Why this exists
Life Without Debt has excellent strategy documents. It has an ACL-rigorous compliance discipline. It has a brand voice ratified by the board. But until now, those things lived in three different documents, enforced by editorial memory. Sprint 44 codifies them into one enforceable envelope that the AI CMO applies on every asset before it goes live.
This page is the source of truth. If a rule here contradicts anything the AI CMO has done in a past sprint, this page wins. If the board updates the Constitution or the Brand Charter, this page is updated to match — not the other way round.
Machine-readable version: /data/marketing-governance.json — this is what the AI CMO reads on every asset build.
The five gates
1
Constitutional congruence
Authority: Constitution §4 · charitable purposes paramount clause
The test: Does this asset serve, or is it ancillary to, one of the charitable purposes in Constitution §4(a-d)?
What §4 says LWD exists to do:
- §4(a) — relieve poverty, sickness, suffering, distress, misfortune, helplessness and disability of persons living with terminal, life-limiting, chronic or serious illness (and their immediate families and primary carers) — including free professional debt-mediation, direct financial relief, debt discharge under the Direct Relief Policy, information/referral/casework, and training of professionals
- §4(b) — educate creditors, medical, legal community, and the Australian public — ancillary to §4(a)
- §4(c) — advance health of §4(a) persons by reducing financial stressors
- §4(d) — incidental or conducive to §4(a-c)
Checklist
- Asset furthers relief for the beneficiary class defined in §4(a) — not a general audience
- Asset does not promise services outside the charitable purposes (no tax advice, legal practice, financial planning)
- Asset does not commit the charity to non-charitable expenditure or activities
- Policy or education framing (§4(b)) is ancillary to the benevolent relief in §4(a) — never primary
Fail action: Do not ship. Route to Laurence + Carla for charitable-purposes review.
2
Compliance envelope
Authority: compliance engine · 41 active rules across 8 regulatory regimes
The test: Does the asset clear all 41 compliance rules with 0 findings when run through the audit engine?
Regime coverage: ACNC (11 rules) · Constitution (10) · DGR (4) · ATO (3) · ASIC (2) · Privacy (2) · AASB (1) · ACL / NCCP (8)
Checklist
- Asset carries ACL disclosure: names Credit Mediation Services Pty Ltd + ACL No. 387398
- Asset carries the outcome-language hedge: "individual outcomes vary" + "mediation is not debt discharge" + "we do not guarantee outcomes"
- Asset does not use forbidden phrases (see Gate 3 table)
- Asset carries ACN: pending + ABN: pending (or real registration numbers if issued)
- Complaints channel discoverable (word "complaints" + link to /enquire)
- Privacy Policy discoverable (word "privacy" + link to /privacy-policy)
- DGR / tax-deductibility claims are conditional: "once DGR endorsement is confirmed" — no unqualified claim until ATO endorsement is on the register
- Asset added to
S14_PAGESso the audit engine tests it - Post-deploy audit run confirms 0 findings on the new asset
Fail action: Do not ship. Remediate wording, redeploy, re-audit. Ship only after 0 findings.
3
Brand & editorial charter fidelity
Authority: brand-editorial-charter.html §§2-3 · cmo-position-brief.html §9 (Language house rules, binding)
The test: Does the copy conform to the five voice principles, the language house rules, and the sentence discipline?
The five voice principles:
- Plain, not simple — intelligent-friend register, not corporate press release, not hospital brochure
- Truthful about limits — never promise; some creditors write off, some reduce, some refuse
- Present, not pitying — beneficiaries are grown adults; we meet them as equals
- Specific, not abstract — "$47,000 across four cards" beats "significant consumer debt"
- Warm, not saccharine — permitted to be moving; not permitted to be maudlin
Language house rules (binding):
| Use | Do not use | Why |
|---|---|---|
| "terminally ill" (adj) · "people with a terminal diagnosis" | "the terminally ill" (as noun) | Person-first language |
| "beneficiary" (internal) · "the people we help" (external) | "client" | Only ACL-holder has clients; RG 96 boundary |
| "debt-hardship negotiation" · "professional debt-mediation" | "debt relief", "debt forgiveness", "debt cancellation", "debt elimination" | ASIC RG 96 outcome-language rule |
| "we work with creditors" · "our licensed practitioners engage creditors" | "we fight creditors", "we take on the banks" | Combative framing invites regulatory scrutiny |
| "the family" · "the household" | "your loved ones" | Saccharine; also imprecise re: STD five pathways |
| "survivors" · "surviving family" | "the bereaved" (as noun) | Person-first; agency-preserving |
| Australian English: -ise, -our, -re, -yse | -ize, -or, -er, -yze (US) | Australian audience; register signal |
Checklist
- Sentences ≤ 20 words in public copy; ≤ 25 in donor copy
- Paragraphs ≤ 4 sentences in beneficiary content; ≤ 6 in donor content
- Flesch reading grade ≤ 8 for beneficiary content; ≤ 10 for donor content
- No exclamation marks except in direct quotations
- No emoji in donor or governance content
- Australian English throughout
- Every forbidden phrase from the house-rules table above absent
- Numbers cited to source; ratios have published definitions
Fail action: Rewrite copy. Rewriting is cheaper than a brand-drift retro.
