Google Analytics 4 (GA4)
Google Analytics 4 is Google's current analytics platform, built on an event-based model. It replaced Universal Analytics in 2023 and tracks users across devices.
Google Analytics 4 is Google’s analytics platform that replaced Universal Analytics, using an event-based data model designed for cross-device tracking and a privacy-first web.
What Google Analytics 4 Means in Marketing
Most marketing teams run on data from Google Analytics. From 2005 to 2023, that meant Universal Analytics. GA4 is the successor, and the transition was not optional: Google stopped collecting new data in Universal Analytics properties in July 2023.
The change is not cosmetic. GA4 is built on a completely different model. Where Universal Analytics grouped everything into sessions and page views, GA4 treats every user interaction as an event. A page view is an event. A scroll past 90% of the page is an event. A video play, a button click, a form submission, a purchase: all events, each with their own parameters.
This model is more flexible because it captures user behaviour that sessions and page views missed entirely. A single-page app that updates the URL without a full page load, for example, barely registered in Universal Analytics. In GA4, every meaningful interaction can be tracked as a distinct event.
How Google Analytics 4 Works
GA4 tracks users across two surfaces: websites and apps, in a single property. That cross-device view means you can trace a path that starts with a mobile app visit, continues on a desktop browser and converts in a mobile session, without stitching together separate properties.
Key concepts that differ from Universal Analytics:
- Events replace hits. Every interaction is an event. Page views, scrolls, clicks and purchases are all events with parameters attached.
- Engagement rate replaces bounce rate. An engaged session is one lasting over 10 seconds, with two or more page views, or a conversion event. Bounce rate is the inverse.
- Explorations. The free-form analysis section allows custom funnels, path reports and segment overlaps that required expensive add-ons in the old platform.
- Data retention. GA4 stores user-level data for a maximum of 14 months by default, after which only aggregated data remains. This is shorter than many teams realise.
Google Analytics 4 Example
A retail brand running both a website and an iOS app previously tracked them in separate properties, which meant any customer who browsed on the app and bought on the desktop looked like two different people. In GA4, with proper user ID configuration, that entire journey can be attributed to one person, making attribution and lifetime value modelling considerably more accurate.
Why Google Analytics 4 Matters for Marketers
GA4 is not just a new interface for the same data. Teams that approach it as a direct replacement for Universal Analytics will miss the point and find the numbers confusing. Teams that learn the event model and build their tracking architecture thoughtfully will have a more complete picture of user behaviour than they have ever had from free analytics.
Start with the basics: ensure key events are firing, conversions are configured and the engagement metrics are understood before drawing conclusions from the data.
Frequently Asked Questions
What is the difference between GA4 and Universal Analytics?
Universal Analytics used a session-based model: it grouped hits into sessions tied to cookies. GA4 uses an event-based model where everything, including page views, scrolls, clicks and purchases, is an event with parameters. GA4 also tracks across devices and platforms in a single property, and it is built around a privacy-first architecture that works better in a world with fewer third-party cookies.
When did Universal Analytics stop working?
Google stopped processing new data in Universal Analytics properties on 1 July 2023. Historical Universal Analytics data became inaccessible from 1 July 2024. Every site that wanted to keep tracking needed to have migrated to GA4 before the first date. Historical GA4 data does not match Universal Analytics data, so comparisons across the transition period are unreliable.
Why does GA4 show different numbers from Universal Analytics?
Because the data models are fundamentally different. A session in UA and an event in GA4 are different constructs, bounce rate has been replaced by engagement rate, and page views are counted differently in single-page apps. If your GA4 numbers differ from UA, it is usually not a measurement error but two different systems measuring overlapping things.