📣 Ad Performance Auditor

Skill Purpose

Analyze the performance of existing ad campaigns based on user-provided data (screenshots, descriptions, or metrics). Evaluate key performance indicators against industry benchmarks, detect creative fatigue and audience overlap, identify wasted budget, and provide prioritized optimization recommendations ranked by expected impact. This is a diagnostic tool—it shows what’s working, what’s broken, and what should be fixed first.

When to Use

• The user wants to audit their existing ad campaigns
• The user shares ad performance data (screenshots, CSV, or described metrics)
• The user asks "why aren't my ads working?" or "how can I improve my ads?"
• The user wants to know whether their ad spend is efficient
• The user has been running ads for 7+ days and wants a performance check
• Activated by /ads audit or /ads audit <plataforma>

Data Collection

Step 1: Gather Performance Data

Ask the user to provide their ad data in any of these formats:

Option A: Key Metrics (Manual Entry) Request these metrics by campaign/ad set:

Option B: Screenshot Analysis If the user shares screenshots of their ad dashboard:

• Extract all visible metrics from the screenshot
• Note which metrics are missing
• Ask follow-up questions about missing critical data

Option C: Platform-Specific Export Guide the user to export the data:

• Goal: Ads Manager > Export > Select columns
• Google: Reports > Download
• TikTok: Analytics > Export
• LinkedIn: Campaign Manager > Export

Stage 2: Establish Industry Benchmarks

Use these benchmark ranges by platform and objective. Compare the user's metrics with these:

Meta (Facebook/Instagram) Benchmarks:

Google Ads Benchmarks (Search):

Google Ads Benchmarks (Display):

TikTok Benchmarks:

LinkedIn Benchmarks:

Adjustments by Industry Vertical:

• E-commerce: CPA benchmarks $15-$45, ROAS target 3-5x
• SaaS/B2B: CPA benchmarks $50-$200, focus on CPL rather than CPA
• Local Services: CPA benchmarks $20-$75, call tracking is critical
• Education/Courses: CPA benchmarks $30-$100, focus on webinar registrations
• Real estate: CPL benchmarks $20-$80, lead quality matters more than volume

Stage 3: Performance Analysis

Run each of these diagnostic checks:

3A: Efficiency Analysis

Calculate and evaluate:

• Cost Efficiency Score: How much value per dollar spent?
• Click Efficiency: CTR vs. benchmark for this platform/industry
• Conversion Efficiency: Conversion rate vs. benchmark
• ROAS Check: Is the return justifying the spend?
• Budget Utilization: Is the daily budget being fully spent? Underspending signals targeting that is too narrow.

3B: Creative Fatigue Detection

Identify whether the ads are fatigued:

Creative Fatigue Diagnosis:

• Light fatigue: CTR fell 10-20%, frequency 2-3 → Update the ad copy and images
• Moderate fatigue: CTR fell 20-40%, frequency 3-5 → New creative concepts needed
• Severe fatigue: CTR fell >40%, frequency >5 → Pause and relaunch with completely new creative

3C: Audience Analysis

Evaluate targeting effectiveness:

• Audience Size: Too broad (>10M on Meta) or too narrow (<100K)?
• Audience overlap: Multiple ad sets targeting the same people?
• Demographic Performance: Which age/gender/location segments convert best?
• Placement Performance: Which placements (feed, stories, reels, search, display) perform best?
• Performance by Device: Conversion differences between mobile and desktop?

3D: Budget Waste Identification

Find where money is being wasted:

3E: Funnel Leak Detection

Identify where conversions are being lost:

Impressões (100%) → Cliques (CTR: X%) → Visualizações da Landing Page (X%) → Conversões (X%)
                                              ^                        ^
                                              |                        |
                                         Velocidade de Carga da LP?  Conversão da LP?
                                         Taxa de Rejeição?           Atrito no Formulário?
                                         UX Mobile?                  Descompasso da Oferta?

Calculate the drop at each stage:

• Impression to Click: Is the ad compelling? (Creative issue)
• Click to Preview Landing Page: Is the page loading? (Technical issue)
• Landing Page Conversion Review: Is the offer converting? (Offer/UX problem)

Step 4: Scoring

Overall Ad Performance Score (0-100):

Score Interpretation:

• 80-100: Excellent — only minor optimizations
• 60-79: Good — solid foundation with opportunities for optimization
• 40-59: Needs Work — significant improvements needed in several areas
• 20-39: Poor — fundamental strategy problems to resolve
• 0-19: Critical — serious issues, consider pausing and restructuring

Step 5: Generate Recommendations

Produce recommendations at 3 levels, ordered by expected impact:

Level 1: Immediate Actions (This Week) Actions you can take today without additional resources:

• Pause underperforming ad sets
• Adjust budgets for top performers
• Add negative keywords (Google)
• Adjust frequency caps
• Fix audience overlap
• Update ad copy with better-performing hooks

Level 2: Short-Term Optimizations (Next 2 Weeks) Actions that require some creative or strategic work:

• Launch new creative variations
• Test new audiences
• A/B test landing pages
• Implement remarketing
• Adjust bidding strategy
• Expand to new placements

Level 3: Strategic Changes (Next 30 Days) Actions that require significant planning and effort:

• Restructure campaign architecture
• Build new funnel stages
• Create new creative concepts
• Expand to new platforms
• Redesign landing pages
• Implement conversion tracking improvements

Step 6: Projected Impact

For each Level 1 recommendation, estimate:

Output Format

Save as ADS-AUDIT.md in the current working directory.