4
Funnel-stage × audience congruence
Authority: cmo-position-brief.html §§audience geometry · marketing-plan-12mo.html §3 (Story · Trust · Data · Compliance)
The test: Does this asset do the RIGHT job for the RIGHT audience at the RIGHT stage of the funnel?
Two audiences, four stages:
| Stage | Goal | Funder KPI | Beneficiary KPI | Story · Trust · Data |
|---|---|---|---|---|
| Awareness TOFU |
Make the problem legible and LWD known | Reach, impressions, brand recall, email captures | Referrer-network reach, earned media, organic SEO | 40 · 30 · 30 |
| Consideration MOFU |
Convert interest into evaluation; earn a seat at the table | Case-for-support downloads, meeting requests, qualified pipeline | Intake-form starts, clinician-guide downloads | 30 · 40 · 30 |
| Conversion BOFU |
Close: donation, MOU, referral, intake | First gift, MOU signed, grant submitted, bequest register entry | Intake form completed, case opened, household onboarded | 20 · 40 · 40 |
| Retention | Keep and grow the existing relationship | Donor retention, second-gift uplift, quarterly report opens | Post-case NPS, outcomes documented for composite bank | 40 · 30 · 30 |
Matrix rules
- Every asset declares its
funnel_stageandaudience_idin a metadata block - Awareness assets MUST NOT ask for the close (no donation CTA on cold ad creative)
- Consideration assets MUST provide a specific next action (download, subscribe, meeting) — no dead-end pieces
- Conversion assets MUST NOT introduce new education content — they close on what earlier stages taught
- Retention assets MUST cite specific past interactions ("you funded 12 hours last quarter") — generic thank-yous prohibited
- Beneficiary-audience assets clear Flesch ≤ 8; funder-audience assets clear Flesch ≤ 10
- Story · Trust · Data weights per stage followed within ± 10 percentage points
Fail action: Refactor the asset to the correct stage/audience, or move to a different stage in the funnel.
5
Industry best-practice
Authority: Fundraising Institute Australia Code of Practice · IAB / DAAS / GARM standards · M+R Benchmarks · Blackbaud Charitable Giving Report · JB Were Support Report Australia · Google Core Web Vitals · WCAG 2.2 AA
The test: Does the asset meet or exceed the observable industry standard for its type?
Principles:
- Give-back-first before ask. 3:1 ratio — three educational touches per one ask (M+R benchmark)
- One primary CTA per asset. Multiple CTAs cannibalise conversion (NN/g, MECLABS)
- Above-the-fold value. Reason to keep reading visible without scroll on mobile (375px) and desktop (1440px)
- Trust signals early. ACL 387398, ACNC status, board, or evaluation partner visible in first screen for funder assets
- Specific numbers over vague claims. "$19.25 = one hour" beats "$X = impact"
- Accessibility WCAG 2.2 AA. Contrast ≥ 4.5:1 · keyboard-navigable · alt text on meaningful images
- Mobile-first responsive. Renders on 375px width. 60%+ of nonprofit traffic is mobile (M+R 2025)
- Page load < 2.5s LCP. Google Core Web Vitals threshold. Bounce rate doubles at 3s+
- Attribution enabled. UTMs on all links; server-side conversion events via Meta CAPI, Google EC, LinkedIn CAPI
- FIA Code of Practice. No manufactured urgency; no implied endorsements; no undocumented mailing lists
- Consent-first storytelling. Real stories require seven-step consent; until 30 consented cases, all stories are labelled composites
- SEO/AEO/GEO alignment. Intent keywords (SEO), answer-engine snippets (AEO), generative-engine citations (GEO)
- Ad creative refresh cadence. Paid creative refreshed every 4-6 weeks per platform
Checklist
- Asset has ONE primary CTA (secondary links permitted but visually subordinated)
- Value proposition visible above the fold on mobile
- Trust signals in first screen (ACL 387398 OR ACNC OR board OR evaluation partner)
- WCAG 2.2 AA: contrast ≥ 4.5:1 · alt text · keyboard-navigable
- Renders on 375px mobile width without horizontal scroll
- UTM parameters + conversion event fires (if paid or landing page)
- If beneficiary story used: labelled composite OR consented per seven-step process
- No manufactured urgency; no implied endorsements; FIA Code compliant
Fail action: Fix the specific gap. Failures at this gate are usually design-side, not copy-side.
