Analytics Tracking

Analytics Tracking - Implementation & Auditing

Category: Marketing Tags: GA4, Google Tag Manager, event tracking, conversion tracking, UTM, analytics audit, consent mode

Overview

Analytics Tracking is the implementation layer for marketing measurement. Bad tracking is worse than no tracking -- duplicate events, missing parameters, unconsented data, and broken conversions lead to decisions based on bad data. This skill covers building tracking right the first time and finding what is broken when it is not.

This skill handles implementation only. For analyzing campaign performance data, use campaign-analytics. For product analytics and in-app behavior, use the product-team skills.

Clarify First

Before building the tracking plan, confirm these inputs. If any is unknown or vague, ASK — do not assume:

• Operating mode — build from scratch, audit existing, or debug a specific issue (selects the entire workflow)
• Platform stack — GA4, GTM, and which ad platforms (Google Ads / Meta / LinkedIn) (drives implementation + conversion tracking setup)
• Key conversions + funnel events — the user actions that matter to the business (drives event taxonomy + conversion configuration)
• Region / consent requirements — EU/EEA users present? (determines whether Consent Mode v2 is required)

Stop rule: ask only the 2-3 that most change the output. If the user says "just draft it," proceed and list your assumptions at the top of the artifact.

Operating Modes

Mode 1: Build From Scratch

No analytics in place. Build the tracking plan, implement GA4 + GTM, define event taxonomy, configure conversions.

Mode 2: Audit Existing Tracking

Tracking exists but data cannot be trusted. Audit coverage, identify gaps, clean up duplicates, fix consent issues.

Mode 3: Debug Specific Issues

Events are missing, conversions do not match, GTM preview shows fires but GA4 does not record. Structured debugging workflow.

Event Taxonomy Design

Get this right before touching GA4 or GTM. Retrofitting taxonomy is painful and expensive.

Naming Convention

Format: object_action (snake_case, past tense verb)

Rules:

1. Always noun_verb order, never verb_noun
2. Snake_case only -- no camelCase, no hyphens, no PascalCase
3. Past tense verbs: _started, _completed, _failed, _viewed
4. Specific enough to be unambiguous, not so verbose it is a sentence
5. Prefix with domain when needed: onboarding_step_completed, billing_plan_selected

Standard Event Parameters

Every custom event should include applicable parameters from this table:

SaaS Event Taxonomy (Reference)

Core Funnel:

visitor_arrived              (automatic page_view in GA4)
signup_started               (user clicked "Sign up")
signup_completed             (account created)
trial_started                (free trial began)
onboarding_step_completed    (params: step_name, step_number)
feature_activated            (params: feature_name)
plan_selected                (params: plan_name, billing_period)
checkout_started             (params: value, currency, plan_name)
checkout_completed           (params: value, currency, transaction_id)
subscription_renewed         (params: value, plan_name)
subscription_cancelled       (params: cancel_reason, plan_name)

Micro-Conversions:

pricing_viewed
demo_requested               (params: source)
form_submitted               (params: form_name, form_location)
content_downloaded           (params: content_name, content_type)
video_started                (params: video_title)
video_completed              (params: video_title, percent_watched)
chat_opened
help_article_viewed          (params: article_name)
invite_sent                  (params: recipient_role)
integration_connected        (params: integration_name)

GA4 Configuration

Data Stream Setup

1. Create property: GA4 Admin > Properties > Create
2. Add web data stream with your domain
3. Enhanced Measurement -- review each:
   • Page views: Keep enabled
   • Scrolls: Keep enabled
   • Outbound clicks: Keep enabled
   • Site search: Enable if you have search
   • Video engagement: Disable if tracking videos manually (avoids duplicates)
   • File downloads: Disable if tracking via GTM (for better parameters)
4. Configure domains: add all subdomains in your funnel
5. Data retention: Set to 14 months (maximum for free GA4)

Key Events (formerly "Conversions")

GA4 renamed "conversions" to key events in March 2024; "conversions" now refers to key events imported into Google Ads. Mark as key events in GA4 Admin > Data display > Events (star icon):

• signup_completed
• checkout_completed
• demo_requested
• trial_started

Rules:

• Maximum 30 key events per standard property (50 on Analytics 360) -- curate carefully
• Marking a key event is not retroactive -- it affects reports from the time it is marked, not historic data
• Do not mark micro-conversions as key events unless optimizing ad campaigns for them
• Counting method: set to "once per session" for lead events, "every" for purchase events

Source: Mark events as key events (as of September 2026).

Custom Dimensions

Register custom dimensions for any event parameter you want to filter/segment by:

Register in GA4 Admin > Custom definitions > Create custom dimension.

Google Tag Manager Implementation

Container Architecture

GTM Container
├── Tags
│   ├── GA4 Configuration (All Pages trigger)
│   ├── GA4 Event Tags (one per custom event)
│   ├── Google Ads Conversion Tags (per conversion action)
│   └── Meta Pixel / LinkedIn Insight (if running ads)
├── Triggers
│   ├── All Pages (Page View)
│   ├── DOM Ready
│   ├── Custom Event triggers (one per dataLayer event)
│   └── Element Click triggers (CSS selector based)
└── Variables
    ├── Data Layer Variables (one per dataLayer key)
    ├── Constants (GA4 Measurement ID, etc.)
    └── Lookup Tables (if needed for mapping)

Implementation Pattern: Data Layer Push

Your application pushes events to the data layer. GTM picks them up and sends to GA4.

Application code:

// Push event when user completes signup
window.dataLayer = window.dataLayer || [];
window.dataLayer.push({
  event: 'signup_completed',
  method: 'email',
  user_id: userId,
  plan_name: 'trial'
});

GTM configuration:

Trigger:
  Type: Custom Event
  Event name: signup_completed

Tag:
  Type: GA4 Event
  Event name: signup_completed
  Parameters:
    method:    {{DLV - method}}
    user_id:   {{DLV - user_id}}
    plan_name: {{DLV - plan_name}}

SPA Handling

Single Page Applications need special attention because page views do not fire automatically on route changes.

Option A: History change trigger (GTM built-in)

• Enable "History Change" trigger in GTM
• Fires GA4 page_view on every pushState/popState

Option B: DataLayer push on route change (more control)

// In your router (React Router, Next.js, etc.)
router.events.on('routeChangeComplete', (url) => {
  window.dataLayer.push({
    event: 'page_view',
    page_location: url,
    page_title: document.title
  });
});

Conversion Tracking: Ad Platforms

Google Ads

Recommended approach: Import GA4 conversions into Google Ads (single source of truth).

1. Link GA4 and Google Ads accounts
2. In Google Ads > Goals > Conversions > Import > Google Analytics
3. Select GA4 key events to import (they become Google Ads conversions)
4. Set attribution model: Data-driven (if 50+ conversions/month), otherwise Last-click
5. Conversion window: 30 days for lead gen, 90 days for high-consideration B2B

Enhanced Conversions: Enable for 15-30% better conversion measurement. Sends hashed first-party data (email, phone) to match conversions that cookies miss.

Meta (Facebook/Instagram)

1. Install Meta Pixel base code via GTM
2. Configure standard events: PageView, Lead, CompleteRegistration, Purchase
3. Conversions API (CAPI): strongly recommended -- client-side pixel loses approximately 30% of conversions due to ad blockers and iOS App Tracking Transparency
4. Deduplication: when using both pixel and CAPI, send the same event_id to prevent double-counting

LinkedIn Insight Tag

1. Install via GTM (Tag type: LinkedIn Insight)
2. Configure conversion events in LinkedIn Campaign Manager
3. Match events to your taxonomy: signup_completed -> LinkedIn "Sign-up" conversion

UTM Strategy

Convention Enforcement

Critical rules:

• Never tag organic traffic with UTMs (overrides GA4 automatic attribution)
• Never tag direct/internal links with UTMs
• Use a UTM builder spreadsheet or tool -- manual entry causes inconsistency
• Lowercase everything -- Google and google are different sources in GA4

Attribution Windows

AI Referral Traffic

Visits from AI assistants (ChatGPT, Gemini, Claude, Copilot, Perplexity, etc.) are a growing, distinct acquisition source. Measure them explicitly instead of letting them blur into Referral or Direct.

GA4 "AI Assistant" Channel (as of September 2026)

Since May 13, 2026, GA4's Default Channel Group includes an AI Assistant channel. When the referrer matches Google's list of recognized AI assistants, GA4 sets medium = ai-assistant and campaign = (ai-assistant), and the session is grouped under AI Assistant. No configuration is needed. The rule is evaluated before Referral. Sources: Default channel group, What's new in Google Analytics.

Known Gaps

Fallback: Custom Channel Group

Create a custom channel group in GA4 (Admin > Data display > Channel groups), add an "AI Assistant (custom)" channel above Referral, with the condition Session source matches regex:

^(chatgpt\.com|chat\.openai\.com|openai\.com|perplexity\.ai|www\.perplexity\.ai|gemini\.google\.com|bard\.google\.com|claude\.ai|copilot\.microsoft\.com|edgeservices\.bing\.com|chat\.deepseek\.com|grok\.com|meta\.ai|chat\.mistral\.ai|you\.com|phind\.com)$

Keep the list in version control and review it quarterly; assistant domains change. The campaign-analytics attribution script uses the same list to classify an ai_assistant channel.

Cross-Domain Tracking

For funnels crossing domains (e.g., acme.com to app.acme.com):

1. GA4 Admin > Data Streams > Configure tag settings > Configure your domains > Add both domains
2. GTM: GA4 Configuration tag > Fields to Set > linker > Add domains
3. Admin > Data Streams > List unwanted referrals > Add both domains

Verification: Visit domain A, click link to domain B, check GA4 DebugView. The session should NOT restart. If a new session starts, cross-domain tracking is broken.