# Auditoria de Desempenho de Anúncios: [Nome do Negócio/Campanha]
**Data da Auditoria:** [Data]
**Período Analisado:** [Intervalo de Datas]
**Plataforma(s):** [Plataformas]
**Gasto Total Analisado:** $[Valor]

---

## Pontuação Geral de Desempenho de Anúncios: [X]/100
[Barra visual de pontuação ou avaliação]

## Resumo Executivo
- **Principal Descoberta:** [Descoberta de maior impacto]
- **Maior Desperdício:** $[Valor] desperdiçado com [o quê]
- **Maior Oportunidade:** [O que poderia melhorar mais os resultados]
- **Melhoria Estimada:** [X]% de melhoria possível com as alterações do Nível 1

---

## Painel de Desempenho

### Principais Métricas vs. Padrões de Referência
| Métrica | Seu Valor | Média do Setor | Status | Diferença |
|--------|-----------|-------------|--------|-----|
| CTR | X% | X% | [status sem emojis] | +/-X% |
| CPC | $X | $X | [status] | +/-$X |
| Taxa de Conversão | X% | X% | [status] | +/-X% |
| ROAS | Xx | Xx | [status] | +/-Xx |
| Frequência | X | X | [status] | +/-X |

### Detalhamento da Pontuação
| Dimensão | Pontuação | Principal Descoberta |
|-----------|-------|-------------|
| Eficiência de Custo | X/25 | [descoberta] |
| Desempenho Criativo | X/20 | [descoberta] |
| Qualidade da Conversão | X/25 | [descoberta] |
| Segmentação de Público | X/15 | [descoberta] |
| Otimização de Orçamento | X/15 | [descoberta] |

---

## Análise Detalhada

### Avaliação de Fadiga Criativa
[Nível de fadiga, evidências, recomendação]

### Análise de Público
[Avaliação de segmentação, problemas de sobreposição, desempenho por segmento]

### Relatório de Desperdício de Orçamento
| Tipo de Desperdício | Desperdício Estimado | Correção |
|------------|----------------|-----|
| [tipo] | $[valor] | [correção] |

### Análise de Vazamentos no Funil
[Análise de queda de conversão etapa por etapa]

---

## Recomendações

### Nível 1: Ações Imediatas (Esta Semana)
| # | Ação | Impacto Esperado | Esforço |
|---|--------|----------------|--------|
| 1 | [ação] | [impacto] | Baixo |

### Nível 2: Curto Prazo (Próximas 2 Semanas)
| # | Ação | Impacto Esperado | Esforço |
|---|--------|----------------|--------|
| 1 | [ação] | [impacto] | Médio |

### Nível 3: Estratégico (Próximos 30 Dias)
| # | Ação | Impacto Esperado | Esforço |
|---|--------|----------------|--------|
| 1 | [ação] | [impacto] | Alto |

---

## Impacto Projetado das Alterações do Nível 1
| Ação | Atual | Projetado | Confiança |
|--------|---------|-----------|------------|
| [ação] | [valor] | [valor] | Alta/Média/Baixa |

## Próximos Passos
1. Implementar as ações do Nível 1 imediatamente
2. Reauditar em 7 dias com `/ads audit`
3. Usar `/ads creative` para gerar novos criativos se a fadiga for detectada
4. Usar `/ads funnel` para reestruturar a arquitetura da campanha, se necessário
5. Gerar novos textos com `/ads copy <plataforma>`

Important Rules

• Never make assumptions about missing data — request it or flag it as unavailable
• Always compare with platform-specific benchmarks, not averages across platforms
• The vertical matters—adjust the benchmarks accordingly
• Frequency is one of the most overlooked metrics — always check it
• ROAS alone is not enough—consider contribution margin and LTV
• Identifying budget waste should always include estimated dollar amounts
• Recommendations should be specific and actionable, not generic best practices
• Tier 1 recommendations should be feasible without requiring new creative assets
• Always account for seasonality when evaluating performance trends
• If the data is insufficient for a full audit, state that clearly and specify what's needed
• A declining CTR with stable conversions may be acceptable — context matters
• Never recommend increasing a campaign budget if there are fundamental creative or targeting issues

--- skills/ads-hooks/SKILL.md ---

name: Scroll-Stopping Hook Generator description: Generates 20+ scroll-stopping ad hooks organized by psychological angle (pain, curiosity, social proof, contrarian, urgency), with platform-specific formatting, A/B variations, and psychological analysis