← Q3 journey review packs

Journey 7 council assessment (Lapsed / churned: triggered win-back, REC-03)

Verdict

RECOMMEND for Gate-1 strategy review as authored-in-parallel. The evidence-to-choice chain is the strongest of the RECOVER packs: every design choice traces to a confirmed figure (the 7-15x propensity gradient, the ~1% ceiling, the -11.8% organic-returner precedent, push 1.12% vs email 0.75%), and the pack refuses to sign the refuted 1.5% full-pool target. It is honest that it cannot ship: pin D1 is launch-blocking and the build waits on the hire. The council's changes remove two cross-pack contradictions and one ruling-conflict, and add the one lens the pack had skipped entirely: Legal/Privacy on contacting the long-lapsed.

Strongest points (keep, do not dilute)

• The restated objective with the double-read: within-band holdouts (journey lift) separated from REC-T01 (targeting value). Without the within-band cell a positive read could be entirely "the model found people coming back anyway".
• Suppression as a treatment and a deliverable: the LOW band, the organic-returner hard suppression, and REC-14's genuine end-state give the programme value even where the offer fails.
• One clock: entry only via the day-60 journey_state hand-off (Sheran-ruled window model), never a second day-count.
• The landing thesis: route deep-link + REC-21 as the conversion surface of every send (73.1% negative cold-landing read), with three packs converging on J7 ownership (DL-014).

Changes applied (each: what · why · who · where · capability · outcome · action)

1. Email consent-age floor added (DL-J7-13, new). What: no outbound email to customers with no engagement and no consent refresh in 24 months; push carries the deep tail where held. Why: email to this pool rests on consent or the PECR soft opt-in taken at purchase and both decay with dormancy; 296.3k churned >240d (confirmed) sit in the backfill's path and no Legal read exists in the repo. Who: the backfill's deep tail; hand-off entrants (60d) are unaffected. Where: eligibility (strategy.md 3.1), channel rules, decisioning matrix, TO-BE map backfill node, Analytics sizing ask. Capability: a consent-age attribute in the backfill audience logic. Outcome: deliverability and regulatory exposure priced before the deep tail is emailed. Action: Sheran commissions the Legal read (delta, new item).
2. Anniversary interim governance aligned to the DL-006 synthesis. What: the REC-30 anniversary slot executes under Journey 9's wave governance (per-cycle anniversary/seasonal mutual exclusion) until D2 lands, then comes home to J7; J7 keeps the trigger's definition and end-state. Why: J7 has no live machinery before D2 while the wave engine ships this month; both packs already agreed the end state; removes CROSS-JOURNEY-REVIEW item 3. Who: churned users whose anniversary falls in a wave window (double-hit risk). Where: Episode 2 (strategy 3.2, brief tab, map, walkthrough, Reference contact rules). Capability: anniversary_of_last_purchase (D5) + wave suppression stack. Outcome: no double contact in a cycle; a clean REC-30 read. Action: Sheran's ruling on DL-006 stays open; the pack no longer contradicts J9's.
3. Decommission home ceded per the DL-007 synthesis; midweek-warm re-home withdrawn. What: the M44/M45 decommission vehicle sits under J9's REC-06 re-scope; J7 consumes the retirement schedule; no re-home of midweek-warm (the 14 Jul STOP-both ruling governs). Why: a Sheran ruling outranks pack preference; the earlier re-home read against it; one home ends the two-owners defect (CROSS-JOURNEY-REVIEW item 4). Who: the M45P blast audience during wind-down. Where: strategy section 5, brief-tab AS-IS/TO-BE row, Process tab CRM actions. Capability: the sequenced retirement plan with per-canvas holdouts (Eng/CRM). Outcome: REC-03 never stacks on a legacy send; the £44k-attributed optics handled by wind-down holdouts. Action: Sheran rules DL-007's final home.
4. Capped-yield economics stated. What: ~1% of the ~317k contactable pool is roughly 3.2k reactivations a year, ~£0.7M at £219.94, labelled assumed (derived from confirmed unit figures on an unreconciled pool). Why: the Executive Sponsor lens needs the prize sized, not only the target refuted; consistent with the campaign's assumed £0.7M line (DL-009). Who: n/a (a governance figure). Where: Objective and KPI (brief tab) + data lens (strategy 4). Capability: none. Outcome: Gate-1 signs a target and a prize that reconcile. Action: none beyond DL-009's reconciliation.
5. Fourteen-point contact-completeness index added (Reference tab). What: all fourteen personalisation/contact fields stated with the home of each; points 7 (preferences and consent, now including consent age) and 10 (required personalisation level: 1:1 route + live price + band register, designed segment-level fallbacks) were previously implicit and are now explicit. Why: review-standard requirement. Who/where/capability: n/a. Outcome: reviewers can audit completeness in one place. Action: none.
6. Smaller writing-standard fixes. L3 class C glossed in plain English (bottom rung of the send-priority ladder); REC-21 convergence (three packs, DL-014) cited on the brief tab and strategy; the Process tab no longer asks for a Gate-1 nod on the two-touch cap (closed, see below); decision references updated to DL-J7-01 to 13.

Working assumptions adopted (escalation discipline: decided, labelled, moved on)

• 24-month email consent-age floor pending the Legal read (DL-J7-13).
• Anniversary slot under J9 wave governance until D2 lands (DL-006 synthesis).
• No midweek-warm re-home; decommission vehicle with J9 (DL-007 synthesis / 14 Jul ruling).
• Two-touch hard cap (DL-107 closed: propagates the ruled J8 shape; flag to the master retained).
• usual_route 12-month staleness rule with dropped-line fallback (DL-106 closed: definition-level, fallback exists either way).
• Double-read design (DL-103 closed: the only design compliant with the standing 9 Jul measurement ruling; Analytics owns sizing).

What genuinely requires approval (unchanged or sharpened, all open in the log)

• DL-101 entry contract: the ruled window model plus the agent-designed metered backfill (now interacting with the consent floor). Sheran with Analytics + CRM.
• DL-102 denominator restatement of the 1.5% objective. Sheran with Analytics; do not sign the refuted target.
• DL-104 band cut-offs fixed before launch (anti post-hoc). Analytics proposes, Sheran ratifies.
• DL-105 REC-14 offer depth and incentive budget ("no offer, suppression only" is an allowed outcome). Sheran + Finance on Analytics sizing.
• DL-006 / DL-007 / DL-014 the boundary-rule cluster: anniversary conflict, decommission home, REC-21 ownership (three-pack agreement, an easy ratification). Sheran.
• DL-J7-13 the Legal read on the consent-age floor. Sheran commissions.
• Pin D1 date (the D2 churn-score family) and the Braze hire: the two gates; deliver or date them.

Evidence gaps (REQUIRES VALIDATION)

1. Pool reconciliation is the top gap: ~600k vs 597,561 vs ~317k contactable vs 1.59M Churned, plus 29.8% computed vs 32% briefed on the 7,979 Lapsed-PU cohort (DL-009; owner Analytics/Ivi). Every economics line in this pack inherits it.
2. Band sizes and REC-T01 power per band (underpowered-band risk pre-flagged; owner Analytics).
3. The consent-age split of the backfill: how much of the deep tail the floor removes (owner Analytics, with the Legal read).
4. Braze content-card instrumentation gap (impressions/clicks not registering, audit): REC-21's read depends on the fix (owner CRM/Eng).
5. No stakeholder input exists on this pack yet; nothing reads as signed except where marked Sheran-ruled.