Product strategy · May 2026 · Confidential
Two funnels. One platform. The strategy on one page.
Two states of mind. Three commercial tiers. Two cornerstone anchors.
The intent moment
Act.
With legalhelp.
Something has happened. The brown envelope. The court letter. The Section 21 notice. The user is being pushed into action; the signature gives them the verb.
Maps to: parking · CCD · debt · eviction · deposit · consumer · subscription · benefits
The avoidance moment
Sorted.
With legalhelp.
Something has been on the to-do list for years. The will. The LPA. The contract you never quite wrote. The user has been meaning to act; the signature gives them the completion.
Maps to: wills · LPA · contracts · subscription audit
Four-stage funnel. Each stage paired 1:1 with the platform mechanism that retains it.
Socials and content across our channels. Rob amplification + charity partner reach + SEO authority + PR. The followers we need to seed the funnel.
High-quality landing pages and on-site content that convert audience members into visitors. Moment-of-need SEO targets; life-event triggered campaigns.
Free products are the conversion lever. Registration is the cost of getting the output; every visitor who completes a free flow becomes a member we own.
Convert members into paying customers with high-quality paid products. Freemium intent (AoS → Defence) and premium cross-sell (wills → LPA → estate planning).
Every concept product on one screen — the menu the customer chooses from.
Channels feed the top; the funnel terminates at Act / Sorted.
Defend yourself. With legalhelp.
★ Cornerstone product
County Court Defence
Defend yourself. With legalhelp.
The brand-defining product on the intent side. The flagship around which Rob's brand and the intent-funnel marketing concentrate. The next four slides cover the opportunity, the regulatory line, the UX architecture, and the competitive position.
Cluster: CCJ Set-Aside · Directions Questionnaire · Disclosure / Part 18 / s.78 CCA pack · Bundle / trial pack · possession defence (s.8 / s.21) · employment tribunal pack · SME claim issue
The system has a structural defendant gap. We close it.
Debt buyers acquire portfolios at 2–5p in the pound. Unit economics depend on volume default. Properly contested cases push per-case cost from £15 to £300–£500.
Citizens Advice — capacity-starved. StepChange — debt-only. MSE forums — peer-to-peer. McKenzie Friends — manual.
Judge Rinder was built on exactly this — the consumer underdog against the institutional bully. The brand stack does its hardest work here.
LSA 2007 s.14 is the bright line. Holding it is a UX problem, not a regulatory one.
Drafting documents for a litigant-in-person is not reserved. Issuing, signing, or conducting on the user's behalf is. The McKenzie Friend / Citizens Advice / direct-access-clerk pattern, codified for AI.
Signature. Address for service. Court-fee payment. Hearing advocacy. Each needs a specific UX answer that keeps the user as the named party on the record — never the platform.
The architecture
User-as-actor. We draft. The user reviews. The user submits. The user is the litigant.
The user-as-actor architecture, in three layers.
AI generates the defence as a tracked-change diff. User reviews each proposed change, accepts or edits, exports the PDF.
Browser extension or one-click prefilled link populates the user's authenticated session on the actual court portal — MCOL, OCMC, Common Platform. User clicks submit on the court's own form.
Every action logged — user clicked, user submitted, AI proposed. Provable trail for any future regulatory challenge.
What the AI does NOT do
Sign documents · pay court fees from a Legalhelp account · receive court correspondence as the user · script hearings live · file in its own name.
Compounding intelligence against repeat-player opponents.
Secure your estate. With legalhelp.
★ Cornerstone product
Wills
Secure your estate. With legalhelp.
The brand-defining product on the avoidance side. The flagship around which the Sorted-side marketing concentrates. The next four slides cover the market, the execution chain challenge, the AI architecture, and the witnessing path.
Cluster: Lasting Power of Attorney · estate planning bundle · HNW estate planning · shareholder agreement · commercial lease review · employment contract pack · trademark filing
A fragmented, brand-shallow market. No operator has done the whole chain.
SRA-regulated, sync appointment, free lifetime storage. £84m revenue (+24% YoY); 500–1,000 FTEs. Strong brand, but the sync model can't take the volume tier.
Async human-checked. ~20,000 Trustpilot @ 4.9★. 400+ charity partners. Sold to Dignity Oct 2024 for £12.9m on $39m raised — standalone wills-only didn't fund the business.
Async online (£90) / phone (£170+). Storage £30/25-year vault as add-on. ~700 Trustpilot @ 4.3★. No regulated review at price.
Solicitor-partner reviewed. SSAFA et al. No storage, no execution support. Channel proven; product stops at the PDF.
The race-to-the-bottom: ActiveWills £19–£99 via affiliate — margin doesn't support quality, not where we play. The dead and consolidating: Beyond Life shut down 2025; customers migrated to Farewill. Pure-play digital wills is consolidating into the bigger operators.
A credible digital operator at £79 covering the whole chain — async + AI-drafted + AI-checked + named-reviewer co-signed + storage referral + executor handoff + probate referral. Plus Rob's halo compressing the brand-trust ramp by 18–24 months. Plus lifetime LTV via LPA + estate planning + probate referral. Nobody has done all of that.
The chain, not the document. Every link can break the will.
Where current operators leave gaps:
Drafting + async review. No storage. No witnessing.
Drafting + storage + appointment review. Doesn't scale.
Drafting + storage add-on. No regulated review at price.
Charity-aligned drafting. No storage. No execution.
All four use humans to draft or check. We are saying AI is now good enough to draft and check a legally binding and valid will end-to-end — and only call in a human reviewer to spot-check the AI's flags. That's not a price-point or a UX nudge; it's a different category of operation.
Multi-agent drafter + checker pipeline. Nobody's done it. It is doable.
Generates from structured user input.
Flags vague references, missing antecedents.
Residuary, gift-over chains, no partial intestacy.
Cross-clause coherence, executor matching.
Tries to break the will at probate.
Human will-writer co-signs in ~5 minutes on the AI's flags only — not from scratch. Cuts review cost from ~£15 (async-only) to ~£4–5. Margin works at £79 without sacrificing the senior-practitioner review gate.
No one is doing this end-to-end today. The AI architecture is doable; the brand and distribution are already being built. The window is the next 12–18 months.
Signing, storage, executor handoff, probate referral — all stitched together.
The close · Co-founder signature
Rob Rinder · co-founder