Cross Platform Audit

Cross-Platform Audit — Unified Paid Media Assessment

Orchestrates parallel scored audits across Google Ads, Meta Ads, and Microsoft Ads, then merges results into a single budget-weighted health score with cross-platform intelligence.

Install

git clone https://github.com/thatrebeccarae/claude-marketing.git && cp -r claude-marketing/skills/cross-platform-audit ~/.claude/skills/

Context Intake (Always Do First)

Before running audits, collect this information. Without it, benchmarks will be generic and recommendations may be wrong.

Ask these questions (combine into one prompt):

1. Business type — E-commerce/DTC, SaaS/B2B, Local Service, Lead Gen, Agency, Other 2. Monthly ad spend — Total and per-platform breakdown (approximate is fine) 3. Primary goal — Revenue/ROAS, Leads/CPA, App Installs, Brand Awareness 4. Active platforms — Which platforms are they advertising on? (Google, Meta, Microsoft, others) 5. Data available — What can they provide? (exports, screenshots, pasted metrics, MCP access)

Use the provided context to:

• Determine which platform audits to run (skip platforms they don't use)
• Calculate budget share per platform for aggregate scoring
• Calibrate severity based on spend level ($5K/mo gets different advice than $100K/mo)

Orchestration Workflow

Step 1: Collect Context

Gather business type, platforms, budgets, goals, and available data per the intake above.

Step 2: Spawn Parallel Audits

Use the Task tool to run platform audits simultaneously. Each audit task should:

1. Load skills/shared/scoring-system.md for the scoring algorithm 2. Load the platform-specific CHECKS.md for the audit checklist 3. Load the platform-specific SKILL.md and REFERENCE.md for diagnostic context 4. Evaluate each applicable check as PASS, WARNING, or FAIL 5. Calculate the platform health score using weighted formula 6. Identify Quick Wins 7. Write results to a structured output

Spawn these tasks in parallel (adjust based on which platforms the user has):

Task 1: Google Ads Audit
  → Read skills/shared/scoring-system.md
  → Read skills/google-ads/CHECKS.md
  → Evaluate 74 checks against provided data
  → Calculate score → Write google-audit-results.md

Task 2: Meta Ads Audit
  → Read skills/shared/scoring-system.md
  → Read skills/facebook-ads/CHECKS.md
  → Evaluate 46 checks against provided data
  → Calculate score → Write meta-audit-results.md

Task 3: Microsoft Ads Audit
  → Read skills/shared/scoring-system.md
  → Read skills/microsoft-ads/CHECKS.md
  → Evaluate 30 checks against provided data
  → Calculate score → Write microsoft-audit-results.md

If the user only has 2 platforms, spawn only those 2 tasks.

Step 3: Merge Results

After all audit tasks complete:

1. Read all *-audit-results.md files 2. Extract per-platform scores and grades 3. Calculate budget-weighted aggregate score:

   Aggregate = Σ(Platform_Score × Platform_Budget_Share)

4. Run cross-platform analysis (see Cross-Platform Analysis below) 5. Merge Quick Wins from all platforms into a unified priority list 6. Generate the unified report

Step 4: Cross-Platform Analysis

Add intelligence that no single-platform audit can provide:

Budget Allocation Analysis

• Is spend distributed optimally across platforms?
• Are high-performing platforms budget-constrained while low performers have excess?
• Does the platform mix match the business type? (e.g., B2B without LinkedIn, ecom without Shopping)
• Compare platform CPAs and flag reallocation opportunities

Tracking Consistency

• Are conversion events defined consistently across platforms?
• Is revenue/value tracking active on all platforms or only some?
• Are attribution windows aligned? (7-day click on Meta but 30-day on Google = inconsistent reporting)
• Is there a single source of truth (GA4) or are platforms reporting independently?

Creative Alignment

• Is messaging consistent across platforms or contradictory?
• Are creative formats appropriate per platform? (vertical video for Meta/TikTok, text-heavy for Search)
• Is creative being refreshed at similar cadences or is one platform stale?

Attribution Overlap

• Are platforms double-counting conversions? (same user converts, Google and Meta both claim credit)
• What percentage of total reported conversions exceeds actual conversions?
• Is there a de-duplication strategy (UTMs + GA4 as arbiter)?

Kill List

• Campaigns/ad sets across ALL platforms with CPA >3x target
• Aggregate wasted spend across platforms
• Quick calculation: total monthly waste = Σ(flagged campaign spend)

Scaling Opportunities

• Which campaigns across all platforms have CPA below target AND are budget-constrained?
• What is the estimated incremental revenue from reallocating waste to winners?

Step 5: Generate Unified Report

Output the unified report in the format specified in EXAMPLES.md.

