Get Started

Turn new planning approvals into paid leads for builders and the specialists they need first

Every week, councils publish approvals for shopfronts, extensions and lofts, each with conditions that need a specialist before work can start. Nobody sells those firms a list. Build a script that turns a week of official open data into a scored digest, and sell it as a subscription firms can say yes to on one call.

190 minIntermediateTools: Free data and tools. Optional postage at about £1.10 a letter, and possibly the ICO fee (£52 a year)Step 1 free
A weekly lead machinea tested script that pulls approved and new applications from official open data (or your hand-copied list), drops the junk, tags each job with the trades it needs, scores it and writes a one-line summary
The law, in plain Englishwhy a planning register is public but a home address is still personal data, and which business model keeps you on the right side of UK GDPR and PECR
A commercial digest firms subscribe to, priced against what Bark, Checkatrade and planning-alert services charge, plus a letter service and an optional opt-in homeowner add-on
  • The conditions angle: approvals turned into work for arboriculturists, drainage engineers and plan writers, with an AI prompt that sorts who's needed first
  • A kit: messages to trades, the homeowner letter and free checklist, legitimate interests assessments, privacy notice, terms outline, pricing calculator, scoreboard and trackers
Before you start
  • uv, to run the script. Step 1 of Make a command-line tool installs it. You don't need to know Python
  • Optional: Ollama with a small model, for the AI summaries on your own PC (Run a ChatGPT-style AI on your own laptop). Without it, the script writes plain summaries from rules. The AI summaries use gemma4:e2b, which needs Ollama 0.20.7 or newer
  • One area you know: a London borough or two, or any council whose weekly list you're happy to read by hand
  • About 6 to 8 hours a week, a phone, and (for homeowner letters) a printer and stamps
01/09Why a planning approval is the warmest lead a builder can get0 done
STEP 01 of 09~20 min

Why a planning approval is the warmest lead a builder can get

See the size of the opportunity, who's already selling it, and what it takes to run.

  1. A planning approval is a homeowner who has already decided to spend moneyThey've paid for drawings, paid the council fee and waited eight weeks. Now they need builders, and they need them soon. The approval is published on the council's register the day it's issued. Most trades find their work through word of mouth, Checkatrade or Bark, where they pay for leads shared with several rivals. Very few read the planning lists themselves.

  2. The numbersCouncils in England decided 145,600 householder applications (extensions, loft conversions and similar) in the year to June 2026, half of all planning decisions, and granted 87% of all decisions (gov.uk planning statistics, April to June 2026 ↗). In London alone, the Planning London Datahub showed 1,281 approved householder decisions made between 1 September and 9 October 2026.

  3. And the money behind each oneThe Federation of Master Builders puts an average single-storey extension at £1,800 to £3,000 per square metre (FMB, March 2026 ↗). A modest 20m² extension is a £36,000 to £60,000 job before the kitchen. A builder who wins one extra job a quarter from your list has paid for a year of it many times over.

  4. You're not the firstNational services already sell planning alerts: Planning Signal ↗ charges £49 + VAT a month for keyword alerts, and PlanWatch ↗ sells one trade's patch, with addresses, for £149 a month. That's proof trades pay for this. It also means "I'll send you planning applications" isn't enough. Your edge is the part they automate away: reading every row, a summary a builder understands in five seconds, local knowledge, the specialists who need the conditions (step 7), and, if you add it, homeowners who've actually asked to be contacted (step 8).

  5. What it takes.

    • Sorting is the product. For Barnet, 1 to 8 October 2026, 68 applications were decided; 22 were tree works or paperwork on older permissions, 8 were refused, and many of the rest were small jobs like windows, porches and driveways. Reading all of that so a builder sees only the good ones is exactly what firms pay for, and the script in step 3 does the heavy lifting.
    • Letters are an add-on, not the core. There's no published reply rate for planning letters; the industry tracker JICMAIL put cold direct mail at a 0.9% average response in its 2024 benchmarks (British Print, April 2024 ↗). Plan on 1 to 3 replies per 100 letters, and lead with the commercial digest.
    • Phone first. Trades are on site all day and wary of lead sites, so calls beat email. Plan to contact 100 or more firms to sign the first few.
    • The law is part of the job. A home address plus "approved loft conversion" is personal data, and the ICO can fine up to £17.5 million under PECR. Step 4 sets out the clean way to do it.
    • Timeline: a week to get the data and script running and make samples, then several weeks of steady outreach before the first paying firm.
    TipWho this suits: you're happy reading a hundred planning descriptions a week, you're comfortable phoning businesses, and you want to sell useful digests, not bulk lists of home addresses (step 4 explains why). About 6 to 8 hours a week covers it. Keep your day job while you sign the first firms.
  6. Everything in this mission is in one kit: download the kit (zip), or open the pieces: kit.md (models, offers, scope, terms, proposal, onboarding, delivery checklist, routine, 30-day plan, prompts), outreach-templates.md, homeowner-next-steps.md, legitimate-interests-assessment.md, privacy-notice.md, pricing-calculator.csv, scoreboard.csv and the script, planning_leads.py.

✓
CheckpointYou've picked one area (one or two London boroughs, or one council outside London) and written down which trades you'd most like to sell to.
If it breaksWhich area should I pick? + 1 more
Which area should I pick?
One you know, with lots of extensions and lofts: suburban boroughs and towns with older houses. In London, check the datahub counts for your borough first (step 2): some boroughs send far more data than others. Outside London, pick a council whose weekly list is easy to read.
Can't trades just read the lists themselves?
They can, and almost none do: the lists are long, unfiltered and written in planning language. You're selling the hour of reading and sorting they won't do, plus a homeowner who wants to hear from them.
Still stuck? Ask the community
Next mission · Make money with AIBuild a Shopify app and earn monthly from the App Store
→