What this document is. The launch-day operational script for Tuesday 2 December 2026 — Giving Tuesday, the peak day of the Christmas Appeal. This is not a strategy document (that is /christmas-appeal-2026). This is what to do minute-by-minute so the campaign launches cleanly, is monitored honestly, and responds to reality without panic.
Ownership. AI CMO (this doc + real-time direction) · AI paid ads ops (Meta + Google unpause and monitor) · AI copywrite (email schedule confirmation) · AI organic socials manager (Lisa daily calendar) · Web-dev (attribution pipeline monitoring) · Carla (single point of escalation if any red-line trip fires).
Governance. 5-gate Marketing Governance Envelope. Reads Sprint 48 attribution (must be live). Reads /christmas-appeal-2026 for the campaign strategy. Reads /christmas-appeal-email-sequence for Email 2 send confirmation.
Contents
The one thing that matters on 2 December. The campaign either goes live cleanly at 6am, or it doesn’t. Everything before 6am is preparation. Everything after 6am is response. This playbook is what “go live cleanly” and “response” look like.
1. Day-before checklist — Mon 1 Dec 2026 EOD
Every item on this list must be green before 6pm AEDT on 1 December. If any item is amber, Carla is informed by 5pm and the launch decision is escalated. Red on any item stops launch.
- Landing page live.
/christmas-appealreturns HTTP 200 with the “coming soon” overlay visible. Underlying page fully authored and reviewed. Server-rendered compliance envelope present. Alt-text on every image. WCAG 2.2 AA verified via Lighthouse. - Attribution pipeline verified. Sprint 48
lwd_sessioncookie firing. D1marketing_sessionstable writing. Meta CAPI test event received in Events Manager (test event code active). Google Enhanced Conversions test event received. Server-sidedonate_successfiring end-to-end in staging. - Email 2 (launch) staged in ESP. Scheduled for 06:00 AEDT 2 Dec. Test send received by Carla + AI CMO. Merge fields verified. Compliance envelope present. Composite label visible. Unsubscribe URL resolves.
- Meta campaign paused-and-ready. All ad sets in Paused status. Creatives approved by Meta (Advantage+ Shopping or standard, depending on ad type). Daily budgets set. Audiences finalised. UTM stack pre-populated in Facebook Ads Manager URL parameters.
- Google campaign paused-and-ready. All 3 ad groups in Paused status. RSAs approved (green disapproval-check passed). Bid strategy set to Maximize Conversions with target CPA capped. Conversion actions imported from Google Analytics via Enhanced Conversions.
- Organic calendar published to Lisa. First 3 posts staged in Buffer / Later / native platform schedulers. Lisa has confirmed by 4pm 1 Dec via WhatsApp / SMS.
- PR pack sent. Press release + spokesperson brief already emailed to the 12-producer target list by 25 Nov. First follow-up 27 Nov. Second follow-up went out 1 Dec 9am. Any bites received — forward to Carla immediately.
- Auto-stewardship configured. Sprint 48 4-hour SLA thank-you template active in ESP transactional path. Test
donate_successevent fires a test stewardship email. Template carries the Christmas Appeal addendum (Sprint 49 §9). - Escalation contacts confirmed. Carla on-call from 5am 2 Dec. AI CMO on-call all day. Web-dev on standby from 5am for any /donate pipeline issues. Escalation ladder (§7) shared.
- Sunset overlay staged. A pre-authored thank-you variant of
/christmas-appealis staged and ready to swap in at 6pm 24 Dec. Deploy path documented. - Carla sign-off received. Written go/no-go by 6pm 1 Dec. Any amber items require explicit acknowledgement.
2. The launch-day timeline — nine checkpoints
- Landing page returns 200 (from within Australia AND from outside via a proxy).
- /donate returns 200. Weekly-giving-calculator returns 200. /give returns 200.
- Attribution: fire a test session (private-browsing tab) — confirm cookie writes, session lands in D1.
- ESP: Email 2 still queued for 06:00. No manual pause introduced overnight.
- Report status in launch channel by 05:30. Green / amber / red.
