Install this Claude Code skill.
Name: Meta Ads Analyzer
Save to: ~/.claude/skills/meta-ads-analyzer.md

---
---
name: Meta Ads Analyzer
description: Analyze Meta ad campaign performance and generate optimization recommendations
---

Analyze the user's Meta (Facebook/Instagram) ad campaign data and provide actionable optimization recommendations.

## Instructions

### Step 1: Gather Campaign Data
Ask the user to provide their campaign data. They can either:
- Share a CSV export from Ads Manager
- Share a screenshot of their Ads Manager dashboard
- Describe their metrics verbally
- Connect the Meta Ads MCP for direct data access

Key metrics to collect:
- Campaign name and objective
- Date range
- Spend
- Impressions and reach
- CPM (cost per 1,000 impressions)
- Clicks and CTR (click-through rate)
- CPC (cost per click)
- Conversions and conversion rate
- CPA (cost per acquisition) or CPL (cost per lead)
- ROAS (return on ad spend) if applicable
- Frequency (average times each person saw the ad)

### Step 2: Health Check
Evaluate against benchmarks:

| Metric | Healthy | Warning | Critical |
|--------|---------|---------|----------|
| CTR (link) | >1.5% | 0.8-1.5% | <0.8% |
| CPM | <$12 | $12-20 | >$20 |
| CPC | <$1.50 | $1.50-3 | >$3 |
| Frequency | <3 | 3-5 | >5 |
| Conv. Rate | >2% | 1-2% | <1% |
| Relevance Score | >7 | 4-7 | <4 |

Flag any metrics in warning or critical range.

### Step 3: Diagnose Issues

**High CPM + Low CTR = Creative fatigue or wrong audience**
→ Refresh creative, test new angles, expand audience

**Good CTR + Low Conversion Rate = Landing page problem**
→ Check page speed, mobile experience, message match

**High Frequency + Rising CPA = Audience saturation**
→ Expand audience, add new lookalikes, test new interests

**Good metrics but low spend = Budget too low or audience too narrow**
→ Increase daily budget by 20%, expand geo or age range

### Step 4: Recommendations
Generate specific, actionable recommendations:

1. **Creative:** Which ads to kill, which to scale, what new angles to test
2. **Audience:** Which audiences are performing, where to expand
3. **Budget:** Where to reallocate spend for best ROAS
4. **Bidding:** Whether to switch bid strategy
5. **Schedule:** Best performing days/hours (if data available)

### Step 5: A/B Test Plan
Suggest 2-3 specific tests to run next:
- Test: [what to change]
- Hypothesis: [why it should improve performance]
- Duration: [how long to run]
- Success metric: [what to measure]

## Output
Present findings as a clear report with a traffic-light system (green/yellow/red) for each metric, followed by a prioritized action list.