The problem Trying reads as risky, so the first try never happens.
Why it happens Trying reads as risky and the whole offer reads as too good to be true. Comprehension and trust, not value rejection: the default posture of a prospect is suspicion.
The behaviour we are driving They decide buy vs bid on the screen where the choice is made, grant marketing permission at a moment of proven value, and complete the first transaction.
The job to be done “Help me decide whether to buy or bid, and make my first try feel safe.”
Q4 marketing model, Business Case tab, exported 29 Sep 2026: volumes checked by Analytics 8 Sep 2026; organic and target rates measured against holdout and confirmed by Analytics; a buyer's month valued at £30 an upgrade, 1.29 upgrades in a first-purchase month and 1.67 after; net is gross at the 23% take rate.
| Model row | Worked volume a month | Rate | Value per active month | Gross a year | Net a year | Q4, driven | Q4, in quarter |
|---|---|---|---|---|---|---|---|
| Cooling never-buyers; the bid or no-bid signal still routes treatment ACTIVATE | 6,831 The number is the monthly inflow into the 61 to 90 day band: the band holds 20,493 people and refills about 4 times a year, so 20,493 x 4 / 12 = 6,831 a month. Ivi replaces this with the directly measured monthly inflow when it lands. | 3.12% vs 2.4% organic | £30 an active month (one upgrade) × 1 months × 12 a year | £14,165 | £3,258 | £544 | £543 |
| Cold never-buyers; bid or no-bid still the routing signal ACTIVATE | 6,168 The number is the monthly inflow into the 91 to 120 day band: the band holds 18,979 people and refills about 3.9 times a year, so 18,979 x 3.9 / 12 = 6,168 a month. Ivi replaces this with the directly measured monthly inflow when it lands. | 2.6% vs 2% organic | £30 an active month (one upgrade) × 1 months × 12 a year | £10,659 | £2,451 | £410 | £409 |
Q4 2026: £954 net on the driven basis (Oct £0, Nov £469, Dec £485); £952 landing inside the quarter. Driven: months active credited to Q4, pro rata by days live from the launch, the basis the Q4 V2MOM carries. In quarter: only the months that fall in October to December (Tony, 10 Sep 2026: the forecast Marketing submits). Net is gross at the 23% take rate. Launch 1 November 2026.
Still to confirm Rates measured against holdout 8 Sep, rough until the holdouts mature (item 6). Two segments, below
Ladders to Drive the First Win: new user to first purchase 20.4%; existing user to first purchase 4.8%. Metric moved New-to-first-transaction %. Evidence B, benchmark.
The planned case, a 30% lift in conversion, as presented to the board on 23 Sep 2026 (the Sizing tab behind the model's Business Case tab): volumes checked by Analytics (Ivi, 8 Sep 2026); organic and target rates measured against holdout and confirmed by Analytics; a buyer's month valued at £30 an upgrade, 1.29 upgrades in a first-purchase month and 1.67 in later months (the 20 Sep change; the earlier £172k counted one upgrade a month); net is gross at the 23% take rate (agreed 8 Sep 2026). Source: Q4 2026 Marketing Model, Business Case (output) and Sizing tabs, exported 2026-09-29 (the live model). Every figure is incremental of the customers who would have returned anyway: worked volume × (target rate − organic rate) × months active × times a year × value per active month.
Each engineering-row chip is coloured by the ask: APP AND JOURNEY OWNED COMPUTED TRAIT · ANALYTICS EXISTS · PIPE TO BRAZE NEW EVENT · ENGINEERING BUILDS
| The moment | Phase 1, ships now | Phase 2, waits on engineering | Who builds it |
|---|---|---|---|
on the map: 1 · The welcome, or the return The welcome, or the return | email One welcome in two versions. Engineering rows D9 | Nothing waits | Braze · Braze buildDesign · copy and creative |
on the map: 5 · The day-one email: how it works, then try it The day-one email | email The day-one how-to email keeps as the one email of the first 48 hours and goes to every new account, bid or no bid. Engineering rows D9 | Nothing waits | Braze · Braze buildDesign · copy and creative |
on the map: 3 · The buy-or-bid decision, in place The buy-or-bid decision, in place | Nothing ships in Phase 1 | native screen or modal Certainty-led copy lands where buy vs bid is decided, for every route (Sheran, 6 Sep: the decision moment never shows the practice screen): the buy-now box… | Analytics · computed traitDesign · copy only |
| The return after a first loss | Nothing ships in Phase 1 | native screen or modal, braze in-app message The app stops greeting a returning loser like a stranger: the auction screen reframes a first loss as bad luck, not proof it does not work,… | App engineering · computed traitBraze · Braze buildAnalytics |
on the map: 4 · The permission ladder The permission ladder | Nothing ships in Phase 1 | native screen or modal, braze in-app message Three uncoordinated asks become one ladder, rungs strongest first: a watch placed (alerts are what make a watch work), a first search with live auctions on… | App engineering · computed trait, new eventBraze · Braze buildDesign · copy and creative |
| The seasonal wave, capped | email, braze push A tiered, capped, first-win-framed seasonal wave replaces the disabled nudge blanket: one honest reason to look, at a moment a trip is plausible, per engagement tier,… | Nothing waits | Braze · Braze buildAnalytics · computed trait, new model |
