gooseworks-ai/goose-skills / skills/shared/meta-ads-analyzer
meta-ads-analyzer
VerifiedDiagnose Meta Ads campaign performance and account gaps using Meta's actual system mechanics — including customer-journey coverage, Breakdown Effect, Learning Phase, Auction Overlap, Pacing, and Creative Fatigue. Use for performance diagnosis, account audits, full-funnel or TOF/MOF/BOF gap analysis, deciding what to test or create next, and producing novice-friendly recommendations without forcing every campaign or ad into a funnel stage.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 18.3 KB |
| references/customer-journey-coverage.md | markdown | 4.6 KB |
| skill.meta.json | config | 146 B |
references/customer-journey-coverage.md
View on GitHub# Customer-Journey Coverage Audit
Use this reference when the user asks for an account audit, a full-funnel review, TOF/MOF/BOF gaps, or what advertising work to create next.
## Non-negotiable guardrail
Treat top, middle, and bottom of funnel as customer-journey jobs, not required campaign containers. Modern Meta accounts may use one consolidated campaign to support several jobs. Do not infer a missing stage from naming, and do not recommend creating three campaigns merely to mirror the framework.
Campaigns, ad sets, audiences, creatives, offers, landing pages, events, and performance are evidence. One item may support several jobs. Only describe an individual item as stage-specific when the evidence is unambiguous.
## 1. Identify the account model
Classify the overall operating model, not every ad:
- **Consolidated** — one or a few broad/Advantage+ campaigns pool delivery and may serve several customer jobs.
- **Funnel-segmented** — campaigns or ad sets intentionally separate prospecting, consideration, and high-intent activity.
- **Hybrid** — consolidated acquisition plus selected retargeting, offer, or retention activity.
- **Unclear** — available structure, audience, creative, or event data is insufficient.
The model is descriptive, not a quality score. Judge whether it works before recommending structural change.
## 2. Build the coverage map
Use the following customer jobs in the report. Spell out the familiar funnel term once for guided users, then prefer the plain-language job.
| Customer job | Funnel shorthand | Evidence that may support it |
|---|---|---|
| **Create demand** | Top of funnel (TOF) | New-customer reach, broad/prospecting delivery, problem awareness, category education, benefit-led hooks, introductions, first-touch landing pages |
| **Build consideration** | Middle of funnel (MOF) | Demonstrations, testimonials, reviews, proof, comparisons, objection education, engaged/site-visitor follow-up, product-detail or comparison pages |
| **Convert intent** | Bottom of funnel (BOF) | Offer and price clarity, guarantees, urgency, objection handling, product/cart/checkout treatment, purchase optimization, high-intent landing pages |
For each customer job, report:
1. What evidence exists
2. Whether there is a gap, weak coverage, or no demonstrated gap
3. Why it matters to this account's objective
4. Confidence in the conclusion
5. The smallest useful test
Do not allocate every ad to exactly one row. Cite the same ad or campaign in more than one row when it genuinely supports multiple jobs.
## 3. Diagnose gap types
Use one or more of these evidence-based gap types:
- **Coverage gap** — the account has sufficient data, but a customer job has no meaningful support.
- **Message gap** — delivery exists, but the creative mix lacks an important message such as proof, demonstration, comparison, offer, or objection handling.
- **Handoff gap** — the ad promise, audience intent, landing page, and optimization event do not form a coherent next step.
- **Delivery gap** — useful coverage exists but cannot spend, learn, or reach enough of the intended audience.
- **Measurement gap** — tracking or downstream data is insufficient to judge whether the journey works.
- **No demonstrated gap** — the available evidence shows adequate support; do not invent work.
Do not call low spend a funnel gap without checking learning state, pacing, audience overlap, bid strategy, and sample size. Do not call weak conversion a creative gap without checking the landing page and conversion event.
## 4. Confidence rules
- **High** — structure, audience, creative, destination, event, and performance evidence agree.
- **Medium** — the conclusion is supported, but one important evidence source is missing or ambiguous.
- **Low** — the conclusion relies on names, partial exports, screenshots, or low-volume data. Frame it as a question or data request, not a finding.
Never convert `Low` confidence into a red/missing verdict. State what would raise confidence.
## 5. Recommendation rules
Recommend the smallest test that addresses the diagnosed gap. Examples:
- Missing proof → test testimonial, review, demonstration, or comparison creative before restructuring campaigns.
- Weak high-intent treatment → test clearer offer, guarantee, objection handling, or landing-page continuity.
- Measurement gap → repair the relevant event or downstream data before judging performance.
- Consolidated campaign with complete coverage → preserve the structure; improve the specific message, asset, or handoff that is weak.
Never modify the account automatically. Present recommendations as hypotheses with expected impact, success criteria, and rollback.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |