agentleFS
narayan-metaflownarayan-metaflow/metaflow-marketing-skills/skills/google-ads-account-audit/SKILL.md

google-ads-account-audit

Skill14 starsChanged 7 months agoMITView on GitHub

Systematic 70+ checkpoint audit covering tracking, structure, keywords, ads, bidding, budgets, audiences, and competitive positioning — with ICE-prioritized recommendations.

4 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 4 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Don't audit without search term data: Search term reports are the single richest source of waste identification.

    line 251
  2. Distinctive:

    Flag AI Max opportunity: If the account has 50+ monthly conversions and isn't testing AI Max for Search (global beta September 2025, avg …

    line 256
  3. Distinctive:

    Account for learning phases: Don't flag bid strategy performance during the 7-14 day learning period after changes.

    line 257
  4. Distinctive:

    Confidence: If audit data is incomplete (missing search terms, no QS data, limited date range), state what conclusions are reliable and …

    line 260

“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Process

  1. Collect data — Request or gather: campaign data export, search term report (90 days), Quality Score report, auction insights, change history (90 days), conversion tracking status, and account-level settings
  2. Run the audit checklist — Work through each section below, flagging issues as Critical (red), Optimization (yellow), or Working Well (green)
  3. Score account health — Calculate overall health score based on findings
  4. ICE-prioritize recommendations — Score each finding: Impact (1-10) x Confidence (1-10) x Ease (1-10)
  5. Deliver action plan — Structured output with prioritized fixes and expected impact

Audit Checklist

Section 1: Tracking & Measurement

CheckpointWhat to CheckFlag If
Google Tag activegtag.js or GTM container firing on all pagesMissing or intermittent
GA4 linkedGA4 property connected to Google AdsNot linked
Conversion actionsList all conversion actions — are they meaningful?Counting micro-conversions as primary
Enhanced conversions (web)Hashing email/phone at tag level for better matchingNot implemented
Enhanced conversions (leads)GCLID-based offline import for lead gen accountsNot implemented for lead gen
Attribution modelData-driven attribution is default and recommendedStill on last-click without reason
Conversion windowsWindows match actual sales cycleDefault windows for long sales cycle
Cross-account conversion trackingUsing manager-level conversion tracking if MCCDuplicate conversion counting across accounts
Offline conversion importUploading CRM data back to Google AdsLead gen account with no offline import
Meta trackingIs the client also running Meta? Check if CAPI is set upMeta tracking gaps — refer to meta-ads-tracking-setup skill

Section 2: Account Structure

CheckpointWhat to CheckFlag If
Campaign-to-objective alignmentEach campaign has a clear strategic purposeCampaigns mixing objectives
Campaign count6-10 campaigns with clear distinctions for complex accountsOver-segmented (20+ campaigns diluting budget) or under-segmented (1-2 campaigns for diverse products)
Search + Display mixingSearch and Display should be separate campaignsMixed in same campaign
PMax asset groupsSegmented by product/service line with distinct audiencesSingle asset group for diverse products
AI Max setupText guidelines configured, Smart Bidding Exploration monitoredAI Max running without text guidelines (launched September 2025, avg 34% conversion improvement)
Demand Gen campaignsPresent if awareness/consideration is a goalMissing from full-funnel accounts (Power Pack strategy: Demand Gen + AI Max + PMax)
Naming conventionsConsistent, parseable naming taxonomyInconsistent or absent naming system
Ad group themingTight keyword-to-ad-group mapping by intentBloated ad groups mixing unrelated keywords

Section 3: Keywords & Match Types

CheckpointWhat to CheckFlag If
Match type distributionBalance of broad, phrase, exact appropriate for account maturity100% broad match without Smart Bidding, or 100% exact match limiting reach
Broad match readinessBroad match requires Smart Bidding + 50+ conversions/monthBroad match with manual bidding or <50 conversions/month
Search term wasteN-gram analysis on search term reportN-grams with $50+ spend and 0 conversions
Negative keyword coverageThemed negative lists: competitors, jobs, free/DIY, irrelevant industriesNo shared negative lists, or lists not updated in 90+ days
Keyword conflictsKeywords in different campaigns competing against each otherSame keyword in multiple campaigns without clear priority
PMax search term cannibalizationPMax stealing queries that Search campaigns should serveHigh-value queries showing in PMax search term insights instead of Search
Low search volume keywordsKeywords flagged as low search volume>20% of keywords with low search volume status
Fractional conversion distortionData-driven attribution splits conversions into fractionsEvaluating keyword CPA on conversions < 1 — filter to conversions > 1 before judging CPA

