#	The asset	New, or a change to something live	What we need from Design	Every version to write
1	Outbid alert PHASE 1 Its send is on the CRM brief. ▸ Today, live: the creative this reworks Today, live The full record ↗ PUSH reaches them outside the app	CHANGE Rewording a push that already sends.	COPY ONLY Title and body.	New bidder Returning bidder A rejected-bid version, or one line honest for both: the same push fires today when the customer's own bid is rejected for being too low, and 'take back the lead' is wrong for someone who never held it. Design picks the answer; splitting reads better and costs an engineering trigger split ▸ The job, the copy and where it lives Where the copy lives NATIVE Engineering ships the wording in a release. It stays native. The job to be done Turn an outbid into a re-bid before close. Never say “you lost” while the auction is still live. And never assert a competitive event that did not happen: a rejected bid is not an outbid. Copy direction, not final words Your maximum is £X, the current bid is £Y, top up to get back in front.
2	Thirty minutes left alert PHASE 1 Phase 1 copy fix, branded replacement in Phase 2 Its send is on the CRM brief. ▸ Today, live: the creative this reworks Today, live The losing variant; the winning one sits beside it in the record. The full record ↗ PUSH reaches them outside the app	CHANGE Rewording a push that already sends. The branded replacement for it is a new build in Phase 2.	COPY ONLY Title and body.	Winning Not winning ▸ The job, the copy and where it lives Where the copy lives NATIVE→BRAZE Engineering ships the Phase 1 rewording. The branded replacement moves it to Braze in Phase 2. The job to be done Stop a missed-window loss without nagging the people already winning. Copy direction, not final words Your auction closes in 30 minutes and you are not in the lead. Top up to get back in front.
3	Auction ended alert PHASE 2 Waits on the backend release that moves the close push to Braze. ▸ Today, live: the creative this reworks Today, live The full record ↗ PUSH reaches them outside the app	CHANGE A rewrite of the push that already sends at the close, with the sell stripped out.	COPY ONLY Title and body.	One version, it reads the same for everyone ▸ The job, the copy and where it lives Where the copy lives NATIVE Engineering ships the wording in a release. It stays native. The job to be done Bring the customer who was away from the app to the end-of-auction screen. It states what happened and sells nothing. Copy direction, not final words Your auction has ended; here is how it closed.
4	Sorry messages PHASE 1 Its send is on the CRM brief. ▸ Today, live: the creative this reworks Today, live The loss email it reworks; the closed-early variant folds in. The full record ↗ PUSH, THEN EMAIL reaches them outside the app	PART NEW Only the early-close cause has a send today. Our technical fault and operator pulled the seats have nothing at all.	COPY ONLY Push and email copy for three causes.	Our technical fault Operator closed it early Operator pulled the seats ▸ The job, the copy and where it lives Where the copy lives BRAZE CRM ships it, no release. The job to be done When the close was not the customer's doing, say what happened, say where their money stands, and give one way forward. Never “you lost” on a close that never ran, and never “it never completed” on one that did. The three share one template, so they have to work as variants of a single design. Copy direction, not final words Technical: the auction never completed, and here is where your money stands. Early close: this one ended sooner than planned; here is your route back.
6	Day-one how-it-works email PHASE 1 Its send is on the CRM brief. ▸ Today, live: the creative this reworks Today, live The full record ↗ EMAIL reaches them outside the app	REBUILD A live email exists, and the ruling is to build it fresh rather than rework it.	COPY AND DESIGN A complete email, both data states, plus the two-question help survey as a tap module. The education slot never repeats a lesson: a repeat loser gets acknowledgement and changed advice.	1 New bidder register: the full lesson, nothing assumed 2 Returning bidder register: same modules, sharper 3 Lesson lead, near miss: the credit test cell rides here (gated on its read) 4 Lesson lead, priced well out: the maximum-bid lesson 5 Lesson lead, bid late: the timing tip leads (ships once Analytics sizes the segment; reads generic until the bid-timing data exists) 6 Lesson lead, repeat loss: acknowledge the pattern, change the advice (threshold with Analytics) 7 The two-question help survey, as a tap module 8 What happened, knocked out: the upgrades for their party size ran out, so no price comparison is named 9 What happened, left the app: they left before the close, and the gap to the winning bid still teaches ▸ The job, the copy and where it lives Where the copy lives BRAZE CRM ships it, no release. The job to be done Teach the loser how to win next time, and tell them how close they were. Day one owns the mechanic: how bidding works, and what happened to you. Education, never recovery. No offer language in the standard versions; the near-miss credit appears only inside its test cell and is never advertised in advance. Copy direction, not final words What happened, how close they were, and how to win the next one. The timing tip: auctions climb near the close, an early maximum beats a late rush. Education, never recovery: the day-three send owns the way back.
