The metrics surface for the 1 Oct 2026 → 31 Dec 2026 launch window. Daily, weekly, and monthly cadences. Every metric names its owner, its trigger threshold, and the escalation it fires. This is the operating layer, not the reporting layer.
Purpose. A single page a founder or triad member can open on Monday morning and know: what am I looking at this week, what does “normal” look like, and what would trigger a call. Companion to /launch-90-days (the ops calendar) and /launch-reactive-playbook (what to do when a metric breaches). Not a public reporting surface — internal working document, noindex, linked from portal only.
Compliance envelope. Every metric preserves: ABN, ACN, ACL 387398, mediation disclosure, complaints and privacy links, dignity floor. No beneficiaries are named or identified. Y1 audited numbers, when referenced as baselines, appear verbatim (31 cases · $1.82M debt resolved · 23-day average resolution · 52% average K10 drop · $1,840 average cost/case · 87c direct-service ratio).
If the triad is only checking one thing, this is it.
Seven signals Corrina reads each morning. If all seven are green she posts a two-line all-clear. If any is amber she posts the metric name and the number. If any is red she pings Lisa directly before the Slack post. No dashboards — the read is manual by design so a human eye is on every number.
/donate in the last 24h. Green: 0. Amber: 1–2 (transient). Red: 3+ or a single 5xx during a scheduled campaign send — fires page to Laurence.complaints@lifewithoutdebt.org.au unread >24h. Green: 0. Amber: 1. Red: 2+ — complaints charter breach, fires call to Chair per /complaints.Two lines to Slack at 9:00 AEDT, every weekday of the 90 days. No exceptions except the dark period (25–31 Dec — see below):
[Date] · Daily read complete. 7/7 green. — the all-clear.
[Date] · Metric [n] amber: [number]. Action: [what Corrina is doing]. — the single-amber post.
[Date] · RED on Metric [n]. Lisa · Laurence pinged directly. Slack post follows. — the escalation-first post.
The Monday triad pack is the primary weekly artefact of the launch window. Corrina compiles Sunday evening from Fri close numbers. Lisa reads before her Monday 08:30 stand. Laurence reads before he opens code. Chair reads before end of Monday.
| Metric | Owner | Green | Amber | Red · action |
|---|---|---|---|---|
| 1 · Founding-donor touch integrity | Corrina | Week's touches sent on schedule, replies routed <24h | Any delay 24–48h | Any delay >48h — escalate to Lisa, Sprint 57 impact review |
| 2 · Christmas Appeal donation total (once live) | Corrina | ≥80% of prior-week rolling avg | 50–79% | <50% — reactive playbook scenario A |
| 3 · New donor count (Christmas Appeal window) | Corrina | Trend flat or up week-on-week | Down 10–25% | Down >25% — reactive playbook scenario B |
| 4 · Donor-list health (unsubs, bounces) | Corrina | Unsub <0.5%, bounce <1% | Unsub 0.5–1.5% or bounce 1–3% | Unsub >1.5% or bounce >3% — list-quality review, playbook scenario C |
| 5 · Referrer follow-up cadence | Corrina | All 5 briefs on T+0/T+8/T+21/T+40 cadence | 1 target behind cadence | 2+ targets behind — Lisa personal outreach |
| 6 · Content-calendar adherence | Corrina | ≥90% of week's scheduled items shipped | 75–89% | <75% — triad review of Sprint 56 scope realism |
| 7 · Voice-fill status | Corrina | All week's send-day pieces voice-filled by Wed prior | Any piece voice-fill Thu-of-send | Any piece shipping without voice-fill — see § voice-fill below |
| 8 · Complaints-log activity | Chair reviews | 0 or all closed within 5 business days | 1 open >5 days | 2+ open OR any open >10 days — Chair Slack to full triad, board note |
| 9 · Privacy incidents | Chair reviews | 0 | N/A — there is no amber | Any incident — immediate OAIC scoping, playbook scenario D |
| 10 · Google Ad Grant compliance (once claimed) | Corrina | ≥5% CTR site-wide, no policy violations | CTR 4–4.9%, no violations | CTR <4% or any policy flag — account-suspension risk, pause spend |
| 11 · Portal uptime | Laurence | ≥99.9% | 99.5–99.89% | <99.5% — incident review, Laurence writes post-mortem |
| 12 · Marketing Hub piece-count · approval status | Corrina | New pieces flow to Hub within 48h of publication, triad-approval status current | Pieces published but not yet on Hub | Pieces bypass Hub entirely — governance review, Sprint 57 process fix |
PDF or web link, no longer than 4 pages, structured:
The board reputation surface. Not a marketing dashboard — it is the surface the board looks at to confirm the marketing function is operating within governance. Chair signs. Filed under board minutes. Companion file: /marketing-governance.
