Analytics Setup

Overview

Establishes the analytics infrastructure that every other agency skill depends on for data. Creates standardized, agency-wide tracking conventions that work across all clients while allowing client-specific customization. Covers GA4 configuration, conversion event taxonomy, UTM standards, attribution modeling, dashboard design, and tag management — ensuring clean, reliable data from day one.

When to Use

• Onboarding a new client (before any audits or campaigns)
• Existing client has messy, unreliable, or incomplete analytics
• Launching a new campaign that needs tracking setup
• Client migrating from Universal Analytics or another platform
• Agency standardizing analytics practices across all accounts
• Data discrepancies are causing reporting issues

Prerequisites

• juma-client-context — business model, conversion types, marketing channels, goals
• juma-onboarding-checklist (recommended) — analytics is part of the onboarding process

Process

Step 1: Audit Existing Setup

Before building, assess what exists:

1. GA4 property: Exists? Properly configured? Data flowing?
2. GTM container: Exists? Organized? Who has access?
3. Conversion tracking: What events are being tracked? Are they firing correctly?
4. Cross-domain tracking: Needed? Configured?
5. Data integrity: Filters, bot traffic exclusion, internal traffic exclusion
6. Historical data: How much useful historical data exists?
7. Platform pixels: Meta, Google Ads, LinkedIn, TikTok — installed and verified?

Document findings and gaps.

Step 2: Design Event Taxonomy

Create a standardized naming convention for all events:

See event-taxonomy.md for the complete agency-wide event naming convention, category definitions, examples, and key conversion event list.

Step 3: Configure GA4

Property-level settings:

1. Data retention: Set to maximum (14 months)
2. Google Signals: Enable for cross-device tracking
3. Data collection acknowledgment: Configure per client's privacy requirements
4. Internal traffic filters: Define by IP range
5. Unwanted referral exclusions: Payment processors, SSO providers
6. Cross-domain measurement: Configure if multi-domain

Data streams:

1. Web stream with enhanced measurement enabled
2. Review enhanced measurement settings (page views, scrolls, outbound clicks, site search, video engagement, file downloads)
3. Disable any enhanced measurement that conflicts with custom tracking

Custom dimensions and metrics:

Dimension/Metric	Scope	Purpose
client_segment	User	Audience segment identification
content_type	Event	Content categorization
campaign_tier	Event	Campaign hierarchy
lead_score	User	Lead quality indicator

Step 4: Set Up GTM Container

See gtm-container-template.md for the complete container structure, naming conventions for tags/triggers/variables, folder organization, and best practices.

Step 5: Define UTM Standards

See utm-standards.md for the complete agency-wide UTM parameter convention, rules, common medium values, and example UTM strings.

Step 6: Select Attribution Model

Recommend attribution model based on business model:

Business Model	Recommended Model	Rationale
Short sales cycle (ecomm)	Data-driven or last-click	Purchase decision is quick
Long sales cycle (B2B SaaS)	Data-driven or linear	Multiple touchpoints matter
Lead gen	Data-driven or position-based	First and last touch both important
Brand + performance mix	Data-driven	Let the algorithm weight touchpoints

Document:

• Primary attribution model for reporting
• Secondary model for comparison/validation
• Lookback windows per channel
• Cross-device considerations

Step 7: Design Dashboards

Executive dashboard (for client stakeholders):

• Overall KPIs: Sessions, conversions, conversion rate, revenue/leads, CPA/ROAS
• Trend lines: Month-over-month, year-over-year
• Goal progress: Actual vs target
• Top-level channel performance

Channel dashboard (for agency team):

• Per-channel deep dive: Traffic, engagement, conversion metrics
• Campaign-level performance
• Audience insights: Demographics, devices, geography
• Content performance: Top pages, landing pages, exit pages

Campaign dashboard (for active campaigns):

• Campaign-specific KPIs
• Budget pacing
• Creative performance
• A/B test status

For each dashboard, specify:

• Tool: Looker Studio, GA4 Explorations, or client's preferred BI tool
• Refresh cadence: Real-time, daily, weekly
• Access: Who can view, who can edit

Step 8: Configure Alerts

Set up automated alerts for:

Alert	Condition	Notification
Traffic drop	>20% decrease vs previous period	Email to account team
Conversion spike/drop	>30% change in conversion rate	Email + Slack
Budget pacing	>110% or <80% of daily budget	Email to media team
404 errors	>10 per hour	Email to SEO team
Goal completion	Monthly target reached	Email to account team