- Deploy the production build of
/christmas-appealwithout the coming-soon overlay. Verify at 05:57 that the primary CTA button renders and links to the correct/donate?plan=weekly&amount=19.25&...URL with the correct UTM stack. - Fire a real (non-charging) test click through the CTA — confirm the /donate page loads with the correct amount pre-selected.
- If any of these fails, ROLLBACK to coming-soon overlay by 05:58 and escalate.
- 06:00 sharp: Email 2 sends automatically per ESP schedule. AI CMO confirms send in the launch channel within 3 minutes.
- 06:01: Meta campaign unpaused. All ad sets active. Daily budget confirmed at approved level. Take a screenshot for the retro.
- 06:02: Google campaign unpaused. All ad groups active. Bid strategy confirmed. Take a screenshot for the retro.
- 06:03: Lisa posts organic Post 2 on LinkedIn + Instagram + Facebook (LinkedIn first, then Instagram, then Facebook — not simultaneous, staggered by ~5 minutes for platform algorithm friendliness).
- 06:10: Ads-ops confirms in Meta Events Manager that the first Purchase event has fired (a test purchase or genuine early conversion). Same for Google Conversions.
- If any of these fails or is delayed > 15 minutes, escalate per §7.
- Read the three numbers from §4 (page views · donate intents · donate successes).
- Compare against day-1 baseline expectations (published in the campaign brief on 1 Dec EOD).
- If the numbers are broadly in range, do nothing — the campaign is running.
- If page-views are strong but donate-intents are weak, the landing page is the bottleneck — queue for review by 09:00.
- If page-views are weak, the ads or the email are the bottleneck — queue for review by 09:00.
- Written status posted to the launch channel by AI CMO. 5 lines maximum: three metrics, one qualitative observation, one action.
- Carla reads and acknowledges. If any action is proposed, Carla approves or defers.
- PR pack: any morning-news bites? Report status.
- Organic: how did Lisa’s launch posts perform in first 3 hours? Comment/reaction volume.
- Meta: what is the current CPA per ad-set? If any ad-set CPA > 3× the campaign average and > 100 impressions delivered, pause that ad-set (creative fatigue guard).
- Google: same principle for ad groups. Bottom-quartile RSA headlines by CTR are automatically deprioritised by Google’s optimisation — do not manually pause headlines before 24h data.
- Any creative rejected by Meta or Google since 06:00? If yes, fix per Meta/Google policy and resubmit. Do not remove RG-96 language to make a rejection pass — if the language is causing rejection, escalate to AI compliance for a variant.
- Lisa: post 2 of the day scheduled? Any DM inbound that needs a human reply (not from a system account)?
- Stewardship: how many
donate_successevents fired since 06:00? How many stewardship emails delivered? Sprint 48 4-hour SLA holding? - If SLA breach observed (any thank-you > 4h after gift), escalate to web-dev immediately — the transactional pipeline needs investigation.
- Full day’s three-number snapshot posted to launch channel. Cumulative page views · donate intents · donate successes.
- Ad spend to date. Cumulative CPA. Compare against day-1 target.
- Email 2 metrics: opens, clicks, conversions attributed via Sprint 48 pipeline.
- Any red-line trip today? If yes, what was the response? If no, note it.
- Overnight monitoring: attribution alerts, ESP bounce alerts, Meta rejection alerts — all routed to AI CMO for 22:00 sweep.
- Final metric snapshot for the day. Written — not just logged.
- Any alerts from Meta, Google, ESP, D1, or CDN? Triaged, actioned, or scheduled.
- Tomorrow (Wed 3 Dec): Lisa’s post 3 queued? Ad spend continues at planned level? Any adjustments approved by Carla for Day 2?
- Shut down. This is a 26-day campaign, not a one-day sprint. Sleep is a campaign asset.
3. Red-line triggers — what stops the campaign
Five conditions trigger an immediate pause of paid spend (Meta + Google) pending investigation. Not review-later — stop now.
