Google Analytics 4 (GA4) is now the standard for website analytics — Universal Analytics was permanently shut down in 2023. But GA4's event-based model and new interface feel very different from what many website owners were used to. This complete guide covers GA4 from scratch: understanding the new data model, setting up tracking correctly, navigating the most important reports, configuring conversions, using Explorations for deep analysis, and connecting GA4 with other tools to get a complete picture of your website's performance.
2026 Update: Google Analytics 4 now requires Consent Mode v2 compliance in the EU, UK, and EEA markets. Sites without proper consent mode implementation risk data gaps and potential regulatory issues. This guide covers consent mode setup in Section 10.
Google Analytics 4 is Google's fourth generation analytics platform, built around an event-based data model instead of the session-and-pageview model of Universal Analytics. Every user interaction — from pageviews and clicks to scrolls, video plays, and purchases — is tracked as an "event" with associated parameters.
| Feature | Universal Analytics | Google Analytics 4 |
|---|---|---|
| Data Model | Sessions + Pageviews | Events + Parameters |
| Cross-Platform | Web only (separate app) | Web + App unified |
| Bounce Rate | Single-page session exits | Replaced by Engagement Rate |
| BigQuery Export | GA 360 only (paid) | Free for all |
| Funnel Analysis | GA 360 only (paid) | Free Explorations |
| Predictive Metrics | Not available | Purchase probability, churn probability |
| Data Retention | Up to 50 months | Up to 14 months default |
Setting up GA4 correctly from the start saves hours of debugging later. Follow these steps in order:
Go to analytics.google.com → Admin → Create Property. Enter your property name (your business name) and select your time zone and currency. If you had Universal Analytics, create a new GA4 property alongside it rather than converting.
In your new property → Admin → Data Streams → Add Stream → Web. Enter your website URL and a stream name. Copy your Measurement ID (G-XXXXXXXXXX) — you'll need this to add tracking to your site.
Choose your installation method:
In Data Streams → your stream → Enhanced measurement, enable: Scrolls (90% scroll depth), Outbound clicks, Site search (if you have a search function), Video engagement (YouTube embeds), File downloads. These auto-track without any additional code.
Open GA4 → Reports → Realtime. Visit your website in another tab. You should see 1 active user appear in Realtime within 30 seconds. If not, check: Is the Measurement ID correct? Is the tag firing (use GA4 DebugView via GTM Preview Mode)?
In GA4, every user action is an event. Understanding event types helps you know what's tracked automatically versus what you need to configure:
Tracked without any setup: session_start, first_visit, page_view, user_engagement
scroll, click (outbound), view_search_results, video_start/video_progress/video_complete, file_download
Standard events Google recommends for specific industries: purchase, add_to_cart, begin_checkout, sign_up, login, generate_lead, view_item
Any event you create with your own name and parameters. Use for business-specific actions: chatbot_opened, pricing_viewed, demo_requested, trial_started. Implement via GTM or gtag.js code.
Reports → Acquisition → Traffic acquisition. Shows which channels drive sessions to your site: Organic Search, Direct, Referral, Organic Social, Email, Paid Search.
Key insight: Add "Session default channel group" as a secondary dimension alongside Engagement rate and Conversions to see which channels deliver the highest-quality traffic, not just the most volume.
Reports → Engagement → Pages and screens. Shows which pages receive the most views, unique viewers, average engagement time, and event counts.
Key insight: Sort by "Views" to find your most popular pages, then check "Average engagement time per session" — low engagement on high-traffic pages signals content/UX problems worth fixing.
GA4 replaced Bounce Rate with Engagement Rate. A session is "engaged" if it lasts 10+ seconds, has a conversion event, or has 2+ page/screen views. Engagement rate = engaged sessions / total sessions.
What's good: 60%+ engagement rate is generally healthy. Below 40% on landing pages suggests the page isn't delivering on what users expected from the traffic source.
Reports → Engagement → Conversions. Shows all events marked as conversions with counts and revenue (if applicable). Filter by channel to see conversion rates by acquisition source.
Key setup: Mark your most important events as conversions in Admin → Events → toggle "Mark as conversion". Common conversions: form_submit, purchase, sign_up, schedule_demo.
Reports → Demographics → Demographic details (age, gender, interests). Reports → Tech → Tech details (browser, device, OS). Essential for understanding your actual audience versus assumed audience.
Accessibility implication: Tech reports reveal what % of users are on mobile vs. desktop — critical for prioritizing your accessibility and performance optimization efforts (mobile users often have slower connections and may use assistive technology).
Explorations (formerly Analysis Hub) provides advanced analysis techniques previously available only in GA 360. Access from the left sidebar → Explore.
