Systems Lab

Agent skill

facebook-ads

Meta Ads (Facebook & Instagram) platform expertise.

slowingNeeds a keyActs undeclared1,084 words

Filed under ABM and paid.

From thatrebeccarae/claude-marketing · 42 skills · 119 · pushed 2026-05-14

What it does when it runs

Meta Ads (Facebook & Instagram) platform expertise. Audit campaigns, audiences, creative strategy, pixel tracking, and CAPI. Use when the user asks about Facebook Ads, Instagram Ads, Meta Ads, social media advertising, Advantage+ campaigns, or Meta pixel/CAPI setup.

Read from the skill and the 6 files bundled beside it. A skill’s own description is written to be selected by an agent, so it describes the job and not the dependencies.

Keys and connectors you must supply
  • META_ACCESS_TOKEN
  • META_APP_SECRET
  • YOUR_ACCESS_TOKEN
Hosts it reaches
  • connect.facebook.net
  • graph.facebook.com
Tool permissions it declares
No allowed-tools in the frontmatter. It does act, so it runs under whatever permissions your session already grants.
Actions present in the files
shell

Ask about facebook-ads

Opens your assistant with this page's verified links already in the prompt.

Is this safe to install?ClaudeChatGPT
Adapt it to my stackClaudeChatGPT
What else do I need for it to workClaudeChatGPT
Rather ask a human? Talk to Cheetah
git clone --depth 1 --filter=blob:none --sparse https://github.com/thatrebeccarae/claude-marketing.git /tmp/claude-marketing
git -C /tmp/claude-marketing sparse-checkout set "skills/facebook-ads"
mkdir -p ~/.claude/skills/facebook-ads
cp -R "/tmp/claude-marketing/skills/facebook-ads/." ~/.claude/skills/facebook-ads/

Picked up without a restart. A project skill of the same name is shadowed by your personal one. For one repository only, swap ~/.claude/skills for .claude/skills. Claude Code docs ↗

The folder is the same in every client that implements the format — 46 of them — so if yours is not above, only the destination changes.

Before you install: this skill will not complete its job on a bare agent. It needs META_ACCESS_TOKEN, META_APP_SECRET, YOUR_ACCESS_TOKEN, which you have to obtain separately.

Reproduced in full from thatrebeccarae/claude-marketing/blob/a8a63ec1341f05ec9c1e9cb52b4edeb14e3bdcba/skills/facebook-ads/SKILL.md, which is licensed MIT (skill frontmatter). 1,084 words, 15 headings.

Meta Ads (Facebook & Instagram)

Expert-level guidance for Meta Ads — auditing, building, and optimizing campaigns across Facebook, Instagram, Messenger, and the Audience Network.

Install

git clone https://github.com/thatrebeccarae/claude-marketing.git && cp -r claude-marketing/skills/facebook-ads ~/.claude/skills/

Core Capabilities

Campaign Auditing & Optimization

  • Full account audit: structure, objectives, audiences, creative, tracking, attribution
  • Identify creative fatigue, audience overlap, and budget inefficiencies
  • Diagnose performance drops (iOS privacy, audience saturation, creative burnout)
  • Advantage+ Shopping and Advantage+ Audience recommendations

Campaign Objectives (Outcome-Based)

  • Awareness — Reach, brand awareness, video views
  • Traffic — Link clicks, landing page views
  • Engagement — Post engagement, messages, conversions on-platform
  • Leads — Instant Forms, Messenger, website lead gen
  • App Promotion — App installs, app events
  • Sales — Website conversions, catalog sales, Advantage+ Shopping

Audience Strategy

  • Core Audiences — Demographics, interests, behaviors, location
  • Custom Audiences — Website (pixel/CAPI), customer list, app activity, video viewers, engagement
  • Lookalike Audiences — 1-10% based on custom audience seed (value-based LALs preferred)
  • Advantage+ Audience — Meta's AI-driven targeting (audience suggestions as signals, not constraints)
  • Broad targeting — No interests/LAL, rely on creative + pixel data + Meta's ML

Creative Strategy

  • Ad format selection: single image, carousel, video, collection, Instant Experience
  • Creative testing frameworks: concept testing vs iterative testing
  • UGC (user-generated content) integration and creator whitelisting
  • Dynamic Creative Optimization (DCO) vs manual A/B testing
  • Creative fatigue signals and refresh cadence
  • Platform-specific creative (Feed, Stories, Reels, Explore)

Tracking & Attribution

  • Meta Pixel — Standard events, custom events, microdata, aggregated event measurement
  • Conversions API (CAPI) — Server-side tracking for privacy resilience
  • Aggregated Event Measurement (AEM) — iOS 14.5+ handling, 8-event prioritization
  • Attribution settings: 7-day click + 1-day view (default), 1-day click, 28-day click
  • UTM parameter strategy for GA4 cross-reference
  • Conversion lift studies and incrementality testing

