gtm-container-auditor

GTM Container Auditor

Purpose

Help agencies and advertisers assess the health and quality of their Google Tag Manager containers. Poor GTM hygiene causes duplicate conversions, inflated metrics, slow page load (from excessive tags), and tracking gaps. A clean container typically improves data accuracy by 15-40%.

When to Use This Skill

Invoke when user mentions:

• Health check: "Audit my GTM container" / "Review my tag setup"
• Duplicate tracking: "My conversions look doubled" / "Are there duplicate tags?"
• Performance: "GTM is slowing my site" / "How many tags do I have?"
• Governance: "We have a messy GTM" / "Can you find unused tags?"
• Migration: "We switched platforms" / "Cleanup after agency handover"
• Compliance: "GDPR audit" / "Which tags fire before consent?"

Required Tools

Tool	Purpose
gtm_list_containers	List all containers across accounts
gtm_audit	Deep audit of a specific container (tags, triggers, variables)

---

Part 1: Container Discovery

Step 1: List Available Containers

gtm_list_containers()

Present a summary table:

Account	Container Name	Public ID	Type	Workspaces
[account]	[name]	GTM-XXXXX	web	[N]

If multiple containers exist, ask the user which to audit, or audit all if requested.

---

Part 2: Deep Container Audit

Step 2: Run Full Audit

gtm_audit(container_path="accounts/[ID]/containers/[ID]")

The audit returns: tags, triggers, variables, workspace summary, built-in variables.

---

Part 3: Health Scoring Framework

Score each area 0-100, then compute overall health:

Tag Health (30% of score)

Issue	Severity	Points Deducted
Tag with no trigger	Critical	-15 per tag
Duplicate tag name	High	-10 per duplicate
Paused tag (> 90 days)	Medium	-5 per tag
Tag firing on All Pages without filter	Medium	-5 per tag
Deprecated tag type	High	-10 per tag
More than 80 tags total	Low	-5

Trigger Health (25% of score)

Issue	Severity	Points Deducted
Unused trigger (no tags)	High	-10 per trigger
All Pages trigger on conversion tags	Critical	-20
Duplicate trigger names	Medium	-5 per duplicate
Regex trigger without testing	Medium	-5 per trigger

Variable Health (20% of score)

Issue	Severity	Points Deducted
Undefined variable referenced in tag	Critical	-20 per var
Unused variable (defined but not used)	Low	-3 per variable
Data Layer variable with typos	High	-10 per var
More than 50 custom variables	Low	-5

Workspace Health (15% of score)

Issue	Severity	Points Deducted
Unpublished changes > 30 days old	High	-15
Multiple workspaces (> 3) active	Medium	-10
No version history visible	Medium	-10

Governance (10% of score)

Issue	Severity	Points Deducted
Tags without notes/descriptions	Low	-3 per 10 tags
Inconsistent naming convention	Medium	-10
No folder organization	Low	-5

Health Score Bands

Score	Rating	Action
90-100	Excellent	Monitor only
75-89	Good	Minor cleanup recommended
60-74	Fair	Scheduled cleanup within 30 days
40-59	Poor	Immediate cleanup required
< 40	Critical	Full audit and remediation urgently needed

---

Part 4: Common Issue Patterns

Duplicate Conversion Tracking

Detected when: Multiple tags firing the same conversion event (e.g., two Google Ads conversion tags for "purchase" or identical GA4 purchase events).

Impact: Doubles reported conversions, distorts ROAS, inflates Google Ads bidding.

Fix: Consolidate to one tag per conversion event. Use GA4 linked conversions instead of separate Google Ads conversion pixel where possible.

All-Pages Firing Tags

Detected when: A tag (especially conversion or remarketing tags) has "All Pages" as its only trigger.

Impact: Fires conversion on every page visit → massive false conversion count.

Fix: Replace with specific page/event trigger (e.g., Thank You Page URL trigger, form_submit event trigger).

Orphaned Tags

Detected when: Tags exist with no trigger assigned (paused indefinitely or accidentally created).

Impact: Clutter, confusion during future edits, potential accidental activation.