Red-line 1 — compliance breach. A live ad, email, or landing surface is caught carrying prohibited language (“debt relief,” “we eliminate debt,” “guaranteed outcome,” “clients”) or missing the ACL 387398 disclosure. Response: pause the offending asset within 15 minutes; escalate to AI compliance for replacement; if replacement not ready in 4 hours, hold the paid channel dark until it is.
Red-line 2 — attribution failure. The donate_success server-side event stops firing, OR the lwd_session cookie stops writing, OR D1 marketing_sessions stops receiving inserts. Response: web-dev investigates immediately. Do not pause paid spend automatically for this trip — conversions are still happening — but do log the outage window in the retro. If the outage lasts > 4 hours, pause Meta + Google to preserve budget for when attribution is restored.
Red-line 3 — unsubscribe spike. Any single email send’s unsubscribe rate > 0.5% within 24 hours of send. Response: stop the next scheduled email (Email 3, if the trip fires on Email 2). Review before proceeding. This is the fatigue guard; the campaign brief accepts a smaller total number of gifts over damage to the house list.
Red-line 4 — spam-complaint spike. Any single email send’s spam-complaint rate > 0.1%. Response: stop the next scheduled email. Investigate. This is the deliverability guard — a spam-marked LWD email risks the inbox reputation for the whole 2027 marketing calendar.
Red-line 5 — SLA breach on stewardship. Any thank-you email sent > 4 hours after its triggering donate_success event, more than 3 instances in one day. Response: escalate to web-dev. The 4-hour promise is a Sprint 48 non-negotiable. This does not pause paid spend — but it does pause the next email send until the pipeline is verified.
4. First-hour metrics — the three numbers
From 06:00 to 07:00 on 2 December, three numbers matter more than everything else. Everything else can wait 24 hours for review.
1. Page views on /christmas-appeal
baseline TBD
Measured via Sprint 48 page_view event. Baseline expectation published to launch channel on 1 Dec EOD. Weak page-views in first hour = ad-delivery lag or email deliverability issue.
2. Donate-intent clicks
baseline TBD
Measured via Sprint 48 donate_intent event (any click on primary CTA). Baseline expectation: roughly 6-10% of page-views. Weak intent-rate = landing page not converting; queue for review.
3. Donate successes
baseline TBD
Measured via Sprint 48 donate_success server-side event (canonical). Baseline expectation: roughly 15-25% of donate-intents. Weak success rate = /donate flow issue; escalate to web-dev.
Baseline TBD deliberately. The specific numeric baselines are set on 1 Dec EOD based on final list size, ad budget, and any peer-benchmark data available. This document does not pretend to know them today. What matters is: the three numbers are watched, the ratios between them are watched, and any hour that is < 30% of expected on any of the three triggers a review.
5. Creative rotation and CPA guardrails
The 24-hour CPA rule
For the first 24 hours of the campaign (06:00 Tue 2 Dec → 06:00 Wed 3 Dec), do not pause creatives on CPA grounds alone. Let Meta’s and Google’s learning phases run. Only pause on the red-line triggers in §3.
The 48-hour CPA rule
From 06:00 Wed 3 Dec, any ad-set (Meta) or ad-group (Google) whose CPA is > 2× the campaign-wide average, over > 200 impressions, is paused. Not deleted — paused — so it can be resurrected if a creative rotation later re-opens the audience.
The mid-campaign creative rotation
Around Day 10-12 (11-13 Dec), fatigue on the initial creatives typically emerges. Sprint 49 does not build a second batch of creatives — instead, the initial set is rotated: video-30s becomes the primary; carousel becomes secondary; still-typographic becomes tertiary. Video-15s is paused (its role was hook in first week; it has been seen). This rotation is executed by paid ads ops without further approval, per this playbook.
The bid-strategy discipline
- Meta. Highest Volume bid strategy for first 3 days (learning). Switch to Cost Cap at day-1-CPA baseline from Day 4. No manual bid tweaking below the ad-set level.
- Google. Maximize Conversions with target CPA cap from day 1. No manual bidding. No keyword-level bid adjustments below the ad-group level.