Step 6: Deck Generation (Optional)

If the user requests a deck (--deck flag, or asks for a presentation/deliverable):

• Take the unified report output
• Pipe to pro-deck-builder skill
• Generate a branded HTML slide deck with:
• Executive summary slide (aggregate score, grade, key findings)
• Per-platform score cards
• Cross-platform analysis slides
• Quick Wins slide
• Prioritized action plan slides
• Budget reallocation recommendation slide

Hard Rules

These constraints apply across all platforms and must never be violated:

1. Never recommend changes during active learning phases on any platform. 2. 3x Kill Rule is universal — flag ANY campaign on ANY platform with CPA >3x target. 3. Tracking must be verified before optimization recommendations — if conversion tracking is broken, that is the only recommendation until fixed. 4. Budget reallocation recommendations must account for minimum thresholds — don't recommend shifting $500/mo to a platform that needs $3K/mo minimum. 5. Attribution overlap must be flagged — if total platform-reported conversions exceed GA4 by >20%, call it out. 6. Special Ad Categories apply cross-platform — if housing/employment/credit/finance on one platform, compliance checks apply to all platforms.

Relationship to Other Skills

This skill orchestrates the following skills — it does not replace them:

How to Use This Skill

Ask questions like:

• "Run a full audit across all my ad platforms"
• "Score my Google and Meta accounts and give me one unified health grade"
• "Audit my paid media and build a deck I can present to my client"
• "I spend $20K on Google, $15K on Meta, and $5K on Microsoft — audit everything"
• "Which platform should I shift budget to based on performance?"
• "Give me a cross-platform audit with a prioritized action plan"

Cross-Platform Audit — Examples

1. Full Multi-Platform Scored Audit

User Request:

"I spend $20K/month on Google, $15K/month on Meta, and $5K/month on Microsoft. Run a full cross-platform audit and give me a unified health score. We're a DTC e-commerce brand doing about $150K/month in revenue from paid."

Analysis Steps: 1. Collect context: DTC e-commerce, $40K/month total, 50% Google / 37.5% Meta / 12.5% Microsoft, revenue goal 2. Spawn 3 parallel audit tasks (Google 74 checks, Meta 46 checks, Microsoft 30 checks) 3. Each task evaluates checks, calculates platform score, identifies Quick Wins 4. Merge results: budget-weighted aggregate, cross-platform analysis, unified action plan

Sample Output:

Cross-Platform Audit — DTC E-commerce — 2026-03-16

Aggregate Health Score: 68/100 (Grade C — Needs Improvement)

Aggregate = (72 × 0.50) + (65 × 0.375) + (58 × 0.125) = 36.0 + 24.4 + 7.3 = 67.6 → 68

Quick Wins (All Platforms — Fix This Week)

Cross-Platform Analysis

Budget Allocation

• Current split: Google 50% / Meta 37.5% / Microsoft 12.5%
• Recommended for DTC: Google 40-50% / Meta 35-45% / Microsoft 5-10%
• Finding: Split is reasonable. However, Google Search campaigns are budget-constrained (top performers hitting cap by 2 PM) while Microsoft has $1,200/month on underperforming syndication. Reallocate $800/month from Microsoft syndication waste to Google Search.

Tracking Consistency

• Google: Enhanced Conversions OFF — estimated 15-20% data loss on Google
• Meta: No CAPI — estimated 30-40% data loss on Meta
• Microsoft: MSCLKID disabled — all conversion data unreliable
• GA4: Linked to Google only, not receiving Meta or Microsoft data
• Assessment: Tracking is the #1 priority. No optimization recommendations are reliable until tracking is fixed on all 3 platforms. Fix tracking first, then re-audit in 30 days.

Creative Alignment

• Google: RSA headlines are product-focused ("Shop Running Shoes"), PMax uses generic lifestyle images
• Meta: Feed ads use benefit-driven copy ("Run Faster, Recover Quicker"), but 3 of 5 creatives are stale (CTR declining 25% over 14 days)
• Microsoft: Identical RSAs to Google (raw import) — not optimized for Microsoft's audience
• Assessment: Messaging is inconsistent between Google (product-focused) and Meta (benefit-focused). Align on core value proposition. Meta creative needs immediate refresh.

Attribution Overlap

• Platforms report: Google 420 purchases + Meta 380 purchases + Microsoft 45 purchases = 845/month
• GA4 reports: 620 purchases/month
• Overlap: 36% over-reporting (845 vs 620)
• Assessment: High overlap — platforms are double-counting ~225 conversions/month. Need UTM parameter consistency and GA4 as single source of truth. Current ROAS calculations on individual platforms are inflated.

