← Q3 journey review packs

Journey 5 council assessment · Never-transacted: first win

Verdict

APPROVE AT GATE-1, WITH CONDITIONS. The journey's design chain (evidence to choice to experience) holds at every touch, the thesis weighting is honest (first win = unlock; value claim = activation volume feeding Journey 1, not revenue; the 857.8k pool is never multiplied), and the Legal/Risk posture on Beginner's Luck is the strongest in the portfolio. The conditions are the five approval items below plus the cross-journey ratifications already in the consolidated log (DL-004, DL-006, DL-065, DL-073).

Strongest points (keep, do not dilute)

1. Pool honesty: the opportunity ladder (in-session deciders, 31.8k onramp, Cooling 20.9k test tier, churned tier as cheap reach only) resists the 850k-times-anything temptation. Founder/Growth lens: this is the correct weighting of the thesis lever.
2. The substantiation rule (population stat only, never a personal prediction, "guaranteed" only on the verified guarantee path, anchors substantiated at fire-time) is applied consistently across every asset, mock and matrix row.
3. State machine over sequence: entry is a state, in-app surfaces are user-initiated and outside arbitration, and the win event (not journey completion) is the exit. The AS-IS audit is honest, including the attributed-not-incremental read on M63.
4. Boundary discipline: D-01 (exit on first transaction; 31.8k dual-listed with J1 B3) and D-05 (one celebration per win event) fix the two worst collision risks before they are built.

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

1. J3 episode window adopted (DL-065). What: J5's outbound layer now states suppression during a Journey 3 recovery episode (7-day window, automatic hand-back). Why: most J3 abandoners ARE J5's audience; unstated, one user gets funnel recovery and activation sends in the same week. Who: never-transacted funnel abandoners. Where: TO-BE map rules, matrix T6 suppression, strategy §3/§5. Capability: journey_state. Outcome: no double-owner contact. Action: DL-065 ratification stands with Sheran (cross-journey).
2. J4 loss hand-back adopted (DL-073, DL-004). What: recovery ownership restated as split at the live bid (J3 pre-bid, J4 post-close), and never-transacted losers exiting J4's day-14 close re-enter J5 with first-timer flags (first_loss, close_reason, route). Why: three packs concur on the split; the pack previously named J3 as sole owner and had no re-entry path for its biggest at-risk cohort. Who: first-time losers. Where: exec summary entry row, matrix live-auction row, map rules, strategy §3/§5. Capability: exhaust flags from J4's canvas. Outcome: the loser gets the bad-luck-not-rejection reframe instead of a cold restart. Action: DL-004/DL-073 ratification (cross-journey).
3. Content-card instrumentation gate surfaced (DL-013). What: the pack now states that card impressions/clicks register zero today (confirmed) and makes the fix launch-blocking for the T3/T4 reads (Analytics brief build row 5; matrix T4; strategy §2/§4/§6). Why: an unreadable test is worse than no test; the sizing ask for "home-card CTR baselines" was impossible as written. Who: all card-surface audiences. Capability: Braze card event instrumentation to warehouse. Outcome: readable tests, honest baselines. Action: DL-013 journeys field extended to J5 (delta).
4. Fourteen-field personalisation completeness table added. What: a one-page index on the Campaign brief tab mapping each of the standard's fourteen personalisation and contact fields to the one place it is defined. Why: the standard requires all fourteen explicitly; contact priority and required personalisation level were previously implicit. Who: reviewers. Where: brief tab, after the decisioning matrix. Outcome: checkable completeness at Gate-1. Action: none.
5. PECR line added to the capture rules. What: marketing-layer capture asks now inherit the PECR classification precedent from Journey 1's T31 ruling (DL-034; unbundling review DL-038), with compliance reviewing capture copy class before build. Why: the cross-journey review names J5 in the PECR cluster; the pack asked for the right consent layer but not the right classification. Who: all capture-surface audiences. Outcome: regulatory exposure managed before build, not after. Action: rides DL-034/DL-038 (J1 group).
6. 23% metric conflict resolved as a working assumption (closes DL-087). What: 23% treated as the TARGET (the portfolio row); interim baseline = the DN 30-day new-user purchase rate, 16.8% (confirmed as stated); the canonical pair still publishes via DL-009. Why: the two sources cannot both be right, and DN's 16.8-20.2% read makes "23% confirmed baseline" implausible as the same cut; sizing was blocked on the conflict. Where: exec summary key metric row, strategy §1/§4. Outcome: test sizing unblocked without inventing a figure. Action: Analytics publishes the canonical pair (validation, not approval).
7. Wave email subject A brought inside the pack's own format rule. What: "Try First Class this summer. You only pay if you win." (53 characters) shortened to "Try First Class. You only pay if you win." (41) in the design brief, reference mock and walkthrough. Why: the pack mandates subject ≤45 characters; the flagship asset broke it. Outcome: brief and experience match. Action: none (copy stays agent-drafted DRAFT).

Working assumptions adopted (labelled, with break conditions)

• 23% = target; 16.8% DN read = interim baseline. Breaks if Analytics' canonical pair differs (DL-009).
• The D1 bid-state/transacted-type sync is treated as required and briefed launch-blocking on the ER board; T2/T3 stay blocked-not-buildable until Eng confirms render-time targeting (closes DL-083 as validation, per ruling 11).
• M63 operationalise-first staging stands; nothing switches off until a winner is declared or the control folds into the onramp holdout; the eventual REPLACE clears at the normal launch gate (closes DL-084).
• ACT-12 split by audience owner: J5 signs only never-transacted variants; returning personalisation waits for Journey 10's scope (closes DL-086; reversible, pre-empts nothing).
• DL-004/DL-065/DL-073 recovery contracts are adopted in the pack ahead of ratification (three packs concur on DL-004; the window and hand-back reuse ruled shapes).

What genuinely requires approval (open, in the consolidated log)

• DL-079 audience boundary + 31.8k dual-listing with J1 B3 (Sheran, with the J1 author; J1 pack needs the mirror edit).
• DL-080 ACT-15 master row + Gate-1 card (Sheran + Airtable write; the pack cannot complete Gate-1 with a code-less moment).
• DL-081 one celebration per win event: a mechanic-ownership change vs the Gate-1 card's dedicated-modal intent, flagged for Sheran to strike.
• DL-082 substantiation rule ratified + a named compliance owner for win-rate, price-comparison and consent-incentive claims (none exists today; ASA/CAP exposure).
• DL-085 ladder-wide extension of the ruled ACT-13 capture cooldown (extends the scope of a Sheran ruling).

Evidence gaps (REQUIRES VALIDATION, owner named)

1. The journey's own denominator and metric are both unreconciled: one Analytics table (DL-009) covering the marketable-pool definition (820k / 850k / 857.8k drift) and the 23% baseline/target pair. Owner: Analytics (Ivi). This is the top gap: the Gate-1 verdict is a bet on moving a number that is not yet real.
2. ~70% guarantee-fires rule + eligibility flags (gates the Beginner's-Luck branch build). Owner: Analytics (Ivi/Ben).
3. Live BUN box copy + real anchor values (Gate-2 blocker). Owner: Product/Pricing.
4. M63 winner declaration or control fold-in (gates REPLACE staging). Owner: Analytics.
5. Content-card instrumentation landing in the warehouse (DL-013). Owner: Eng, Analytics verifies.
6. Canonical barrier-mix coding (IN 35/27/20/15/7.5 vs BG 33/25/7). Owner: Research.