What Is GA4 (Google Analytics 4)?
GA4 is the fourth major iteration of Google Analytics, launched in 2020 and made the default platform in 2023 when Universal Analytics (UA) was retired. The most significant change from its predecessor is the shift from a session-based data model to an event-based data model. In Universal Analytics, the fundamental unit of measurement was the session, which contained hits such as page views and events. In GA4, every interaction is itself an event with associated parameters, giving marketers far more granular data about exactly what users do.
Out of the box, GA4 automatically collects a set of recommended events including page views, scrolls, outbound clicks, video engagement, and file downloads. These enhanced measurement events fire without any custom code, simply by enabling them in the GA4 settings. Custom events can be added through Google Tag Manager or directly in the GA4 interface using the DebugView to verify they fire correctly. This flexibility makes it possible to track nearly any user interaction on a website, from clicking a phone number link to using a product filter or watching a product video.
One of GA4's most significant improvements for South African businesses is its approach to cross-channel attribution. The Traffic Acquisition and User Acquisition reports show how different marketing channels contribute to conversions, and the Advertising workspace provides a path report that maps the sequence of touchpoints a user took before converting. This multi-touch view is far more useful than the last-click attribution that dominated Universal Analytics, helping businesses understand the true contribution of SEO, paid advertising, email, and social media to their overall results.
GA4 is also designed with privacy in mind. It uses modelled conversions and behavioural modelling to fill gaps in data caused by users who reject cookies, use browser-level tracking prevention, or are in regions with strict privacy laws. While no analytics platform is fully immune to data gaps in the post-cookie landscape, GA4's machine learning-based gap-filling produces more accurate overall conversion counts than Universal Analytics was able to provide. For South African businesses with POPIA (Protection of Personal Information Act) compliance obligations, GA4's configurable data retention and IP anonymisation settings provide relevant controls.
GA4 In Practice
The scenario below is an illustrative example, not a Juicy Designs client result. The figures indicate the scale of effect that GA4 event tracking and funnel reporting typically produce, so treat them as indicative rather than measured.
Picture a Cape Town e-commerce store that migrates from Universal Analytics to GA4 and configures custom events to track add-to-cart actions, checkout initiations and purchases. Using GA4's Funnel Exploration report, the team might discover that around 60% of mobile users who add an item to the cart drop off at the shipping address step. Digging into that step could point to a form usability problem specific to mobile, such as the autofill prompt being blocked by an overlay element. After fixing an issue like that in the next site update, mobile checkout completion rates could plausibly increase by close to 30%, a revenue impact that would be traceable to GA4's event-level reporting.
For businesses that previously relied on Universal Analytics, the transition requires attention. GA4 data cannot be compared directly to historical UA data because the measurement models are different. Metrics such as sessions, bounce rate, and users are calculated differently. The recommendation is to maintain GA4 data going back at least 14 months before relying on year-on-year comparisons, and to extend data retention settings from the default two months to 14 months in the GA4 admin settings as one of the first configuration steps after setup.
How GA4 differs from Universal Analytics
Google Analytics 4 (GA4) is the current version of Google Analytics, which replaced the older Universal Analytics, and it differs from its predecessor fundamentally in how it models data. Universal Analytics was built around sessions and pageviews; GA4 is built around events, treating almost every interaction, a pageview, a click, a scroll, a purchase, as an event, which gives a more flexible, behaviour-focused view and unifies measurement across websites and apps. GA4 also emphasises engagement-based metrics, such as engaged sessions and engagement rate, over the old bounce rate, and is designed with a more privacy-conscious approach as tracking becomes harder, relying more on modelling. These differences mean GA4 metrics are not directly comparable with Universal Analytics figures, and reports look and work differently, which is why moving to it required relearning rather than a simple upgrade.
Key concepts in GA4
Understanding GA4 rests on a few concepts. Events are the foundation: everything measured is an event, and marketers designate the important ones, a purchase, an enquiry, as conversions, which then appear in conversion reports and can feed advertising. Engaged sessions and engagement rate, sessions lasting over ten seconds, converting, or with multiple pageviews, replace bounce rate as the headline measure of whether visits were worthwhile. GA4 spans web and app in one property, giving a unified view where relevant. It uses data-driven attribution and modelling to fill gaps left by privacy limits, so some figures are estimated rather than directly tracked. And its exploration reports allow flexible custom analysis beyond standard reports. Grasping that GA4 is event-based, engagement-focused and privacy-aware, rather than session-and-pageview-based like the old tool, is what makes its reports and metrics make sense.
FAQ
Why are GA4 metrics different from Universal Analytics?
GA4 uses an event-based model and counts unique users and engaged sessions differently from Universal Analytics. The bounce rate is replaced by an engagement rate, sessions are counted differently because GA4 resets the session at midnight, and some metrics like pages per session are no longer defaults. These differences mean historical Universal Analytics data cannot be directly compared to GA4 data.
Do I need GA4 if I already have Universal Analytics data?
Yes. Universal Analytics stopped collecting new data in July 2023. If you are still relying on old Universal Analytics reports, you are making decisions based on data that is over two years old. GA4 is the only active Google Analytics platform and the one required for integration with Google Ads, Search Console, and all current Google marketing products.
Do you need GA4 if you have Universal Analytics data?
Yes, GA4 is the current, supported version, and Universal Analytics has stopped processing new data, so GA4 is needed to continue measuring. Historical Universal Analytics data cannot be carried into GA4 because the models differ, which is why moving to GA4 early to build history was important.