Kill List

Scaling Opportunities

Net reallocation: Cut $5,400 waste → Reinvest $5,000 in winners = $400/mo savings + better performance

Prioritized Action Plan

Phase 1: Fix Tracking (Week 1)

1. Deploy Meta CAPI via Gateway 2. Enable Google Enhanced Conversions 3. Enable Microsoft MSCLKID auto-tagging 4. Set up consistent UTM parameters across all platforms 5. Link GA4 to Meta and Microsoft for unified reporting Expected impact: Accurate data across all platforms within 7 days

Phase 2: Stop the Bleeding (Week 2)

6. Pause Kill List campaigns ($5,400/mo waste eliminated) 7. Exclude Microsoft syndication partners from brand campaigns 8. Create and apply Google negative keyword lists 9. Add brand exclusions to Google PMax 10. Set Meta attribution to 7-day click / 1-day view Expected impact: $5,400/mo waste eliminated, brand cannibalization stopped

Phase 3: Reallocate and Scale (Week 3-4)

11. Increase Google Brand Search budget by $1,500/mo 12. Increase Google Shopping - Top SKUs budget by $2,000/mo 13. Increase Meta Advantage+ Shopping budget by $1,500/mo 14. Refresh stale Meta creatives (3 ad sets with fatigue) 15. Optimize Microsoft Ads beyond Google import (new negatives, LinkedIn targeting, bid adjustments) Expected impact: 15-25% CPA improvement as budget shifts from waste to winners

Phase 4: Re-Audit (Day 30)

16. Re-run cross-platform audit after tracking fixes have 30 days of clean data 17. Recalculate health scores with accurate conversion data 18. Adjust budget allocation based on true platform CPA (not inflated by attribution overlap)

Expected Outcomes (90-Day Projection)

---

2. Two-Platform Audit (Google + Meta)

User Request:

"Audit my Google and Meta accounts. $30K/month on Google, $20K on Meta. SaaS company, goal is demo requests."

Analysis Steps: 1. Context: SaaS B2B, $50K/month, 60% Google / 40% Meta, lead gen goal 2. Spawn 2 parallel audit tasks (Google + Meta only) 3. Merge with budget-weighted aggregate 4. Cross-platform analysis focused on lead gen attribution

[Output follows same format as Example 1, with 2 platforms instead of 3]

---

3. Audit with Deck Generation

User Request:

"Run a full paid media audit and build me a deck I can present to my client. They're on Google and Meta, spending $60K/month total on DTC."

Analysis Steps: 1. Run standard cross-platform audit (Example 1 format) 2. After unified report is generated, invoke pro-deck-builder skill 3. Generate branded HTML slide deck with:

• Title slide with aggregate score
• Executive summary (aggregate score, top 3 findings, estimated waste)
• Per-platform score cards (1 slide each)
• Cross-platform analysis (budget, tracking, creative, attribution)
• Quick Wins slide
• Kill List + Scaling Opportunities
• Prioritized 90-day action plan
• Expected outcomes with projected metrics

Deck invocation:

"Pipe this audit to pro-deck-builder. Use dark cover slide, warm cream content slides. Include the aggregate score prominently on the title slide."

[Audit output feeds directly into pro-deck-builder, which generates the HTML deck]

Cross-Platform Audit Reference

Budget Allocation Benchmarks

Platform Mix by Business Type

Minimum Viable Spend by Platform

If a platform is below minimum viable spend, recommend consolidating to fewer platforms rather than spreading thin.

Attribution and Tracking Consistency

Recommended Attribution Setup

Attribution Overlap Indicators

UTM Requirements for Cross-Platform Deduplication

All platforms must use consistent UTM parameters for GA4 to arbitrate attribution:

utm_source: google | meta | microsoft
utm_medium: cpc | paid-social | cpc
utm_campaign: {campaign_name}
utm_content: {ad_id} or {creative_name}
utm_term: {keyword} (search only)

Cross-Platform Performance Benchmarks

CPA Comparison by Vertical (2025-2026)

ROAS Comparison by Vertical

Budget Reallocation Framework

When to Shift Budget Between Platforms

Reallocation Rules

1. Never reallocate more than 25% of a platform's budget in one week — dramatic shifts reset learning phases 2. Reallocate from waste first — use Kill List spending before cutting working campaigns 3. Account for platform minimums — don't reduce a platform below its minimum viable spend unless fully exiting 4. Monitor for 14 days after reallocation before making further changes 5. Keep a testing budget — 10-15% of total should fund platform/audience/creative experiments

Cross-Platform Creative Alignment

Messaging Consistency Checks

Format Requirements by Platform

Kill List Criteria

Universal Thresholds (Apply Across All Platforms)

Waste Aggregation

Total monthly waste across all platforms:

Total Waste = Σ(monthly spend on campaigns flagged as Kill List)
Waste % = Total Waste / Total Ad Spend × 100