What actually gets approved in London’s small-site planning

The independent evidence base for small-site planning outcomes in London: what gets approved, and what a site is therefore worth. 15,533 applications across 33 London boroughs, 12,458 decided, every refusal reason classified and evidence-graded. The same answer whoever is asking, developer or seller, agent or authority, in place of the comparable-led guess that usually stands in for it.

For councils + registered providers

GLA Small Sites Release Fund. Submissions close 10 July.

Comparable-evidence data for site investigations, due diligence, and feasibility. £250,000 available per applicant. The fund pays for evidence, not application drafting.

See the evidence offer →

The current acquisition model is buy first, learn second. You commit £15,000 to £25,000 on drawings, consultants and a planning submission, then wait eight weeks at best. In Hammersmith & Fulham the median runs to twenty-two. The whole exercise is a punt on a question the borough already has a settled view on.

43% of London small-site applications are refused. Half miss the eight-week statutory deadline. In Havering, Croydon and Waltham Forest the refusal rate runs closer to seven in ten. Most of that outcome is, in principle, knowable in advance.

Of the schemes that do get through, 1 in 4 loses units during determination. The typical cut is 2 units. At a six-figure residual per unit, that’s the margin of the whole acquisition.

Published policy tells you what a borough wants.
The data tells you what it approves.

37%
Where beats what
Inner London refuses 37% of small sites. Borough is the dominant variable, before design or agent.
Verified
−42%
Cross unit 10, residual falls
Affordable-housing obligations bite at the 10-unit threshold. Per-unit residual drops 42% the moment you trigger them.
Verified
Pre-app is the biggest lever
Same scheme, double the approval odds. Consistent across every borough with enough data to test.
Verified
25%
One in four lose units
1 in 4 approved schemes lose units during determination. The typical cut is 2 units, and which schemes it happens to is largely knowable in advance.
Verified
Track record

We publish how well we predict. Misses included.

A retrospective hold-out, plus a tamper-evident ledger of calls committed before the council decides. The same evidence standard, turned on our own predictions.

See the track record →
First look
Sample Site Assessment

See the depth before you commit.

£125 · 24 hours
A real, anonymised Site Assessment from a recent London project, delivered straight to your inbox. All sections present. £125 credited against your first full Site Assessment if commissioned within 30 days.

All sections present. Anonymised real project. Credit applied to a full Site Assessment within 30 days.

Buy a sample →
For a borough
Borough Intelligence Report

What does this borough actually approve?

£1,500 · or £3,500 for three
A full intelligence read of one borough: approval rates by site type and area, refusal taxonomy, unit-count and density sweet spots, case studies, process intel. Pick three boroughs and the bundle saves £1,000.

Single borough £1,500. Three-Borough Bundle £3,500. Delivered within 5 to 7 working days.

Commission a Borough Report →
Free

What’s happening in this borough?

Free · no sign-up
Every London borough has a public dashboard: approval rates, scheme types, refusal patterns, and 13 quarters of trend. Updated each quarter from the council’s own register.

All 33 boroughs. Interactive area maps. Quarterly refresh.

Open the dashboards →
01

Every application coded

Three years of small-site decisions per borough. Each one coded by site type, area, PTAL, density, conservation status, and outcome.

02

Every document read

Officer reports and decision notices parsed for refusal reasons, conditions, case officers, and design commentary. Over 30,000 documents processed.

03

Every finding graded

Four evidence tiers based on sample size and statistical significance. Small samples are flagged openly, and every claim is qualified.

What you actually get for £450.

We tested the Site Assessment scoring against 4,646 real London decisions the model didn’t see in advance. Training stopped at December 2024; every decision since is a blind test. Here’s what held up, framed the way you’d actually use it.

3 in 4

Bid with confidence. Sites we flag Low Risk get approved. London average sits at 1 in 2.

±2 units

Bid the right number. The capacity model calls the approved unit count within 2 units, 74% of the time. Six-figure swing on residual, quantified.

£20k

Or don’t bid at all. Typical abortive cost on a refused small-site application. One £450 assessment pays itself back the first time we flag a loser.

Abre Etteh

I built the tool I wished existed.

I’ve sat on every side of this table. Designing schemes as an architect. Assessing them at a London council. Acquiring sites and managing planning as a developer. Fourteen years of watching the gap between what policy says and what officers actually approve.

The planning system publishes the data. Nobody was synthesising it. So I built Perfect Scale: every application coded, every document read, every finding graded.

Better data → better schemes → more homes.

Abre Etteh, ARB Registered Architect

The data exists. Use it.

15,533 applications across 33 London boroughs, 12,458 decided. Every finding evidence-graded. The same evidence, whoever is asking.