Fix: Delete confirmed-unused tags or add a descriptive note + archive label.

Data Layer Race Conditions

Detected when: Variables reference dataLayer keys (e.g., ecommerce.purchase.revenue) but no dataLayer push is confirmed in the trigger configuration.

Impact: Variables evaluate as undefined → broken conversions, empty event params.

Fix: Ensure dataLayer.push() fires BEFORE GTM fires the tag. Use "DOM Ready" or custom event triggers instead of page load triggers for dynamically loaded data.

Consent Violation Risk

Detected when: Marketing/analytics tags (Google Ads, Meta Pixel, LinkedIn Insight) fire without a consent check trigger or Consent Mode configuration.

Impact: GDPR non-compliance. Potential fines up to 4% of annual revenue.

Fix: Implement Consent Mode v2. Add "consent granted" trigger to all marketing tags. Ensure tags default to denied before consent.

Consent Mode v2 is mandatory for EEA advertisers since March 2024. Consent Mode v2 adds two new required parameters beyond the original v1:

• ad_user_data: Controls whether user data can be sent to Google for advertising
• ad_personalization: Controls whether ads can be personalized for the user

Both must be explicitly set. Without them, Google may restrict access to audience features and Smart Bidding in the EEA. Check that your CMP (Cookiebot, OneTrust, Usercentrics, etc.) is configured to pass all four parameters: ad_storage, analytics_storage, ad_user_data, ad_personalization.

---

Part 5: Tag Type Classification

Classify all tags found:

Category	Tag Types	Risk Level
Analytics	GA4, UA, Adobe Analytics	Low
Advertising	Google Ads, Meta Pixel, LinkedIn Insight, TikTok Pixel	High (consent)
Remarketing	Dynamic remarketing tags, Custom audience pixels	High (consent)
Conversion	Google Ads conversion, GA4 conversion events	Critical (accuracy)
Utility	Custom HTML, Custom Image, Hotjar, Clarity	Medium
Deprecated	UA/GA3 tags, old conversion linkers	High (remove)

Flag any deprecated tag types immediately — Universal Analytics ended July 2024.

---

Part 6: Output Format

================================================================================
               GTM CONTAINER AUDIT REPORT
               Container: [GTM-XXXXX] — [Container Name]
               Account: [Account Name]
               Audit Date: [YYYY-MM-DD]
================================================================================

OVERALL HEALTH SCORE: [N]/100 — [Rating]

SUMMARY
───────────────────────────────────
Total Tags:       [N] ([N] active, [N] paused)
Total Triggers:   [N] ([N] unused)
Total Variables:  [N] custom + [N] built-in
Workspaces:       [N] ([N] with unpublished changes)
Container Type:   [web/ios/android/amp]

CATEGORY SCORES
───────────────────────────────────
Tag Health:        [N]/100
Trigger Health:    [N]/100
Variable Health:   [N]/100
Workspace Health:  [N]/100
Governance:        [N]/100

═══════════════════════════════════
CRITICAL ISSUES (Fix Immediately)
═══════════════════════════════════
[List issues with Severity: Critical]

Example:
⛔ 2 conversion tags fire on ALL PAGES — false conversions guaranteed
   Tags: "GA4 - Purchase", "Google Ads - Conversion"
   Fix: Replace trigger with "Thank You Page" URL contains /thank-you

⛔ Google Ads conversion tag has no trigger
   Tag: "Ads Conversion - Lead"
   Fix: Assign appropriate trigger or delete if unused

═══════════════════════════════════
HIGH PRIORITY (Fix Within 7 Days)
═══════════════════════════════════
[List high-severity issues]

⚠️ 4 orphaned triggers (no tags using them)
   Triggers: [list names]
   Fix: Delete unused triggers to reduce confusion

⚠️ Universal Analytics tag detected — GA3 deprecated since July 2024
   Tag: "UA - Pageview"
   Fix: Remove. Ensure GA4 replaces all UA functionality.

