By name: meta-ads-performance-optimizer description: Analyze and optimize running Meta Ads campaigns - diagnose underperformance, make scaling decisions, reallocate budgets, identify when to kill vs scale ads, and troubleshoot issues. Use when users ask about campaign performance, ROAS/CPA issues, scaling, why ads aren't working, or optimization recommendations. Triggers include "optimize campaign", "why is my ROAS dropping", "should I scale", "campaign not performing", "CPA is too high", "ads not spending", "learning limited", or any performance troubleshooting.
Performance Optimizer Skill
Analyze, diagnose, and optimize Meta Ads campaigns for ecommerce.
Skill Overview
The Performance Optimizer skill helps diagnose campaign issues, make data-driven optimization decisions, and systematically improve performance. It covers the full optimization lifecycle from diagnosis to scaling.
Core Principle: Andromeda-Era Optimization (2026)
OPTIMIZATION HAS SHIFTED
OLD APPROACH:
- Micro-manage audiences
- Constant bid adjustments
- Manual A/B testing at ad set level
- Interest/behavior tweaks
MODERN APPROACH (2026):
- Focus on creative refresh
- Let algorithm optimize targeting
- Minimize edits during learning
- Scale through creative diversity
- Monitor, don't micro-manage
KEY INSIGHT: 70% of optimization impact comes from creative
Audience and bid optimization impact is declining with Andromeda
When to Use This Skill
Use performance-optimizer when:
- Campaign performance is declining
- CPA/ROAS not meeting targets
- Deciding whether to scale or kill campaigns
- Diagnosing why ads aren't spending
- Dealing with "Learning Limited" status
- Analyzing creative performance
- Planning budget reallocation
- Troubleshooting delivery issues
Reference Files
When to Load Each Reference
| File | Load When |
|---|---|
references/diagnostic-frameworks.md |
Diagnosing any performance issue, initial analysis |
references/scaling-playbook.md |
Deciding to scale, how to scale, scaling troubleshooting |
references/budget-optimization.md |
Reallocating budgets, budget decisions, spend issues |
references/creative-analysis.md |
Analyzing creative performance, fatigue, refresh decisions |
references/learning-phase.md |
Learning Limited issues, new campaign optimization |
references/metrics-benchmarks.md |
Understanding if metrics are good/bad, benchmark comparisons |
references/troubleshooting-guide.md |
Specific problems (not spending, rejected, delivery issues) |
Diagnostic Decision Tree
PERFORMANCE ISSUE IDENTIFIED
│
├─► What's the symptom?
│ │
│ ├─► CPA/ROAS degrading
│ │ ├─► Sudden drop (>30% in 3 days)?
│ │ │ └─► Check: Algorithm change, creative fatigue, competition
│ │ │ Load: diagnostic-frameworks.md
│ │ │
│ │ └─► Gradual decline over weeks?
│ │ └─► Check: Audience saturation, creative fatigue, market changes
│ │ Load: creative-analysis.md
│ │
│ ├─► Not spending budget
│ │ ├─► New campaign?
│ │ │ └─► Check: Learning phase, bid strategy, audience size
│ │ │ Load: learning-phase.md
│ │ │
│ │ └─► Established campaign?
│ │ └─► Check: Audience exhaustion, bid caps, creative approval
│ │ Load: troubleshooting-guide.md
│ │
│ ├─► Learning Limited
│ │ └─► Load: learning-phase.md
│ │
│ ├─► High CPM
│ │ └─► Check: Audience competition, seasonality, relevance
│ │ Load: diagnostic-frameworks.md
│ │
│ ├─► Low CTR
│ │ └─► Check: Creative relevance, hook quality, audience mismatch
│ │ Load: creative-analysis.md
│ │
│ └─► Low Conversion Rate
│ └─► Check: Landing page, offer, tracking, audience intent
│ Load: diagnostic-frameworks.md
Quick Diagnostic Framework
Step 1: Identify the Symptom
PRIMARY METRICS TO CHECK:
□ ROAS / CPA - Are you profitable?
□ Spend - Is budget delivering?
□ CPM - What are you paying for impressions?
□ CTR - Are people clicking?
□ CVR - Are clicks converting?
