DEFERRED 2026-05-01T18:30 per phase-fit reversal. This memory file records the original Σ.κ ★ lock decision but the lock has been DISCARDED. Will Update Architecture was prematurely framed as Phase 3 work; correct phase is Phase 6 (per-brand integration — IW). The original content below is preserved for forensic continuity + re-use when Phase 6 opens. Do not act on this memory’s recommendations as if locked.
Outcome: Rich-locked Σ.κ ★ on 2026-05-01T16:00 BST per Rich directive (“mlp use clio manage, clio grow crm, clio draft, lexus protege AI; huge opportunity to offer will conversion services at www.inheritwills.com”).
A-numbered amendment: A-143 (Tier-1 arch-state v3.28 → v3.29).
SECOND ζ.2 multi-lock reversal in single decision (after MQ-001 R2 + A-142 ζ-Q15 REVISED). Supersedes:
- Σ.η ★ Hybrid B2B-only (prior ζ-Q16 pick from earlier this session; margin 0.27 above 0.20 threshold — first ζ.2 above-threshold decision since ζ-Q14)
New locked configuration:
- Same Π.θ ★ L4M Stage-2 Extraction mechanism (per A-142) serves BOTH buyer segments
- B2B partner-firm migration: Clio Manage + Clio Draft + Clio Grow APIs primary (MLP-confirmed); Word + SharePoint secondary; LEAP Phase-1.5+ trigger; per-matter solicitor signoff MANDATORY + 10% AI-judge sample; tiered Pilot ≤50 £25-75K / Steady-State per-matter £25 / Multi-firm Enterprise unlimited; A-113 DO-per-firm preserved
- B2C InheritWills.com flagship per brand-architecture v1.16 §IW: Word + PDF + scanned + handwritten + WillSafe DIY + photographed paper wills; tiered DIY £39 / Lite £99 / Standard £299 / Premium £499 / Deluxe £999 + Living Monitoring £30-90 p.a. recurring; A-113 extended to DO-per-user
- Cross-brand ecosystem: AwkwardEstates SEO+content → InheritWills primary funnel → MyFamilyInherits cross-sell → WillsInFaith faith-led upgrade per brand-architecture v1.16 §6.2; Update Your Will Week (March) sponsorship with Association of Lifetime Lawyers (campaign umbrella); InheritKitBench v0.1 quality-gating
- Lexis+ Protégé interop NOT competition: cross-vendor Skills passive (per Π.θ ★); MLP solicitors use both; zero competitors in succession slot per T39+T47
- Customer-facing terminology lock (Rich 2026-05-01 mid-flight): “Will update” B2C / “Migration” B2B / “Will conversion” TT-internal
NO reversal of: A-142 Π.θ ★ L4M 4-stage; A-113 sovereignty (extended to DO-per-user); A-114 Cedar+Catala substrate; A-141 SSSOM verificationTier; T1 MCP architecture; brand-architecture v1.16 IW positioning (already locked; Σ.κ EXECUTES it); Cloudflare AI Gateway + Haiku 4.5.
Sensitivity: 8-option scoring; Σ.κ ★ 4.82 Rich-pick vs Σ.ι 4.78 (margin 0.04; below 0.20 threshold; sensitivity-driven). Margin Σ.κ vs Σ.η ★ prior pick = 0.27 — ABOVE 0.20 threshold (first ζ.2 above-threshold decision since ζ-Q14 margin 0.42). 5/8 perturbations preserve Σ.κ ★. Σ.η wins 2/8 under maintainability + cost priority. Σ.α wins 1/8 under “no-migration discipline”.
Susskind 2008+2023 dual-grounding (mid-session library ingestion):
- Susskind 2008 §2.1 mass-customization + §2.5 main-contractor + §4.6 open-sourcing prediction
- Susskind 2023 §22 AI fallacy rejection of LLM-mimicry + §6 rules-as-code + §1 more-for-less + §17 acquirer-narrative
- 5 concrete commercial benchmarks (£34K/sub A&O ARPU; “make money while they sleep”; Big-4 + RELX + Thomson Reuters acquirers; EY-Riverview-Law-2018 precedent; 3,000-4,000 lawtech startups)
- 7 vocabulary alignments (more-for-less / externalized / decomposition / rules as code / embedded legal knowledge / mass customization / make money while they sleep)
- 2 acquirer-narrative anchors (Big-4 + Thomson Reuters/RELX)
Methodological observations:
- SECOND ζ.2 multi-lock reversal: supersedes Σ.η ★ Hybrid B2B-only prior pick from earlier today.
- FIRST ζ.2 above-threshold decision since ζ-Q14: margin Σ.κ vs Σ.η ★ = 0.27.
- BREAKS recent below-threshold sensitivity-driven decisions pattern: commercial-moat priorities outweigh maintenance/cost.
- Brand-architecture v1.16 already-locked positioning EXECUTED, not amended: Rich’s directive elevated architectural priority of an already-locked product positioning. Σ.κ specifies how locked branding is technically realised.
- Susskind 2023 grounding via already-indexed library (notes-complete since 2026-04-21) — no book purchase; Rich-noted “old book” expectation accurate.
- Customer-facing terminology lock (“will update”) aligns with Update Your Will Week 2026 campaign umbrella for marketing-channel synergy.
- 17 of 18 ζ.2-related decisions remain convergent.
5 NEW risks (in risk register additions):
- R-NEW-IW-LAUNCH-DEPENDENCY (IW Q4 2027 launch deliverable)
- R-NEW-CONSUMER-UX-CRAFT (B2C UX investment)
- R-NEW-LIVING-MONITORING-CHURN (subscription churn management)
- R-NEW-B2C-EXTRACTION-PRECISION (L4M precision on consumer-uploaded artefacts unverified; richard-task #208 spike)
- R-NEW-TRIGGER-EVENT-DATA-DEPS (UK life-event data sources required for Living Monitoring; richard-task #209 Phase-1.5+)
Risk register count 35 → 40.
**10 NEW richard-tasks 205-214**: IW product positioning + Clio integration + B2C UX + L4M precision spike + trigger-event detection + Update Your Will Week sponsorship + Lexis+ Protégé positioning + WIF spec + AE content + MFI cross-sell.
Spike file: ~/off-github/library/projects/inherit/T-will-conversion-product-line-spike-2026-05-01.md v1.2 (B2B+B2C unified architecture + Susskind 2008+2023 dual-grounding + customer-facing terminology lock).
Cascade artefacts: arch-state v3.29 A-143; IK arch-state v1.3; IAS arch-state v1.4; INDEX v2.19; this memory file; active-work-log; richard-tasks-tt v1.16→v1.17.
Phase 3 status: 8 Q answered + cascaded (ζ-Q9..ζ-Q16-REVISED); 1 deferred (ζ-Q15-original t-box layering → Phase 5); ζ-Q17+ candidates surfacing.
NEXT: ζ-Q17 candidate = Living Monitoring trigger-event detection architecture (richard-task #209 elevated to Q-question; gates Σ.κ ★ recurring revenue moat; UK GDPR + A-113 sovereignty implications + cross-brand IW+MFI Living Monitoring tier integration).