═══════════════════════════════════
MEDIUM PRIORITY (Fix Within 30 Days)
═══════════════════════════════════
[List medium-severity issues]

═══════════════════════════════════
TAG INVENTORY
═══════════════════════════════════
| Tag Name | Type | Trigger | Status | Issue |
|----------|------|---------|--------|-------|
| GA4 Config | GA4 | All Pages | Active | ✅ OK |
| Ads - Purchase | Google Ads | Thank You | Active | ✅ OK |
| Old UA Tag | UA | All Pages | Paused | ⚠️ Deprecated |

═══════════════════════════════════
CONSENT COMPLIANCE CHECK (Consent Mode v2)
═══════════════════════════════════
Marketing tags found: [N]
Tags with consent trigger: [N]
Tags WITHOUT consent check: [N] ← [Compliant / RISK]

Consent Mode v2 status:
  ad_storage configured:        [Yes / No / Unknown]
  analytics_storage configured: [Yes / No / Unknown]
  ad_user_data configured:      [Yes / No / Unknown]  ← v2 required
  ad_personalization configured:[Yes / No / Unknown]  ← v2 required

[List non-compliant tags if any]

Note: All four Consent Mode v2 parameters are mandatory for EEA advertisers
(required by Google since March 2024). Missing ad_user_data or ad_personalization
will restrict Smart Bidding and audience features in Google Ads.

═══════════════════════════════════
REMEDIATION ROADMAP
═══════════════════════════════════
Week 1 — Critical fixes:
1. [Specific action]
2. [Specific action]

Week 2 — High priority:
3. [Specific action]

Month 1 — Medium priority:
4. [Specific action]

ESTIMATED EFFORT: [X] hours
EXPECTED DATA ACCURACY IMPROVEMENT: [X]%

---

Part 6b: Server-Side Tagging Assessment

Server-side GTM (sGTM) is now mainstream and directly relevant to container audits. When auditing a web container, assess whether server-side tagging should be recommended.

Signs Server-Side Tagging Would Help

Observation	Recommendation
15+ tags on All Pages	Move heavy tags (Meta Pixel, GA4) to server-side
Ad blockers impacting conversion data	Server-side bypasses most ad blockers
ITP/cookie restrictions on Safari	Server-side extends cookie lifetime
CAPI or Enhanced Conversions not implemented	sGTM is the easiest path to server-side
Page load > 3s with GTM contributing	Offload tag execution server-side

sGTM vs Web GTM Responsibilities

Function	Web GTM	Server GTM
Data layer push	Yes	No
Event collection (GA4)	Can be replaced	Preferred
Ad pixels (Meta, TikTok, LinkedIn)	High risk (consent/blocking)	Better — consent-aware server call
Conversion tracking (Google Ads)	Fine	Also fine via Enhanced Conversions
Custom HTML tags	Full control	Limited (server-side only)

When a web container has many marketing tags AND no server container is in use, flag this in the audit with a recommendation to consider sGTM as a Phase 2 improvement.

---

Part 7: Industry Benchmarks

Healthy Container Metrics

Metric	Healthy Range	Warning	Critical
Total tags	< 50	50-100	> 100
Unused triggers	< 5%	5-15%	> 15%
Tags without notes	< 20%	20-50%	> 50%
Paused tags	< 10%	10-25%	> 25%
Active workspaces	1-2	3-4	> 4

GTM Performance Impact

Each additional tag on All Pages adds ~5-15ms to page load time:

• 20 tags: ~100-300ms overhead
• 50 tags: ~250-750ms overhead
• 100 tags: ~500ms-1.5s overhead (significant SEO and UX impact)

Recommend: Audit tags that fire on All Pages. Can most fire on specific events instead?

Optional: Enrich with Live Data

If the user has connected their GTM account, run a live container audit instead of relying on manual review:

# Pull a full audit of the container — tags, triggers, variables, and issues
gtm_audit(container_path="accounts/YOUR_ACCOUNT_ID/containers/YOUR_CONTAINER_ID")

This returns tag counts, firing rules, variable usage, and flags common issues (missing notes, paused tags, All Pages over-triggers). Use the audit output to populate the framework above with real data rather than estimates.