer, larger detached houses outperform more, smaller semis — even when the maths clearly says they do.

A 50 sqft change on a two-bedroom house shifts the £/sqft and changes which types are optimal. A three-bedroom semi at 950 sqft might hit the ceiling price at £285,000. Make it 1,000 sqft and the price stays at £285,000 — you've added cost with no revenue gain. Reduce it to 900 sqft, separate it into a detached, and suddenly it's worth £295,000 at better margin.

People will pay more for a smaller detached house than a larger semi-detached house at the same headline price. The market values separation over size, consistently.

Most appraisals test one scenario. The architect draws a layout, the surveyor values it, the appraiser inputs it. Nobody asks: is this the optimal mix, or just the first feasible mix?

The LVA tests thirty scenarios. Different house types. Different sizes. Different tenure splits. Different densities. It holds every variable — planning policy, market ceiling prices, build costs, levy structures — and finds the configuration that maximises residual land value.

Not maximum GDV. Maximum residual land value. The two are not the same.

Ceiling Price Blindness

Every house type in every location has a ceiling price — the maximum the market will pay regardless of size.

Oversizing beyond the ceiling is pure value destruction: build cost increases but revenue doesn't.

Architects instinctively maximise gross internal area. They want to give buyers more space, more rooms, more features. The instinct is well-meaning. The financial impact is often disastrous.

A four-bedroom house in a £400,000 market might have a ceiling at 1,500 sqft. Build it to 1,700 sqft and the price stays at £400,000. You've added 200 sqft of build cost — perhaps £40,000 — with zero revenue uplift. The £/sqft has dropped from £267 to £235. The residual land value has fallen.

This happens on scheme after scheme. Developers build oversized semis that bump into ceiling prices. The same revenue could have been achieved with smaller, cheaper units. Or the same square footage could have been reconfigured into detached units at higher £/sqft.

The fix requires market research. Not automated valuations. Not regional averages. Actual phone calls to local agents. What does a 750 sqft two-bed achieve? A 1,050 sqft three-bed detached? A 1,500 sqft four-bed? Map the ceiling prices for every type in the local market, then design to those ceilings.

Ceiling price blindness is often the single most expensive mistake on a site.

Less Is More (But Not Always)

Government levies — CIL, S106, BNG, affordable housing, public open space — are unit-count-based.

Fewer larger units means the same total sqft but fewer levies. The savings can be dramatic.

A 100-unit scheme with 30% affordable housing delivers 30 affordable units. Reconfigure the same developable area into 85 units and the affordable obligation drops to 25 or 26 units. Same square footage. Five fewer affordable units. On a site with £15,000/unit S106 and £8,000/unit CIL, that's over £100,000 of levy savings before factoring in the reduced affordable land take.

The 'Less Is More' principle works when unit-count levies are high and market values are strong. Reduce units, increase size, push up the £/sqft, and the residual land value increases despite lower GDV.

But this doesn't apply universally.

In lower-value areas, sometimes you need more units, not fewer. If you're achieving £180/sqft and the levy burden is modest, adding units increases revenue faster than it increases costs. And sometimes sites are genuinely under-plotted — the layout has been overly conservative and there's capacity for more homes without breaching density policy.

The maths tells you which applies. Test both directions. Run the numbers at 90% of the proposed unit count and at 110%. If land value increases in one direction, explore further. If it peaks in the middle, you've found the optimum.

Don't assume less is always more. Test it.

Finance Model

On one recent FVA, switching from simple interest to monthly cashflow modelling reduced finance costs from £1.12M to £443K — a 60% saving that changed the viability conclusion entirely.

Simple interest overstates cost on every phased scheme.

The formula most appraisals use: total build cost × interest rate × build period. If you're building £30M of housing over 2.5 years at 7% interest, that's £5.25M of finance cost. Except it isn't. Because you don't borrow £30M on day one. You draw down construction finance progressively — month by month as the build advances. And you start repaying debt as soon as units sell.

Proper cashflow modelling accounts for this. Month 1: borrow for groundworks and infrastructure. Month 6: borrow for first building phase. Month 18: first sales complete, debt begins reducing. Month 30: final units sell, loan clears.

The average outstanding debt is nowhere near the total build cost. It peaks mid-build and falls rapidly during the sales period. The actual interest paid is 40-60% lower than the simple interest formula suggests.

If your appraisal uses simple interest, your numbers are wrong. And if you're submitting a Financial Viability Assessment to argue for reduced affordable housing or CIL, the District Valuer will challenge it immediately. They know the difference. Your appraisal needs to reflect reality, not shortcuts.

Monthly cashflow modelling isn't exotic. It's standard practice in corporate finance. Development appraisals should be held to the same standard.

GDV Does Not Equal Land Value

Increasing GDV does not automatically increase land value.

Build costs, S106, profit margin, and affordable housing can absorb the entire uplift. The correct optimisation target is maximum residual land value, not maximum GDV. The two are not proportional.

I've reviewed appraisals where an architect added an extra floor of flats — increasing GDV by £2M — and the residual land value dropped by £400K. Why? Because the additional build cost (£1.8M), the additional CIL (£250K), the additional S106 (£180K), and the additional affordable housing obligation (4 more units at transfer value) consumed more than the revenue increase.

The developer would have been better off building less.

This is the fundamental error in most appraisals: they optimise the wrong variable. They assume more units or more GDV equals better outcome. It doesn't. What matters is residual land value — revenue minus costs minus profit minus obligations. That's the number that determines what you can pay for the site and whether the scheme stacks up.

Every change to the scheme — more units, fewer units, different mix, different tenure — should be tested against its impact on residual land value, not GDV. If you're chasing GDV, you're chasing the wrong number.

Before You Submit Your Viability or Make Your Bid, Test It

Not against one scenario — against every scenario the site could support.

Test more units. Test fewer units. Test different house types. Test different sizes. Test the affordable housing percentage. Test the finance model. Cross-check every assumption — land value, build cost, sales rates — against multiple sources.

The numbers matter too much to get wrong.

Most appraisals are professionally competent and internally consistent. But competent and optimal are not the same. An appraisal can be thorough, well-presented, and defensible — and still leave £500K or £2M or £7M on the table because the mix wasn't tested or the density wasn't challenged or the finance model used the wrong formula.

If you're about to spend £5M on a land deal or submit a viability case that determines whether your scheme is viable, the cost of getting independent verification is trivial compared to the cost of getting it wrong.