FORMULA: CPA = CPM ÷ (CTR × CVR × 1000)
If CPA is high, one of these is the problem
Step 2: Isolate the Cause
HIGH CPM (Impression cost issue):
→ Competition, audience overlap, low relevance, seasonal
LOW CTR (Creative/relevance issue):
→ Weak hook, audience mismatch, creative fatigue, ad blindness
LOW CVR (Conversion issue):
→ Landing page, offer, audience intent, tracking issues
Step 3: Apply the Fix
CPM FIX: Broaden audience, improve relevance, reduce overlap
CTR FIX: New hooks, refresh creative, test audiences
CVR FIX: Landing page optimization, offer testing, audience refinement
Optimization Workflow
Daily Checks (5 min)
□ Spend pacing (on track for daily budget?)
□ CPA/ROAS vs target (within acceptable range?)
□ Any delivery issues (paused, rejected, learning limited?)
□ Frequency check (audience fatigue?)
ACTION THRESHOLD:
- Only act if >20% deviation from baseline
- Wait 24-48 hours before making changes
- Document observations
Weekly Analysis (30 min)
□ Week-over-week performance comparison
□ Creative performance ranking
□ Audience performance comparison
□ Budget allocation review
□ Frequency and saturation check
□ Identify winners and losers
ACTIONS:
- Kill clear losers (>30% worse than average)
- Increase budget on winners (max 20% increase)
- Plan creative refresh if fatigue detected
- Reallocate budget from underperformers
Monthly Review (2 hours)
□ Month-over-month trend analysis
□ Full creative audit
□ Audience strategy review
□ Account structure assessment
□ Competitive landscape check
□ Goal and target adjustment
ACTIONS:
- Major budget reallocations
- Campaign restructuring if needed
- New creative strategy development
- Audience expansion planning
Kill vs Scale Decision Framework
When to Kill
KILL CAMPAIGN/AD SET WHEN:
✗ CPA >50% above target for 7+ days
✗ Spent 3x target CPA with 0 conversions
✗ ROAS <50% of target for 7+ days
✗ CTR <0.5% (severe relevance issue)
✗ Frequency >5/week with declining performance
✗ No improvement after 2 optimization attempts
DON'T KILL IF:
- Still in learning phase (wait for 50 conversions)
- Only 24-48 hours of data
- Just made changes (wait 3-5 days)
- Seasonal dip expected
When to Scale
SCALE CAMPAIGN/AD SET WHEN:
✓ CPA at or below target for 7+ days
✓ ROAS at or above target consistently
✓ 50+ conversions in learning period
✓ Frequency <3/week (room to scale)
✓ Stable or improving trend
HOW TO SCALE:
- Vertical: 15-20% budget increase every 3-5 days
- Horizontal: New audiences, geos, or creative
- Never increase >30% at once
Andromeda-Era Optimization Priorities
HIGHEST IMPACT (Focus here):
1. Creative refresh and diversity (70% of impact)
2. Landing page optimization
3. Offer/value proposition
MODERATE IMPACT:
4. Budget allocation between campaigns
5. Bid strategy selection
6. Campaign structure
LOWER IMPACT (Don't over-optimize):
7. Audience micro-adjustments
8. Placement tweaks
9. Schedule optimization
Quick Reference: Common Issues
| Issue | First Check | Quick Fix |
|---|---|---|
| CPA rising | Creative fatigue | New creative variations |
| Not spending | Audience size, bid | Broaden targeting, raise bid cap |
| Learning Limited | Conversions/week | Consolidate, broaden, raise budget |
| High CPM | Competition, overlap | Broaden audience, check overlap |
| Low CTR | Creative relevance | New hooks, test creative |
| Low CVR | Landing page | LP audit, offer test |
| High frequency | Audience saturation | Expand audience, refresh creative |
Integration with Other Skills
| Skill | Integration Point |
|---|---|
| campaign-launcher | Diagnosis informs new campaign setup |
| audience-architect | Audience performance informs targeting changes |
| creative-studio | Performance data drives creative briefs |
Optimization Principles
1. PATIENCE: Wait for data before acting (50+ conversions)
2. INCREMENTALISM: Small changes, measure impact
3. DOCUMENTATION: Track what you changed and why
4. CREATIVE-FIRST: Most problems are creative problems
5. TRUST THE ALGORITHM: Don't micro-manage in 2026
6. CONSOLIDATION: Fewer campaigns = more data = better optimization