The asset lifecycle
Every marketing asset the AI CMO builds passes through these seven steps. No shortcuts, no skipping.
0BriefAudience, stage, KPI, Story/Trust/Data weights, primary CTA
1DraftAI CMO authors first pass, running Gate 3 inline
2Self-checkRun Gates 1-5 checklist, produce governance report
3AuditAdd to S14_PAGES, deploy, run audit, 0 findings required
4Human reviewOptional for high-stakes assets (major gift, EOFY, PR)
5DeployWrangler deploy, verify canonical URL, commit, push
6ReportPlain-English close-out to Carla with provenance
Retrospective — grading Sprint 43.1 against the five gates
Before I apply this framework going forward, I've graded the seven case-for-support one-pagers I built in Sprint 43.1 against the five gates. This is the discipline: if the framework can't grade the past honestly, it can't govern the future.
Sprint 43.1 · 7 Case-for-Support One-Pagers PASS
| Gate | Verdict | Evidence |
|---|---|---|
| 1. Constitutional | PASS | All 7 target beneficiary class from §4(a); Foundations/Corporate/Bequests/International serve fundraising for §4(a) work; Government serves §4(b) education framed as ancillary; Health Referrals serves §4(a)(iv) referral pathway; Regular Giving funds §4(a)(i) direct mediation |
| 2. Compliance | PASS | Audit #34 · run-17e9b54adc837cda · 107 pages · 41 rules · 0 findings |
| 3. Brand | PASS | Sentence length observed, no exclamation marks, no forbidden phrases, Australian English throughout, "beneficiary" used correctly, "outcomes vary" hedge on every page |
| 4. Funnel | PASS | All 7 are consideration-stage assets targeting funder-audience (except Room F which targets clinician-referrer). Story/Trust/Data weights approximately 30/40/30 as required. Each has ONE primary CTA (a room-specific enquire link) |
| 5. Best practice | PASS | One primary CTA per page · trust signals in first screen (room tag + case-for-support v1.0 marker) · mobile-responsive CSS with 640px breakpoint · PDF fallback for offline · WCAG contrast on bronze/cream palette · composites labelled "composite illustration" |
What this changes going forward
- Sprint 43.2 (Corporate Prospectus) — runs through the five gates before I ship it. If any gate fails, I remediate before deploy.
- Sprint 43.3 (EOFY email sequence) — each email graded individually. Awareness emails (early sequence) can't ask for close; conversion emails (late sequence) close on what earlier ones taught.
- Sprint 43.4 (Regular Giving conversion page) — must meet all Gate 5 web performance metrics (LCP < 2.5s, WCAG 2.2 AA, mobile-first). This is a BOFU asset, not an awareness page.
- Every future asset — I publish a governance report as part of the close-out, showing which gates passed and any remediation applied to reach PASS.
What triggers a framework update
The framework updates when the source of authority updates:
- Constitution amended by board resolution → Gate 1 updates
- New compliance rules added to the audit engine → Gate 2 updates
- Brand Editorial Charter revised (currently v0.1 DRAFT) → Gate 3 updates
- Marketing plan repositioned → Gate 4 stage weights and KPIs update
- Industry standards move (e.g. new WCAG version, updated FIA Code) → Gate 5 updates
Framework version is tracked in marketing-governance.json and every deploy of a new asset references the framework version it was graded under.