If you ran the free kit, you have a sheet with a Flagged column and a bad feeling about it. Three, maybe five out of every ten businesses came back unverifiable, and the honest advice was: do not mail these. That advice was correct. It was also incomplete. Most Flagged leads are recoverable — and of the leads that survive, one is worth ten of the others. This kit does both jobs.
The free kit's Scout checked the website — and small owner-operators frequently don't put their own name on their website. But owners hide in public records: the Secretary of State's business registration, the state trade licence board, the chamber of commerce, LinkedIn, the archived About page that existed in 2019. Real street addresses hide on Google Business Profiles while the website shows a PO Box.
The Fixer looks where the first pass didn't — without ever lowering the standard. And then the Priority Pass answers the second question the free kit couldn't: of the leads that survive, which do you write to first? A plumber paying Google forty dollars a lead and letting half of them ring out is worth ten quietly-fine shops. On the free kit's sheet they look identical.
The recovery playbook. Targeted, gap-by-gap research routes — state registrations, licence boards, archived pages — that repair Flagged rows into mailable leads with proof URLs. Recovery rates run highest on address gaps; a lead that survives unfixed genuinely can't be verified from the open web, and now you know it.
HOT / WARM / COOL — decided the way an experienced salesperson would, by walking a decision tree: who already pays for calls, whose reviews complain nobody answers, who is busy and small. Plus the pitch line to open with, chosen from the evidence.
Everything the free kit tracked plus demand signals, recipient routing, decision-maker email, source trails — growing to 24 columns by the last pass. Built for you, in Excel, by the prompts. You never type a header.
The full-schema versions of the free kit's prompts, with the still-trading check, demand signals, and self-audits that force every claimed fact to carry a proof URL — or lose the claim.
Lane A batches ten businesses a sitting for screening; Lane B runs one lead at a time for the letters you'll actually send. Every prompt written twice — you copy, paste, and upload the file it hands back.
Works on claude.ai, free or paid plan. No scrapers, no list vendors, no subscriptions. The only rule is the one the whole method stands on: a fresh chat for every prompt, and never talk it out of a COULD NOT PROVE.
This is a digital product — nothing is shipped. No PDF, nothing to download either: the kit is a private page on this site that unlocks the moment your payment lands, plus a personal access link in your inbox that opens it on any device, work anywhere. You pay once; there’s no subscription and nothing renews. Keep it open next to Claude while you work.

$47
one-time · instant access · no subscription
£37 in the UK · €44 in Europe — you’re charged in your own currency at checkout.
You get access the moment payment lands — the full kit opens on this site, and your receipt and access link arrive by email. Checkout is handled by Digistore24 as merchant of record. Run the free kit first and buy this only if you want the recovery and ranking passes. Our refund policy covers you either way.
Run it first, yes — it's free, it's the same Finder → Scout → Checker spine, and it's how you find out whether this method suits you before spending a dollar. The Starter Kit contains the complete pipeline (including upgraded, full-schema versions of those three prompts), so nothing breaks if you start here — but the free kit is the honest try-before-you-buy.
Neither. It's a page on this site that unlocks when you buy — prompts with copy buttons, the playbooks, the decision tree, the troubleshooting table. Your access link opens it on any device. It reads in about twenty minutes and then works as a reference while you run batches.
Ask within 60 days and you get your money back. No questions, no conditions, no hoops — just email us and we'll process it. Your access is switched off and that's the end of it. We'd still rather you run the free kit end to end first and buy this only if you want what it adds, but if you're not happy you're not stuck. Full terms are in our refund policy.
Then it genuinely can't be verified from the open web, and mailing it would be a guess — the kit tells you that honestly instead of inventing an owner name. Expect to recover a chunk of your Flagged pile, not all of it. Recovery runs higher for address problems than decision-maker problems, because addresses are recorded in more public places than people are.
No — a free plan works. The batch lane is light on usage; the one-lead-at-a-time lane uses roughly ten times more, so on a free plan you might hit a limit mid-batch and wait it out. Every prompt needs web search and code execution switched on; the kit shows you both toggles.