7	Day-three come-back email PHASE 1 Its send is on the CRM brief. New: nothing sends today, so there is no live creative. EMAIL reaches them outside the app	NEW Nothing sends at day three today. Every post-loss send fires at the close.	COPY AND DESIGN A complete email, both data states, plus the fallback layout and a push variant.	New bidder, first loss New bidder, lost before Returning bidder, first loss Returning bidder, lost before A no-figure fallback carrying the next auction alone, which is the likely default A push version, only where email is unconsented A return-leg version where a return is detectable, with a £5-off test cell on that return upgrade ▸ The job, the copy and where it lives Where the copy lives BRAZE CRM ships it, no release. The job to be done Give one good reason to look again, to the people who have not already come back on their own. Day three owns the way forward: when the next chance is, and what a winning number looks like on that route. No urgency, and no comparison against the walk-up fare. The lost-before versions acknowledge the pattern rather than repeating the same commiseration. A typical winning price rides as a tip only where the numbers support it. Copy direction, not final words There is another auction on your route on {{next_auction_day}}, and recent upgrades on it went for {{price_low}} to {{price_high}}.
11	Outbid email PHASE 1 Its send is on the CRM brief. ▸ Today, live: the creative this reworks Today, live The full record ↗ EMAIL	CHANGE A template change on a live email, not a new send.	COPY AND DESIGN The class-named headline and benefit lines, on the live template. The fight-back line comes out.	1 First Class 2 Standard Premium ▸ The job, the copy and where it lives Where the copy lives BRAZE Both outbid variants are live Braze campaigns and ours to edit. The job to be done Tell them they are outbid, name the class, and make the way back one tap. Copy direction, not final words The class-named headline, benefit lines per class, and no price line ever showing a blank.
12	Close email, merged PHASE 1 Its send is on the CRM brief. ▸ Today, live: the creative this reworks Today, live The full record ↗ EMAIL, TWO TEMPLATES	CHANGE Both close templates exist and both stay live; the rule chooses between them.	COPY AND DESIGN Two branch layouts on the existing templates, tips block out; the loss branch gains a See your next auction button; the buy-now branch renders only where the offer is confirmed open and an upgrade confirmed available for the party, and drops to the loss branch otherwise.	1 The buy-now branch: an upgrade is still available and the offer is open 2 The loss branch: nothing left to sell, or nothing we can confirm 3 The repeat-loser version on the loss branch: the reassurance line varies on the prior-loss count, never the same cheer twice; and where the loss history is sell-outs or early closes the copy never implies the bid was the problem, another service leads ▸ The job, the copy and where it lives Where the copy lives BRAZE Both close emails are live Braze campaigns. The job to be done One outcome message per close, never two contradicting each other. Copy direction, not final words One send at the close, two branches: buy-now where an upgrade is confirmed available and the offer still open, the loss branch where it is not. Tips block out of both. No separate buy-now email follows.

