Original LeadRestore methodology

The LeadRestore Recovery Signal Framework

Old data is not automatically a sales opportunity. This is the framework LeadRestore uses to separate records that may deserve another look from records that should stay closed, paused or suppressed.

Written by Callum Hanson, Founder of LeadRestore · Published 4 September 2026 · Methodology, not a performance claim

Why publish the framework?

Lead recovery should be explainable. A business should be able to understand why an old enquiry has been prioritised before anyone contacts the customer. LeadRestore is designed around that principle: AI can help organise and draft, but the commercial decision and final customer contact stay with a person.

This page documents the practical signals behind that approach. It does not claim that any one signal guarantees a sale, and it does not turn every historical contact into a lead.

The eight recovery signals

1. Unfinished intentThe strongest records often contain a clear next step that never happened: a requested callback, revised quote, finance option, installation date, design change or promised answer.
2. Specific buying questionsQuestions about price, timing, availability, finance, specification or delivery can show that the person progressed beyond casual browsing.
3. Evidence of real engagementMultiple meaningful exchanges, a quote, survey, consultation or detailed requirement can carry more signal than a name collected with no conversation history.
4. Reason the opportunity paused“Not yet”, timing, budget timing or a temporary delay is different from a definite no, completed purchase elsewhere or explicit opt-out.
5. Recency in contextAge matters, but it is not judged in isolation. The sensible window for a fitted kitchen, recruitment brief or home-improvement quote is different from a short-cycle purchase.
6. Opportunity valueValue helps a team decide where human follow-up time is most commercially sensible. It does not make a weak record strong by itself.
7. Quality of the historic notesUseful notes let the next message refer to what the customer actually cared about. Thin or contradictory notes reduce confidence and should reduce automation.
8. A legitimate next actionA record is more useful when the history suggests a small, relevant next question rather than a generic “are you still interested?” chase.

The exclusion layer comes first

Prioritisation should never override suppression. Records that are completed, clearly lost, opted out or otherwise inappropriate to contact should be excluded before recovery scoring. Businesses remain responsible for lawful basis, consent where required, their own do-not-contact rules and any sector-specific obligations.

The framework is designed to answer one question: is there enough unfinished commercial context here to justify a careful human review?

How the signals work together

LeadRestore does not treat the framework as eight independent tick boxes. A recent enquiry with no buying context may be weak. An older quote with detailed notes, a clear delayed timeline and an unfinished next step may deserve more attention. The useful signal comes from the combination.

That is why LeadRestore returns a priority, a reason and a suggested next message rather than simply producing a list of every old contact.

What the framework deliberately does not do

It does not predict guaranteed revenue. It does not automatically send messages. It does not treat silence as consent. It does not invent missing history. And it does not replace the judgement of the team that owns the customer relationship.

As LeadRestore gathers genuine, appropriately anonymised product evidence, we intend to publish aggregated research about which recovery patterns appear most often. Until then, this methodology is published as the operating framework, not as a claimed conversion study.

See the framework in practice

Try LeadRestore before uploading your own data.

The 30-second example uses fixed sample enquiries and needs no email, upload or card. If the logic makes sense, the Free Audit can then analyse up to 50 of your own old enquiries.

See the 30-Second Demo