← Q3 journey review packs

Journey 1 council assessment: First-time winner, 1st to 2nd win

Verdict: CONDITIONAL

The strategy is sound and the review found no unsupported claims: every figure carries its label, the consolidation case is verified against the signed matrix rows and TO-BE walks, the ruled contracts (STACK ban, capture ladder, referral strip, travel-completion floor) are correctly encoded, and the one-headline-test discipline is held. The pack could not stand as authored because three cross-journey facts had moved since it was written: Journey 6 is now authored (the pack's interim rule rested on it being unauthored), Journey 6 recommends the opposite M39_E disposition (the pack presented its own verdict as uncontested), and the Journey 5 pack minted two win-surface contracts (dual-listing, one-celebration-per-win) that this pack must mirror or the two packs contradict. All three are now fixed in the pack.

Strongest points: the hazard-curve evidence (81.2% of instant-win quick returns land by D+1, confirmed) driving the D+1 default with a channel-separation rule rather than a blanket block; the honest reach ladder whose bottom row admits the not-armed majority has no habit channel until the placement ships; suppression-as-treatment for the power band; and the entry-population correction to the confirmed 108,821 mix.

What the council changed in the pack (applied, rebuilt, validated)

1. Journey 6 premise corrected; DL-089 boundary adopted. What: replaced "Journey 6 unauthored, interim exit no-send" with the DL-089 two-part boundary as the working assumption (J1 owns forward asks for its 30 canvas days via grw01_canvas_active; J6 owns trip-anchored reflection/review from trip 1), no-send retained only as the fallback if ratification slips. Why: the premise was stale the day it was written; J6's pack landed 20 Jul (CROSS-JOURNEY-REVIEW item 2). Customers: every D+30 non-converter and every first-trip traveller in the canvas window. Where: TO-BE map rules, D+30 map node, strategy §3d/§5. Requires: the grw01_canvas_active journey-state key (D7, already pinned). Improves: no first-winner gets J1 D+3 and J6 T+24h forward asks in the same week. Action: Sheran ratifies DL-089.
2. The M39_E contested verdict is now on the pack's face. What: the brief and map now state that Journey 6 recommends the opposite shape (rework in place as the post-trip T+24h slot) and that one Sheran ruling covers both packs (DL-002). Why: the sharpest inter-pack disagreement in the portfolio; presenting IN ADDITION as settled would have misled the review group. Customers: all winners in the D+1 window. Where: existing-asset table, map open questions, strategy §5. Requires: nothing new. Improves: honest decision surface at Gate-1. Action: Sheran rules DL-002; Lisa + Analytics confirm trigger coverage first.
3. The Journey 5 win-surface contracts written in (DL-079, DL-081). What: B3 dual-listing (ACT-19 Rows 5/7 run as the in-app surface of B3 under the shared win_moment_canvas_active key, one owner, no second build) and one celebration per win event (Beginner's Luck as a copy branch of the owning surface, never a second modal). Why: without the mirror, a never-won close-window winner takes two celebrations and two asks at one event, and 31.8k/yr transacted-never-bid users sit in two journeys' practice arcs (both confirmed cohorts). Customers: never-won winners and the 31.8k onramp. Where: matrix rows 1 and 3, TO-BE rules, strategy §5. Requires: win_moment_canvas_active key; Q5 audit. Improves: enforces the STACK ban and single-ask contract at the highest-trust moment. Action: Sheran ratifies both; DL-081's mechanic strike is flagged explicitly.
4. Content-card instrumentation paired launch-blocking with the placement (DL-013). What: every appearance of the placement gate now carries its pair: card tracking registers ZERO today (confirmed), Eng fixes, Analytics verifies warehouse landing before the holdout starts; added as row 6 of the Analytics build table and to the Eng actions. Why: a placement that ships unmeasured burns the majority cohort's test; the read is the point. Customers: the not-armed ~56% of IW first-winners. Where: brief status row, channel rules, Analytics brief, Process tab, strategy §4/§6. Requires: the instrumentation fix (Eng, briefed). Improves: the not-armed arc becomes readable, protecting the journey's primary KPI claim. Action: Sheran sequences the paired build (DL-036 + DL-013).
5. Pre-travel slot ownership rule added (DL-052). What: while a won trip is pre-travel (S1), Journey 2 owns every marketing slot (layered rule 6 L1); the T+0 win beat is exempt as the entry hand-off; rebook touches floor at travel-completion + 1 day. Why: for the ~85% short-lead majority the post-win and pre-travel windows are the same days; the pack held the floor but not the ownership half. Customers: every winner travelling within the canvas window. Where: TO-BE map rules, strategy §5. Requires: rule 6 layer model (ratified) + journey_state. Improves: cap arbitration by design, not by accident. Action: Sheran ratifies DL-052.
6. Fourteen-field completeness index added. The brief tab now carries the standard's fourteen personalisation and contact fields with where each is answered; all fourteen were present in substance, field 12 (adaptation to response) was the least explicit and is now stated.
7. Plain-English pass. BUN, IW, L0, PECR and MDE expanded at first use per tab; no other writing-standard breaches found (sections lead with conclusions; no em-dashes; figures labelled).

Working assumptions adopted (closed rather than escalated)

• DL-039 closed: superseded; the DL-089 boundary is the working assumption, no-send only as fallback.
• DL-040 closed: travel-completion proxy = scheduled departure timestamp + 1 day, swap when trip_completed lands; Eng's timestamp-reliability confirm moves to the validation plan. The ruled floor stands; the proxy is the only buildable reading.
• DL-043 closed: T21_P/T14_P exclusivity is a Braze audit task (tracked on the Process tab, Lisa), not a decision; the shared dedupe key and STACK ban make the design safe under either answer.

What genuinely requires approval

Sheran: DL-005 (paired J1+J2 ratification, the pack's headline ask) · DL-002 + DL-089 (one post-trip ruling across J1/J6) · DL-081 + DL-079 (win-surface contracts; DL-081 strikes a card mechanic) · DL-035 (C1 re-prompt delete) · DL-037 (D+1 default) · DL-038 (ACT-13 fork re-order) · DL-041 (headline test) · DL-052 · DL-085 (ladder-wide counter) · backbone dials. Build hard gates: DL-001/Q5 (Eng/PMM audit) and DL-034 (T31 PECR class ruling, Lisa + compliance). Investment: DL-036 + DL-013 (placement + instrumentation); DL-044 stays open (celebration screen commit/defer, Product + Sheran).

Evidence gaps (feeding the validation plan)

1. The canonical 2nd-transaction KPI (DL-042), the journey's top gap: two definitions circulate (portfolio 5% to 7.5% vs M1 repeat-bid 7.03% to 8.8%); the headline test cannot be sized or read until Analytics (Ivi) defines one, together with the 2-8d vs 6wk median reconciliation.
2. M39_E trigger coverage (auction wins vs BUN only), TBC: Lisa + Analytics; decides whether the D+1 beat exists for close-window winners and shapes DL-002.
3. M29 opt-in baseline volume and the early-winner first-timer split (ask ㉘), both still missing (Analytics).
4. Content-card baseline: none exists and none can until the instrumentation fix lands (DL-013).
5. Consultations not yet held, stated per the standard: Product on the interim savings-modal anchor for BUN celebration-lite (surface_status LOW CONFIDENCE); CS on the refunded-win hand-off script.
6. Departure-timestamp reliability at canvas evaluation time (Eng), underwriting the DL-040 proxy.