N-gram analysis quick method: Break search queries into 1-grams, 2-grams, 3-grams. Aggregate clicks, spend, conversions, and CPA by n-gram. Action framework:

  • N-gram with $50+ spend, 0 conversions → Add as negative keyword
  • N-gram with strong CPA → Promote to exact-match keyword in dedicated ad group
  • N-gram with high impressions, low CTR → Investigate intent mismatch

For automated n-gram analysis, use the google-ads-scripts skill (Brainlabs Search Query Mining Script updated for GAQL by Nils Rooijmans, March 2025).

Section 4: Ads & Creative

CheckpointWhat to CheckFlag If
RSA completeness15 headlines + 4 descriptions per RSAUnder-filled RSAs (<10 headlines)
RSA asset performanceAsset-level reporting (2025 feature)Multiple "Low" rated headlines/descriptions not replaced
Ad strengthGoogle's ad strength indicator"Poor" or "Average" — target "Good" or "Excellent"
Headline diversityHeadlines cover different angles (features, benefits, urgency, social proof)All headlines say the same thing with minor variations
Pin strategyHigh-intent or brand headlines pinned to position 1No pins and key messages buried, or over-pinning limiting combinations
Extensions/assetsSitelinks (4-6), callouts, structured snippets, images, price, callMissing extension types — each adds CTR
Landing page relevanceAd message matches landing page headline and CTAMessage mismatch between ad and landing page
Mobile optimizationLanding pages load <3 seconds, mobile-friendlySlow load or poor mobile experience

Section 5: Bidding & Budget

CheckpointWhat to CheckFlag If
Bid strategy appropriatenessStrategy matches campaign maturity and conversion volumeSee Bid Strategy Maturity table below
ECPC checkECPC deprecated March 31, 2025 — campaigns not migrated now run on Manual CPCAny campaign still showing ECPC status
tCPA/tROAS targetsTargets realistic vs actual performanceTarget CPA set 50%+ below actual CPA (algorithm can't deliver)
Smart Bidding readinessMinimum conversion volume for strategy tiertCPA with <30 conversions/month, tROAS with <50 conversions/month
Portfolio bid strategiesShared strategies across related campaignsRelated campaigns on individual strategies that could benefit from pooling (typical lift: 19-27% ROAS improvement)
Budget sufficiencyDaily budget adequate for bid strategyPMax budget < 3x target CPA daily, or < $50-100/day
Budget cappingCampaigns limited by budgetImpression share lost to budget > 20% on profitable campaigns
AI Max learning phase7-14 days for new AI Max campaignsChanges made during learning phase (September 2025 global beta)
Impression share analysisSearch impression share vs targets<30% = likely underbidding; >95% with low ROI = likely overspending
Impression share lost (rank)Lost to ad rank specificallyHigh % lost to rank = Quality Score or bid issue
Impression share lost (budget)Lost to budget specificallyHigh % lost to budget = increase budget or reduce targeting

Bid Strategy Maturity Table:

Monthly ConversionsRecommended StrategyNotes
0-15Manual CPC or Maximize Clicks (with bid cap)Not enough data for automation
15-30Maximize ConversionsBuilding data, algorithm learning
30-50Maximize Conversions or tCPAEnough for tCPA if CPA is stable
50+tCPA or tROASFull Smart Bidding — tROAS needs 50+, ideally 100+
50+ with broad matchBroad match + tCPA/tROASRecommended 2025 default for mature accounts

Section 6: Quality Score Deep Dive

Quality Score = expected CTR + ad relevance + landing page experience. Analyze at component level, not just the composite number.