6. Lisa’s organic day plan for 2 December
Lisa runs organic in her own voice. The AI CMO does not draft her posts. Lisa’s day 1 plan below is a suggestion of angle and timing — not a script.
| Time | Platform | Angle |
|---|---|---|
| 06:15 | Launch announcement. Lisa in her own voice on why $19.25 a week matters for a household in a hard year. | |
| 06:20 | Re-cut of the same LinkedIn post shortened. Story: 1 slide with the $19.25 number, link sticker to /christmas-appeal. | |
| 06:25 | Cross-post from LinkedIn with 20% re-write for Facebook audience tone. | |
| 10:00 | Optional: repost or quote-boost a supportive comment from the launch post. | |
| 13:00 | Instagram Stories | Story 2 — behind-the-scenes: a picture of the office, the audit binder on the desk, no beneficiaries. |
| 16:00 | Late-afternoon post — the one number: 87 cents in the dollar. One sentence. | |
| 19:30 | Instagram Stories | Story 3 — end-of-day gratitude for early donors. Numbers OK if delivered honestly (“X people gave today so far”). |
Never post a beneficiary photo. Never use first-name intimacy for cases. Never name a real donor without written consent. These rules apply to organic as strictly as paid. Sprint 49 will lose a few conversions to this discipline — that is the acceptable cost.
7. Escalation ladder
| Level | Trigger | Contact | Response SLA |
|---|---|---|---|
| L1 | Any red-line trip (§3) | AI CMO, immediate | 15 minutes to acknowledge; 4 hours to remediate or escalate |
| L2 | L1 unremediated > 4 hours; OR compliance-language breach in a live asset | Carla, by SMS or WhatsApp | 1 hour to acknowledge; join the launch channel |
| L3 | Regulatory enquiry received (ASIC, ACNC, OAIC); OR media enquiry with legal risk | Carla + Board Chair + external counsel | 2 hours to convene; nothing said publicly until convened |
| Payment / /donate outage | /donate returns 5xx or Stripe declines > 10% for > 15 minutes | Web-dev, immediate | Pause paid spend until restored; AI CMO holds email 3 if outage still open at Email 3 send time |
| Attribution outage | D1 marketing_sessions not writing; Meta CAPI test event failing; Google Enhanced Conversions failing | Web-dev | Investigate immediately; do not pause spend for < 4h outage; log for retro |
8. End-of-day retro — what we capture
Every day 2-24 Dec, a short retro is written into the campaign log by 22:00. The daily retros feed into the campaign-close retro published 2 Jan 2027.
Daily retro template (5 lines)
- Day + date. “Day 3 · Thu 4 Dec 2026”
- Three numbers. Page views · donate intents · donate successes · and the cumulative $ raised.
- What worked. One sentence.
- What did not. One sentence. Honest — not spun.
- Tomorrow adjustment. One sentence, or “no change.”
Campaign-close retro (published 2 Jan 2027)
- Numbers, honest. Total gifts, split weekly vs one-off. Total dollars. Retention forecast for the weekly cohort at 90 / 180 / 365 days.
- Cost. Total paid ad spend. Meta vs Google split. Blended CPA. Blended $ raised per $ spent.
- Attribution. First-touch mix. Last-touch mix. Assisted-conversion mix. What we can and cannot say about channel-level ROI.
- Compliance. Any breaches (should be zero). Any near-misses. Any Meta / Google disapprovals and how they resolved.
- Retention hand-off. The new-donor cohort is handed to Sprint 51 (retention infrastructure). Data columns exported to the CRM for stewardship cadence.
- What we’d do differently. Written plainly. This document informs Sprint 54 (EOFY 2027) and Sprint 62 (Christmas Appeal 2027).
The retro is public. The campaign-close retro is published at /christmas-appeal-2026-retro in early January 2027, linked from /marketing-kits as a Sprint 51 discovery card. The numbers are honest whether they are impressive or humbling. This is a design decision, not a communications tactic: a charity that publishes its Christmas Appeal retro plainly is a charity donors trust in Year 2.