#	The asset	New, or a change to something live	What we need from Design	Every version to write
8	Auction-close screen PHASE 2 An app build: the end-of-auction screen is a native surface. The app’s own header: Auction ended. Capture: Closed · Lost. IN THE APP inside the app	CHANGE The card slot already exists: today it carries the buy-now offer, the cross-sell or the practice-auction promo. The change is its copy and branching, not a new surface.	COPY ONLY Copy for the eight closes in two registers, and the won ending in three. The built-in fallback line per ending (eight): the app’s own wording when live data cannot load. Stands alone: no name, no price, no data. On money: a losing bidder is never charged, and the copy may say so in Design’s words. The technical-fault ending stays silent until Product confirms no failure path can take payment.	Eight endings, each in a new-bidder and a returning-bidder version, so sixteen The won ending, the ninth branch of the same card: new-customer reassurance, what-comes-with-it for a returning winner, designed silence for power users (drafts on the App experience page). The post-win journey itself stays the first-win journey's Plus the nine built-in fallback lines, one per ending: the version the app shows when Braze’s words are missing or stale, written to stand alone with no data Plus the never-charged line where an ending speaks about money, in Design’s words: a losing bidder is never charged Two endings change with a render-time availability check, so each needs both states: outbid with the trip ahead hands off to the screen's buy-now only while the seat is still on sale, and renders the next-trip ending where it is not; our technical fault offers the seat where one remains, otherwise one button back to a live auction ▸ The job, the copy and where it lives Where the copy lives NATIVE→BRAZE The card is native and always renders; its words move to the Braze copy layer in Phase 2, which is what makes the versions shippable without app releases. The job to be done Tell everyone who did not get the upgrade, honestly, how the auction closed, and give exactly one thing to do next. Sold out and a failed close are not losses and must not read as one. One job and one action per ending. Nothing celebratory, nothing salesy. The won branch: reassure or add value, never re-sell what they just bought. The buy button stays as it is today (changed 16 Sep 2026). Where seats remain, the card’s words say seats may be limited and that fewer can be upgraded; the checkout confirms. Where none remain, the sold-out treatment shows. The card carries the one recommended action for its ending, and no button at all where the screen’s own button already is that action. That rule decides half the copy. Copy direction, not final words Here is how your auction closed, and the one thing you can do next.
9	Lock-screen bid status PHASE 2 Waits on the lock-screen status app build (iOS Live Activity). No capture held; the October design and drawn states: the App experience page. ▸ Today, live: the creative this reworks Today, live The full record ↗ LOCK SCREEN outside the app, on the device	CHANGE A lock-screen stream already exists and Suze designed its states last October. This extends it rather than starting it.	COPY ONLY State copy, once the audit settles how many states there are.	Audit against the existing October design before anything is scoped as new: it covers seven states, this pack asked for four ▸ The job, the copy and where it lives Where the copy lives NATIVE An app build. Braze governs entry, holdout and caps, but the surface is ours. The job to be done Carry bid status passively, so status pushes can retire without bid volume falling. It states, it never sells. And the starting state must stand alone: the card can freeze at its first state if the update token never uploads, so the first frame has to be true and useful with no updates ever arriving. Copy direction, not final words Your live bid status, at a glance.
10	Knocked-out message: party bigger than the seats left PHASE 2 Waits on the knocked-out rework: a backend release that carries party size. No capture held; drawn representation on the App experience page. New: no screen exists today. IN THE APP (A BRAZE IN-APP MESSAGE) inside the app	NEW The pushes that name this exist, but there is no screen that guides the way out.	COPY ONLY Copy for the cause and the two ways forward.	Two ways forward, side by side Knocked out at open reads as a register variant of the same message ▸ The job, the copy and where it lives Where the copy lives NATIVE→BRAZE The push is native today and moves to Braze. The screen itself is a Braze in-app message. The job to be done Name the cause plainly and guide the trade-off, so a resolution action replaces a dead end. Never blame the customer, and never say increase your bid: a party-size knock-out cannot be fixed by bidding. Start from Suze’s existing Avanti West Coast screen rather than from scratch. Copy direction, not final words Fewer upgrades are left than your party size, and here are your two ways forward. Where both classes run on the service, one suggestion line may note the party can sit across classes: a scenario, never a button.

If it is not in these two tables, it is not in this round.

Versions fork on how new the bidder is, on what happened, or on whether they were winning. Travel personas decide who gets sent what, never what the words say. Every asset ships in two data states, personalised and a designed fallback, never a broken merge field, and any figure that cannot be substantiated drops out rather than softening. Where a row changes something live, today's version sits greyed beneath the proposal on the Live experience walkthrough, in the customer's own sequence.