ComponentWhat Drives ItFix If Below Average
Expected CTRHistorical CTR vs competitors for same queryImprove ad copy relevance, test new headlines, add negative keywords to filter irrelevant impressions
Ad RelevanceHow closely ad matches search intentTighten ad group themes, write intent-specific headlines, use keyword insertion where appropriate
Landing Page ExperiencePage relevance, speed, mobile-friendliness, navigation easeImprove message match, speed (<3 sec), mobile UX, add relevant content

Action priority: Landing page experience has the widest impact (affects all keywords pointing to that page). Fix landing pages first, then ad relevance, then expected CTR.

Section 7: Audiences

CheckpointWhat to CheckFlag If
Remarketing listsActive, properly sized, segmented by recencyNo remarketing lists, or lists expired
Customer matchCRM lists uploaded and refreshedNo customer match lists for accounts with CRM data
Audience segmentsIn-market, affinity, custom segments appliedNo audience signals on PMax/Demand Gen campaigns
Observation vs targetingAudiences in correct mode per campaign goalObservation mode where targeting is intended (or vice versa)
Audience performanceSegment-level performance reviewSegments with high spend and no conversions not excluded

Section 8: Competitive & Change History

CheckpointWhat to CheckFlag If
Auction Insights trendsImpression share, overlap rate, position above rate over 90 daysNew competitor gaining significant share
Change history correlationLast 90 days of account changes vs performance shiftsPerformance drop aligns with a specific change (bid strategy switch, new campaign, budget cut)
Competitor ad copyGoogle Ads Transparency Center reviewCompetitors testing offers or messaging you're not responding to

Health Score Calculation

Score each section 0-10 based on findings:

SectionWeightScore
Tracking & Measurement20%/10
Account Structure10%/10
Keywords & Match Types20%/10
Ads & Creative10%/10
Bidding & Budget20%/10
Quality Score10%/10
Audiences5%/10
Competitive5%/10
Weighted Total/10

Interpretation: 8-10 = Healthy, 6-7.9 = Needs optimization, 4-5.9 = Significant issues, <4 = Critical — major restructuring needed.

Output Format

## Google Ads Account Audit

### Health Score: [X.X / 10] — [Healthy / Needs Optimization / Significant Issues / Critical]

### Critical Issues (fix immediately)
1. **[Issue]** — [What's wrong, specific data] — ICE: [I×C×E = score]
   - Action: [Specific fix]
   - Expected impact: [Estimated improvement]
2. ...

### Optimization Opportunities (fix within 2 weeks)
1. **[Issue]** — [What's wrong, specific data] — ICE: [score]
   - Action: [Specific fix]
   - Expected impact: [Estimated improvement]
2. ...

### Working Well (maintain)
- [What's working and why]
- ...

### Priority Action Plan (next 30 days)
| Priority | Action | ICE Score | Timeline | Owner |
|----------|--------|-----------|----------|-------|
| 1 | [Highest ICE action] | [score] | Week 1 | [who] |
| 2 | ... | ... | Week 1-2 | ... |
| 3 | ... | ... | Week 2-4 | ... |

### 2025 Platform Change Flags
- [ ] ECPC migration status (deprecated March 2025)
- [ ] AI Max for Search adoption opportunity (September 2025)
- [ ] PMax negative keywords enabled (January 2025)
- [ ] PMax search term reporting reviewed
- [ ] Enhanced conversions implemented
- [ ] Power Pack strategy evaluated (Demand Gen + AI Max + PMax)

Example

Input: "Here's my Google Ads data for last 30 days. I'm spending $15K/month across 4 Search campaigns and 1 PMax campaign. CPA went from $45 to $72 over the past 6 weeks. Help me figure out what's wrong."

Output:

Google Ads Account Audit

Health Score: 5.2 / 10 — Significant Issues

Critical Issues

  1. ECPC still active on 2 Search campaigns — ECPC was deprecated March 31, 2025. These campaigns reverted to Manual CPC, losing Smart Bidding optimization. — ICE: 9×9×9 = 729
    • Action: Migrate to Maximize Conversions (both campaigns have 35+ conversions/month — sufficient for tCPA)
    • Expected impact: 15-25% CPA reduction once algorithm re-learns (7-14 day learning phase)
  2. Search term waste: $2,340/month on zero-conversion n-grams — N-gram analysis shows "jobs," "salary," and "careers" 2-grams account for $1,800 in wasted spend. "Free" modifier adds another $540. — ICE: 9×10×10 = 900
    • Action: Add negative keyword lists: jobs/careers, free/DIY. Review remaining zero-conversion n-grams with $50+ spend
    • Expected impact: ~$2,340/month savings = 15.6% of total spend recovered
  3. PMax campaign budget at $25/day with $72 CPA — Below minimum threshold (daily budget should be 3x target CPA = $135+/day minimum). Algorithm can't optimize with insufficient budget. — ICE: 8×8×7 = 448
    • Action: Increase PMax daily budget to $150/day or pause PMax and consolidate into Search campaigns
    • Expected impact: PMax either performs as designed or stops wasting budget at sub-scale

