Google Analytics 4 vs. Universal Analytics

Google Analytics 4 is not Universal Analytics with a different interface. Its event-based data model changes how interactions are collected and how familiar metrics are calculated.

Universal Analytics has since been retired for normal data collection, so this comparison is most useful when reading historical reports or explaining why old and new numbers do not match.

Sessions are no longer the centre of everything

Universal Analytics organised most activity around sessions and pageviews. GA4 records page views, scrolls, purchases, and other interactions as events with parameters.

That makes web and app measurement fit into one model and allows more flexible event reporting. It also means an old UA goal or event cannot always be copied across without deciding what the interaction should mean in GA4.

GA4 can collect some interactions through enhanced measurement, including scrolls, outbound clicks, site search, video engagement, and file downloads. Check the property settings rather than assuming every event is present.

Familiar names can hide different calculations

Several metrics look similar while measuring different things:

• GA4 reporting often emphasises active users, while UA commonly emphasised total users.
• GA4 combines web page views and app screen views in a shared event model.
• GA4 conversions are selected events; UA used goals and ecommerce tracking.
• Session counts can differ because the products start and attribute sessions differently.
• GA4 bounce rate is based on sessions that were not engaged, rather than UA’s single-page session definition.

These differences mean the migration date creates a break in the data. I would annotate that break in reports instead of pretending two differently defined metrics form one clean trend line.

Identity is still imperfect

Both products can use browser identifiers, and GA4 can also use signals such as a site-provided User-ID where configured and permitted. This does not create a magic identifier that follows every visitor accurately across devices.

Consent, browser restrictions, deleted cookies, implementation choices, and modelling all affect the result. Analytics numbers are measurements with assumptions, not a census.

The reporting workflow changes too

GA4 includes standard reports, explorations, integrations with products such as Google Ads, and predictive features where a property meets the relevant requirements. The interface and terminology can take time to learn, particularly if a team built years of reporting around UA.

The sensible migration was therefore not just installing a new tag. It involved defining events and conversions, validating them, updating dashboards, and documenting metric changes.

When comparing GA4 with old Universal Analytics data, compare definitions before values. A familiar label is not proof that the number was calculated the same way.