Development feasibility explained: from land bid to financial model

Development feasibility is the process of working out whether a proposed development makes financial sense. At its core, it answers one question: given what I expect to sell the units for, what can I afford to pay for the land?

This is called the residual land value approach — and it is the foundation of almost every development appraisal.

The residual method

Start with the Gross Development Value (GDV): the total expected sales revenue from the scheme. Subtract the Total Development Cost (TDC): construction, professional fees, finance, and contingency. The remainder — adjusted for the developer's target margin — is what the land is worth.

Residual Land Value = GDV − TDC − Developer Profit

If the residual is above the land ask, the deal works. If it is below, you either need to renegotiate the land price, reduce costs, or change the unit mix.

Where analysts spend their time

Building a feasibility model from scratch takes time. You need comparable sales data to anchor GDV assumptions. You need construction cost benchmarks for the cost stack. You need to model finance costs over the development period. And you need to run sensitivities — what happens to the margin if build costs increase 10%, or if sales prices fall 5%?

Each of these steps requires data retrieval, normalisation, and calculation. In a typical workflow, this takes a full day or more.

How AI changes the process

AI agents can compress the data-gathering and structuring phase significantly. Given a site brief, an agent can retrieve relevant comparable sales, apply the residual methodology, build a cost stack from benchmarks, and produce a sensitivity table — in minutes rather than hours.

The analyst's role shifts from building the model to reviewing it: checking assumptions, applying local knowledge, and making the judgment calls that require expertise.

That is a better use of an analyst's time.