on the map: 6 · The stall, and who owns it The fresh-start offer | email A loss is never chased twice, so the one follow-up is money off the first upgrade, framed as a fresh start, never as compensation. Engineering rows D3 | Nothing waits | Braze · Braze buildAnalytics · computed trait, decision neededDesign · copy and creative |
on the map: 8 · The first-win celebration The first-win celebration | Nothing ships in Phase 1 | native screen or modal, braze in-app message The first win stops passing through the generic confirmation every winner sees: the celebration lands the proof (what they paid against the walk-up price, substantiated at… | Braze · decision neededAnalytics · computed traitDesign · copy and creative |
on the map: 6 · The stall, and who owns it The check-in, day 7 | email A silent first week earns one value-led check-in instead of a calendar drip: the £5 against a real price for credit holders, the practice invitation for… Engineering rows D9 | Nothing waits | Braze · Braze buildAnalytics · logic changeDesign · copy and creative |
| The occasion touch inside the window | email The middle of the window is owned by relevance: the wave where one lands, the occasion fallback where none does. | Nothing waits | Braze · logic change |
on the map: 7 · The expiry warning The expiry warning, day 57 | email, braze push One deadline, honestly enforced: the strongest-converting send class fires where urgency is true, then the incentive story ends. | Nothing waits | Braze · logic changeDesign · copy and creative |
| The stalled searcher | This journey registers the abandon feed and stays silent for the window. | Nothing waits | Marketing · copy only |
on the map: 1 · The welcome, or the return The hand-back after a first loss | Nothing marks the hand-back, by design: the customer already heard the recovery arc's day-one and day-three messages. | Nothing waits | Marketing · copy only |
| The genuinely unengaged | No wave touch: sending to the dead half of the pool costs sender reputation for every live send the estate makes. | Nothing waits | Marketing · copy only |
The words are on the journey map above. A D number opens that row of the Engineering brief; the engineering board uses the same numbers. Live sends not named here continue unchanged.
| The question | Owner |
|---|---|
| Rule which surface owns each win type's celebration (auction won, instant win, auto-win, buy-now), with Product co-approving the in-app halves. | Blocks launch Sheran, at Gate 1 |
| Commission the permission ladder's master record (the asks, surfaces, copy and shared-memory rules in one signed design), before any ladder surface ships. | Blocks launch Sheran, at Gate 1 |
| Rule the fresh-start offer, pending cost sign-off and Analytics sizing. | Sheran, at Gate 1 |
| Declare the long-running education-message test's winner, or fold its control into the onramp holdout. | Blocks launch Analytics: Ivi and Ben |
| One canonical definition of the marketable never-transacted pool: three figures circulate (820k, 850k, 857.8k), and the engagement-tier model behind the wave. | Blocks launch Analytics: Ivi and Ben |
| Confirm where every button lands: sign the register 'Where every tap lands' on this matrix against the deep-link availability list, one pass for the journey (Sheran, 6 Sep 2026, ruling J5-12). | Product |
| # | The ask | Hand back | Start here |
|---|---|---|---|
| 1 | Decide the education-message test: declare the winner or fold its control into the onramp holdout. The standing rule: running tests get decided. | The verdict, and what the winning message is. | The session-slot row on the Decisioning matrix. |
| 2 | One canonical definition of the marketable never-transacted pool: three figures circulate (820k, 850k, 857.8k). Due before the wave calendar freezes. | The definition and the number. | The wave row on the Decisioning matrix. |
| 3 | The engagement-tier model behind the wave, with contactability per channel, and the deliverability line for the genuinely unengaged. | The tier definitions and per-tier volumes. | The wave row and the suppressed rows. |
| 4 | Size the onramp holdout: baseline, minimum detectable effect, significance and runtime, read on first real bid conversion among never-bid customers. Conversion, never sessions. | The sized test design. | The measurement position on the Campaign brief. |
| 5 | Build the two selection traits: bid state and transacted type at render, and the never-won flag at the win, nightly on the standing pattern. | The traits live in the sender. | The Engineering brief’s traits band. |
| 6 | Size the fresh-start offer test, and the £5 consent-incentive lead if ruled in: both money questions carry a sizing before any send. | The test designs. | Rows 5 and 7 on the Decisioning matrix. |
| 7 | Ratify the five outcome audiences the film and walk now carry (signed up in the last 30 days; searched, never bid; bid in one auction, never won; bid in two or more, never won; not seen for 112 days or more), settle their overlap and priority order, and join the push and email consent split to each. Counts pulled from the user segmentation model 6 Sep 2026, one basis and one date. | The ratified definitions, and the per-outcome channel split. | The outcome chips on the film’s title card and the walk’s outcome picker. |
The do-nothing baseline: the 30-day new-user purchase rate, 16.8%, down from 20.2% in under a year. The journey is judged on moving it toward 23%.