Step 9: Document Everything

Create a client-specific analytics documentation:

1. Property and container IDs: GA4, GTM, platform pixel IDs
2. Event taxonomy: Full list of custom events with descriptions
3. Conversion events: Which events are marked as conversions
4. UTM convention: Client-specific examples
5. Dashboard locations: URLs and access instructions
6. Alert configurations: What triggers what notification
7. Access log: Who has access to what, at what permission level
8. Change log: History of configuration changes

Output Format

# Analytics Setup: [Client Name]
**Prepared by**: [Agency Name] | **Date**: [Date]

## Setup Summary

| Component | Status | Details |
|-----------|--------|---------|
| GA4 Property | [Configured/Existing/New] | Property ID: [ID] |
| GTM Container | [Configured/Existing/New] | Container ID: [ID] |
| Conversion tracking | [N] events configured | [List key events] |
| Platform pixels | [List platforms] | [Status each] |
| UTM standards | Documented | [Link to UTM builder] |
| Dashboards | [N] created | [List with URLs] |
| Alerts | [N] configured | [List key alerts] |

## Event Taxonomy

| Event Name | Type | Description | Conversion? | Parameters |
|-----------|------|-------------|-------------|------------|
| [event_name] | [Custom/Enhanced] | [Description] | [Yes/No] | [key params] |

## Conversion Events

| Event | Type | Value | Counting |
|-------|------|-------|----------|
| [event] | [Primary/Secondary] | [$X or N/A] | [Once per session/Every] |

## UTM Standards

| Parameter | Convention | Client Examples |
|-----------|-----------|-----------------|
| utm_source | [Convention] | [Examples] |
| utm_medium | [Convention] | [Examples] |
| utm_campaign | [Convention] | [Examples] |

## GTM Container Structure

### Tags
| Tag Name | Platform | Type | Trigger | Status |
|----------|----------|------|---------|--------|
| [Name] | [Platform] | [Type] | [Trigger name] | [Active/Draft] |

### Key Triggers
| Trigger Name | Type | Conditions |
|-------------|------|------------|
| [Name] | [Type] | [Conditions] |

## Attribution Model
- **Primary model**: [Model name]
- **Rationale**: [Why this model]
- **Lookback window**: [N days click, N days view]

## Dashboards

| Dashboard | Audience | Tool | URL | Refresh |
|-----------|----------|------|-----|---------|
| Executive | Client stakeholders | [Tool] | [URL] | [Cadence] |
| Channel | Agency team | [Tool] | [URL] | [Cadence] |
| Campaign | Media team | [Tool] | [URL] | [Cadence] |

## Alerts

| Alert | Condition | Recipients | Channel |
|-------|-----------|------------|---------|
| [Alert name] | [Trigger condition] | [Who] | [Email/Slack] |

## Access Log

| Platform | Account/Property | User | Permission Level |
|----------|-----------------|------|-----------------|
| [Platform] | [Account] | [User/Email] | [Admin/Editor/Viewer] |

## Documentation & References
- Event taxonomy spreadsheet: [Link]
- UTM builder: [Link]
- GTM change log: [Link]
- Analytics playbook: [Link]

Common Mistakes

• No naming convention — Without a standard, event names become inconsistent across clients and team members. Establish and enforce the convention
• Over-tracking — Tracking everything creates noise. Focus on events tied to business outcomes
• Forgetting enhanced measurement conflicts — GA4's built-in enhanced measurement can conflict with custom GTM tags, causing double-counting
• No internal traffic filter — Agency and client team visits skew data. Always filter internal traffic
• Missing cross-domain setup — If the client has multiple domains in the user journey (e.g., blog on subdomain, checkout on different domain), cross-domain tracking is essential
• Skipping the documentation — When the team member who set it up leaves, undocumented analytics becomes a black box. Document everything
• Not testing before launch — Always use GTM Preview mode and GA4 DebugView before publishing changes

Related Skills

• juma-onboarding-checklist — Analytics setup is a key onboarding step
• juma-reporting — Clean analytics data enables accurate reporting
• juma-ab-test-plan — Testing requires proper conversion tracking
• juma-campaign-plan — Campaign measurement depends on analytics setup
• juma-paid-media-plan — Conversion tracking and attribution for paid media
• juma-channel-audit — Accurate channel data requires proper tracking