Visualize the steps users take toward a conversion and see exactly where they drop off. Create funnels for: checkout process (add_to_cart → begin_checkout → purchase), signup flow, or any multi-step user journey.
See which pages users visit after or before a specific page. Understand organic navigation patterns — where do users go after your homepage? What page do people visit immediately before they convert?
Analyze groups of users who share a common characteristic (first visit date, first conversion date). Track retention: what % of users who first visited in week 1 return in week 2, 3, 4?
Compare up to 3 user segments to find intersections. Example: what users are both from organic search AND have high purchase intent (viewed pricing page)?
Analyze users' lifetime value and behaviors from acquisition through their entire history on your site. Essential for subscription and ecommerce businesses.
Build custom pivot table reports with any combination of dimensions and metrics. The most flexible report type for answering specific business questions.
GA4 conversions replace Universal Analytics Goals. Any event can be marked as a conversion, and GA4 shows conversion data across all standard reports.
form_submit, purchase)| Business Type | Recommended Conversions |
|---|---|
| Lead Generation | generate_lead, form_submit, schedule_demo, contact_form_sent |
| Ecommerce | purchase, add_to_cart, begin_checkout, add_to_wishlist |
| SaaS / Software | sign_up, trial_started, pricing_page_view, demo_requested |
| Content / Media | newsletter_signup, video_complete, file_download, article_read |
| Services / Agency | phone_call, contact_form_sent, consultation_booked, portfolio_viewed |
Audiences are segments of users that match conditions you define. GA4 audiences serve two purposes: analysis (filtering reports by audience) and remarketing (using audiences in Google Ads campaigns).
Create audiences in Admin → Audiences → New audience. Audiences are retroactive — they start populating from creation date, not historical data.
Linking GA4 with Google Search Console (GSC) brings organic search keyword data directly into GA4 reports — showing which search queries drive traffic and how those users engage with your site.
Once linked, GA4's "Queries" report shows which Google search queries sent visitors to your site — including clicks, impressions, CTR, and average position alongside GA4 engagement metrics. This is powerful for identifying high-impression, low-CTR keywords where better title tags or meta descriptions could significantly increase traffic.
GA4's default data retention period is only 2 months for user-level data and 14 months maximum (configurable in Admin). After that, historical data in Explorations is no longer accessible.
Action required: Immediately increase data retention to 14 months in GA4 Admin → Data Settings → Data Retention. The default 2-month setting means you lose year-over-year comparison data in Explorations after just 2 months.
Standard reports (Acquisition, Engagement, etc.) aggregate data and retain it indefinitely. Only Explorations (which require raw user-level data) are affected by the retention limit. To keep historical raw data beyond 14 months, export to BigQuery — the free daily export is one of GA4's most valuable features.
Consent Mode v2 is required for sites targeting users in the EU, UK, and EEA as of March 2024. It tells GA4 how to behave based on user consent choices.
| Consent Signal | What GA4 Does |
|---|---|
| analytics_storage: granted | Full tracking — cookies and full event data collected |
| analytics_storage: denied | Cookieless pings only — no user identifiers, data modeled via ML |
| ad_storage: denied | Google Ads remarketing and conversion tracking limited |
| ad_personalization: denied | Ads personalization disabled, no remarketing audiences |
Implement Consent Mode via your CMP (OneTrust, Cookiebot, CookieYes) or directly in Google Tag Manager using the Consent Initialization trigger. Without proper consent mode, GA4 data for EU users may be incomplete and your use of GA4 for advertising may violate GDPR.
Default is 2 months — you lose historical Exploration data immediately. Fix in Admin → Data Settings → Data Retention.
Your own visits inflate metrics. Create an internal traffic filter in Admin → Data Streams → Configure Tag Settings → Define internal traffic, then add an IP exclusion filter.
Without conversions, you can't measure what's working. Mark at least one meaningful business action (form submission, purchase, sign-up) as a conversion immediately after setup.
If users move between your main site and a separate checkout domain, GA4 creates a new session at the domain boundary. Configure cross-domain measurement in Tag Settings.
GA4 calculates metrics differently (sessions, bounce rate, conversions). The numbers are not directly comparable — expect GA4 to show lower sessions and higher conversion rates for the same traffic.
GA4 tells you how users behave on your site. PageGuard tells you why your site may be underperforming technically — slow page loads, accessibility barriers, and SEO issues that drive users away before GA4 can even record an engaged session.
GA4 shows visitor behavior after they arrive. PageGuard reveals technical issues — slow load times, accessibility barriers, SEO problems — that prevent visitors from arriving and converting in the first place. Get a free website health scan in 30 seconds to complement your GA4 analytics.