Optimization Opportunities

  1. No enhanced conversions — Missing 10-15% of conversion attribution. — ICE: 7×8×8 = 448
    • Action: Implement enhanced conversions for web (tag-level email/phone hashing)
    • Expected impact: Better attribution → better Smart Bidding decisions → lower CPA
  2. Quality Score: 4 ad groups with avg QS < 5 — Landing page experience "Below Average" on 3 groups, ad relevance "Below Average" on 1. — ICE: 7×7×5 = 245
    • Action: Audit landing pages for message match and page speed. Rewrite ads in low-relevance group
    • Expected impact: QS improvement from 4→6 typically reduces CPC 15-25%
  3. No audience signals on PMax — PMax running without remarketing lists or customer match. — ICE: 6×7×8 = 336
    • Action: Add remarketing audiences and customer match list as audience signals
    • Expected impact: Faster PMax learning, better targeting accuracy

Working Well

  • Campaign naming conventions are consistent and parseable
  • Sitelink and callout extensions present on all Search campaigns
  • GA4 properly linked with data-driven attribution active

Priority Action Plan (next 30 days)

PriorityActionICE ScoreTimelineOwner
1Add negative keyword lists (jobs, free/DIY)900Day 1-2Media buyer
2Migrate ECPC campaigns to Maximize Conversions729Day 1-3Media buyer
3Implement enhanced conversions448Week 1-2Developer + media buyer
4Fix PMax budget or consolidate448Week 1Media buyer
5Add audience signals to PMax336Week 2Media buyer
6Landing page QS improvements245Week 2-4Dev + media buyer

2025 Platform Change Flags

  • ECPC migration status — NOT MIGRATED on 2 campaigns (Critical)
  • AI Max for Search — not yet adopted, evaluate after fixing current issues
  • PMax negative keywords — enabled but no negatives added
  • PMax search term reporting — not yet reviewed (do this week)
  • Enhanced conversions — not implemented
  • Power Pack strategy — evaluate after stabilizing current campaigns

Guidelines

  • Don't audit without search term data. Search term reports are the single richest source of waste identification. If the user only provides campaign-level data, request the search term report before completing the audit.
  • Don't recommend tROAS with <50 conversions/month. The algorithm needs volume. Recommending it prematurely leads to delivery collapse.
  • Don't panic over PMax opacity. PMax now has campaign-level negatives, full search term reporting, and channel performance reports (all 2025). Use them before concluding PMax is wasteful.
  • Don't ignore change history. A CPA spike that coincides with a bid strategy change or budget cut has a different fix than one caused by competitive pressure or seasonality.
  • Flag ECPC immediately. Any campaign still showing ECPC was deprecated March 31, 2025. This is always a Critical finding.
  • Flag AI Max opportunity. If the account has 50+ monthly conversions and isn't testing AI Max for Search (global beta September 2025, avg 34% conversion improvement), note it as an optimization opportunity.
  • Account for learning phases. Don't flag bid strategy performance during the 7-14 day learning period after changes.
  • Fractional conversion trap. Data-driven attribution creates fractional conversions. A keyword showing 0.3 conversions isn't necessarily bad — it contributed to a conversion. Filter to conversions > 1 before evaluating keyword-level CPA.
  • Cross-references: For Meta tracking issues, use the meta-ads-tracking-setup skill. For Meta account auditing, use the meta-ads-audit skill. For automated audit scripts, use the google-ads-scripts skill.
  • Confidence: If audit data is incomplete (missing search terms, no QS data, limited date range), state what conclusions are reliable and what needs more data. Never fabricate findings to fill gaps.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.