Key Benchmarks

MetricGoodGreatWarning
CTR (Link Clicks)1-2%3%+<0.8%
CPC (Link Click)$0.50-$1.50<$0.50>$2.00
CPM$8-$15<$8>$20
Conversion Rate (Landing Page)3-5%8%+<2%
ROAS (ecommerce)3-4x6x+<2x
Cost Per LeadIndustry dependentBelow industry avgRising trend
Frequency (prospecting)1.5-2.51-1.5>3.5
Frequency (retargeting)3-62-3>10
Thumb-Stop Rate (video)25-35%40%+<20%
Hook Rate (3s video views %)30-40%50%+<25%

Campaign Structure (Modern Best Practice)

Ad Account
├── Advantage+ Shopping Campaign (ecommerce)
│   ├── Broad targeting (Meta AI optimizes)
│   ├── Existing customer budget cap (10-20%)
│   └── Mixed creative (image, video, UGC, carousel)
├── Prospecting Campaign (Conversions/Sales)
│   ├── Ad Set: Broad (no targeting, trust the pixel)
│   ├── Ad Set: Lookalike 1-3% (value-based)
│   └── Ad Set: Interest stacks (if needed)
├── Retargeting Campaign (Conversions/Sales)
│   ├── Ad Set: Website visitors 1-30 days
│   ├── Ad Set: Engaged (video viewers, IG/FB engaged)
│   └── Ad Set: Cart abandoners
├── Lead Gen Campaign (if applicable)
│   ├── Ad Set: Lookalikes of converters
│   └── Ad Set: Interest-based
└── Brand / Top-of-Funnel (optional)
    ├── Video Views (awareness content)
    └── Traffic (blog, resources)

Modern trend: Consolidate ad sets. Meta's algorithm performs better with fewer, broader ad sets and more creative diversity within each.

Workflow: Full Account Audit

When asked to audit a Meta Ads account:

  1. Pixel & CAPI Health — Event tracking, CAPI coverage, event match quality score, AEM config
  2. Account Structure — Campaign consolidation, objective alignment, naming conventions
  3. Audience Strategy — Overlap analysis, LAL seeds, audience size, Advantage+ adoption
  4. Creative Analysis — Format mix, creative fatigue (frequency + declining CTR), testing velocity
  5. Budget & Bidding — CBO vs ABO, budget allocation, bid strategy (lowest cost vs cost cap vs bid cap)
  6. Funnel Coverage — Prospecting vs retargeting balance, exclusions between funnels
  7. Attribution — Window settings, cross-platform reconciliation (vs GA4), CAPI deduplication
  8. Placement Optimization — Advantage+ placements vs manual, creative per placement
  9. Shopping / Catalog — Feed quality, product sets, dynamic ads setup
  10. Recommendations — Prioritized with expected impact and creative direction

Post-iOS 14.5 Best Practices

  • Implement CAPI alongside pixel (aim for 90%+ event match quality)
  • Prioritize 8 conversion events per domain in Events Manager
  • Use value optimization when possible (Purchase over Add to Cart)
  • Broader audiences outperform narrow (give Meta's ML room to optimize)
  • UTM tracking + GA4 as secondary attribution source
  • Creative is the new targeting — invest in creative testing over audience testing

How to Use This Skill

Ask me questions like:

  • "Audit my Facebook Ads account performance"
  • "My ROAS is declining — what should I investigate?"
  • "Help me set up Conversions API (CAPI)"
  • "Design a creative testing framework for my brand"
  • "Should I use Advantage+ Shopping or manual campaigns?"
  • "Build a full-funnel campaign structure for my DTC brand"
  • "My frequency is too high — how do I manage audience fatigue?"
  • "Plan a lead gen campaign for B2B on Facebook/Instagram"

For detailed Meta Ads API reference, pixel implementation, and advanced configurations, see REFERENCE.md.

Hard Rules

These constraints must never be violated in recommendations:

  1. Pixel + CAPI both required. Post-iOS 14.5, browser-only tracking loses 30-40% of conversion data.
  2. Event deduplication must be active (event_id matching) — without it, conversions are double-counted.
  3. Event Match Quality ≥8.0 for Purchase event. Below 6.0 is critical.
  4. Budget must be ≥5x target CPA per ad set — below this, the algorithm cannot exit learning phase.
  5. Never recommend edits during active learning phase — wait for ~50 conversions/week or intentional reset.
  6. Creative fatigue = action required. CTR decline >20% over 14 days with frequency >3 = replace creative immediately.
  7. <30% of ad sets in Learning Limited — above this threshold, consolidation is mandatory.
  8. Purchasers/converters must be excluded from prospecting campaigns.
  9. Special Ad Categories must be declared before campaign creation for Housing, Employment, Credit, and Financial Products.

Scored Audit

When performing an account audit, load skills/shared/scoring-system.md for the weighted scoring algorithm and CHECKS.md for the 46-check Meta Ads audit checklist. Produce a health score (0-100, grade A-F) with Quick Wins and a prioritized action plan.

Files bundled with it

These load only when the skill asks for them, so they cost nothing until it runs.

Other skills for the same job

Different authors, same problem. Matched on the words in the skill name, across every library in the catalogue except this one.

Need help setting it up?

This page tells you what facebook-ads does and what it needs. Cheetah builds the agent setup it runs inside: data, CRM, sequencing and the guardrails.

Book a call →

The directory stays free. There is nothing gated behind this.