Getting Started with GA4: What Every Digital Marketer Needs to Know

GA4 has been the default Google Analytics product since July 2023, when Universal Analytics was permanently sunset. And yet, a year later, I still encounter companies running on incomplete migrations, broken event schemas, and dashboards that tell half the story.

This guide is the one I wish had existed when I started. Let's fix that.

---

Why GA4 is Fundamentally Different

The biggest mistake people make is treating GA4 as a "UA upgrade." It isn't. It's a completely different data model.

Universal Analytics used a session-based model:

• A session was the primary unit of measurement
• Hits (pageviews, events, transactions) happened inside sessions
• Everything was scoped to sessions or users

GA4 uses an event-based model:

• Every interaction is an event
• There are no "categories", "actions", or "labels"
• Parameters carry the meaning

This sounds abstract, but the practical implication is huge: your UA reports will not map to GA4 reports. You need to rethink your measurement framework from scratch.

The Core Concepts You Must Understand

1. Events and Parameters

In GA4, everything is an event. A pageview is an event (page_view). A purchase is an event (purchase). A button click is an event you define yourself.

Each event can carry up to 25 custom parameters that describe the action:

// Sending a custom event
gtag('event', 'cta_click', {
  cta_label: 'View Projects',
  cta_location: 'hero',
  page_path: '/home'
})

2. Automatically Collected Events

GA4 automatically collects these events without any code changes:

• session_start — when a new session begins
• first_visit — on first visit
• page_view — on each page view
• scroll — when a user scrolls 90% of a page
• click — on outbound clicks

3. Recommended Events

Google provides a schema of recommended events for common use cases. Use them where possible — they enable Predictive Analytics and future integrations:

Event	Use Case
purchase	E-commerce transactions
generate_lead	Lead generation forms
begin_checkout	Checkout funnel
view_item	Product page views
login	User authentication

---

Setting Up GA4 Correctly

Step 1: Create a GA4 Property

1. Go to Google Analytics Admin
2. Click Create Property
3. Select your industry and reporting time zone
4. Create a Web Data Stream and enter your URL

You'll get a Measurement ID (format: G-XXXXXXXXXX).

Step 2: Implement via Google Tag Manager

Don't add the GA4 snippet directly to your code. Use GTM:

1. In GTM, create a new Tag → Google Tag
2. Enter your Measurement ID
3. Set trigger to All Pages
4. Publish

Step 3: Configure Data Retention

GA4 defaults to 2 months of event data retention. Change it to 14 months in Admin → Data Settings → Data Retention. Do this immediately — you can't recover data retroactively.

Step 4: Connect BigQuery

GA4's free BigQuery export is one of the best things about it. Connect it immediately in Admin → BigQuery Linking. This gives you access to raw event data for advanced analysis.

---

Common GA4 Mistakes to Avoid

Not enabling Google Signals

Google Signals enables cross-device reporting and demographic data. Enable it in Admin → Data Settings → Data Collection.

Relying on Realtime for validation

For validation, use the DebugView — enable debug_mode: true in GTM or use the GA4 Debugger Chrome extension. Realtime has a ~30 second delay and excludes some events.

Not having a tracking plan

Before you start tagging, document your measurement framework. What decisions does this data support? What events map to those decisions? A tracking plan saves weeks of rework.

---

What to Track First

If you're starting from scratch, here's my recommended event priority:

1. Page views (automatic)
2. Scroll depth — 25%, 50%, 75%, 90%
3. Outbound clicks (Enhanced Measurement)
4. CTA clicks — your most important buttons
5. Form submissions — lead generation, newsletter, contact
6. Purchase / lead events (if applicable)

---

Conclusion

GA4 is more powerful than UA when properly implemented. The event-based model, combined with BigQuery export and predictive audiences, opens doors that UA never could.

The key is to invest in a proper measurement framework upfront — not retrofit tracking after launch. Start with your business objectives, work backwards to events, and document everything.