| Metric | Definition · source of truth | Board flag |
|---|---|---|
| 1 · Y1-numbers integrity month total | Total published content this month citing Y1 numbers; verbatim-check pass-rate. | <100% pass — board note |
| 2 · Exclude-phrase presence month total | All public-facing pages published this month with the exclude phrase verbatim in the footer. | Any absence — board note |
| 3 · Beneficiary-naming audit month total | Zero-tolerance metric. Confirms 0 beneficiaries named across all content this month. | Any breach — extraordinary board meeting |
| 4 · Complaints resolution SLA | All complaints received this month, resolved within 5 business days (per /complaints charter). | Any >5 days — board note |
| 5 · Privacy incident month count | Any breach or near-miss involving personal information. | Any — board escalation |
| 6 · ACL 387398 disclosure integrity | All content this month referencing mediation carries the ACL disclosure verbatim; CMS-vs-LWD distinction preserved. | Any omission — board note |
| 7 · Marketing Hub governance coverage | All content published this month has a Hub filing row with triad-approval status recorded. | <100% — process fix required |
| 8 · Cost/case efficiency snapshot | Y1 baseline: $1,840 / case. Current-month projected trend from live pilot data, if any. Reported alongside Y1 baseline verbatim. | Report-only, no flag threshold in Y1 launch window |
When a metric fires, this is what happens. No metric fires without a named human being pinged — dashboards do not raise alarms, humans do.
The dark period is a governance decision, not a metrics gap. Between 25 Dec (post-Christmas) and 31 Dec 2026 the marketing function operates in low-activity mode. No public sends. No paid ad activity. Only the founding-donor Christmas warmth touch (23 Dec, sent day-of before dark period opens) and Q1 impact snapshot (30 Dec) proceed — both governed under founding-donor cadence.
What Corrina reads: Metrics 3 (complaints), 6 (beneficiary-naming), 7 (exclude-phrase), and 9 (privacy) only. All others suspend. If Corrina takes leave in this window (encouraged), Chair covers the four zero-tolerance reads via phone-only escalation. The dashboard resumes full read on Mon 4 Jan.
Why: the 90-day launch is not a sprint at 100% intensity for 90 days. Steward-mode weeks are stewardship, not throughput. Dark-period discipline is what makes the 12-week peak sustainable.
Voice-fill is the single most constraining metric on Sprint 56 delivery. It is not on the daily read because it does not need daily surveillance — it needs a weekly forward-look and a monthly retrospective.
Weekly forward-look (Corrina, Monday pack Page 3): for every scheduled send this week, is voice-fill in hand by end-of-Wednesday? [Lisa voice-check pending] or [Laurence voice-check pending] markers on scaffolds count as voice-fill NOT in hand.
Monthly retrospective (Chair, board pack): what fraction of the month's shipped content shipped with voice-fill complete vs shipped with placeholder retained? Y1 launch-window target: 100% voice-fill by end of Sprint 56 for founding-donor and Christmas Appeal streams; 80% for referrer briefs (some retain composite voice by design).
Escalation: if any founding-donor touch would ship without Lisa voice-fill, Corrina pings Lisa personally 72 hours before send. If Lisa cannot voice-fill in that window, the touch delays — it does not ship generic. Founding-donor stewardship cadence has no acceptable generic-scaffold fallback.
Every metric has a named human owner. This is not a role table — it is a name table. Roles rotate; names commit.
| Owner | Watches | Escalation authority |
|---|---|---|
| Corrina Marketing Operator | All 7 daily metrics · compiles Monday pack · sends all Slack posts · routes founding-donor replies to Lisa | Amber posts to triad. Red pings named owner. Cannot make triad-halt decision — escalates. |
| Lisa Founder-voice / Chair | Founding-donor cohort integrity (metric 1) · voice-fill status (all voice-carrying content) · content-calendar realism at monthly board read | Can halt any founding-donor send. Can voice-fill or delegate. Chairs board reputation surface monthly. |
| Laurence Founder / CTO surface | Portal uptime (metric 11) · donate-page reliability (daily metric 2) · Marketing Hub piece-flow (metric 12) · voice-fill on his own voice-carrying pieces | Can halt any send whose donate-flow or infrastructure is not ready. Post-mortem author on any red-portal event. |
| Chair Board Chair, distinct from Lisa | Complaints (metric 8, weekly) · privacy (metric 9, weekly) · monthly board reputation surface (all 8 monthly metrics) · extraordinary escalations | Sole authority on any zero-tolerance red (beneficiary-naming, exclude-phrase, privacy). Can convene extraordinary board meeting. |
| Chris CMS mediation counterparty | Not in the marketing metric surface. Referenced only when ACL 387398 disclosure integrity (monthly metric 6) requires cross-check with CMS. | Consultative only in marketing surface. Full authority within CMS mediation surface. |
The dashboard is not a report. It is a governance surface. Its purpose is not to make the marketing function look good on paper — its purpose is to make sure the marketing function stays inside the envelope. Green all-clear posts are what “governed” looks like. Amber is what “working as designed” looks like. Red is what “the system detected something before it became a story” looks like.
What comes off this dashboard for external audiences (donor updates, board notes, referrer conversations) is a filtered subset with narrative wrap. That translation happens in /marketing-governance. It never happens automatically.