GA4 vs Universal Analytics: what changed and what it means

Universal Analytics is no longer a choice. According to Google, the old version stopped processing data in July 2023 and the historical data was permanently deleted roughly a year later. GA4 is therefore quite simply the only Google Analytics still running. The real difference is not a choice between two tools, but a completely different measurement model: where Universal Analytics thought in sessions and pageviews, GA4 thinks in individual events. In this article you will read what changed, why your old UA numbers no longer hold up and how to avoid steering on the wrong figures.

Why did Universal Analytics disappear?

Universal Analytics, internally also known as GA3, was the standard for years. But the tool dates from an era of desktop traffic, third-party cookies and a far simpler web. As users started moving across apps, mobile and multiple devices, UA’s session model ran into its limits.

So Google pulled the plug. The free Universal Analytics properties stopped measuring in July 2023, and the old data was deleted afterwards. Anyone still reporting today on numbers from their old UA account is working with figures that no longer exist or that were never comparable to what GA4 shows. That is exactly where much of the confusion comes from.

With a well-considered data & analytics approach, every trajectory starts here: making sure you measure what brings in customers, rather than comparing old apples with new pears.

What is the difference between GA4 and Universal Analytics?

The fundamental difference lies in how the two tools look at the world.

Universal Analytics organised everything around the session: a visit with a beginning and an end, built up from pageviews. That worked fine for a classic website, but it left gaps. A video that was watched, a click on a button or a scroll did not automatically count.

GA4 revolves around the event. According to the Google Analytics 4 documentation, every interaction in GA4 is an event: a pageview, a click, a scroll, a form start, a download. Those events are independent of the session and you can measure and analyse them flexibly. The result is that you see far more than before, but also that you cannot simply lay old and new reports side by side. A “session” in UA and an “event” in GA4 measure different things.

A few concrete consequences of that switch:

• One property for web and app. In UA you needed separate properties for that. GA4 brings website traffic and app data together in a single overview.
• More privacy-friendly measurement. GA4 leans on first-party cookies and fills data gaps with modelling, so it aligns better with rules such as the GDPR. If you want to know why that matters, also read our article on first-party data.
• No more fixed hit limits. Where UA capped the number of hits per property, GA4 does not apply those ceilings in the same way in practice.

Why do your old UA statistics no longer add up in GA4?

This is the source of most frustration. People open GA4, look for a familiar number and cannot find it, or find something that looks completely different.

The clearest example is bounce rate. In UA it was a core metric. GA4 pushed it into the background and focuses on engagement instead. Rather than counting who clicks away immediately, GA4 measures so-called engaged sessions: visits that last longer than ten seconds, contain at least one conversion event or view multiple pages. That is a more positive and fairer signal, because a visitor who reads one long page for ten minutes counted as a “bounce” in UA.

Other things also work differently or have disappeared. Filters, data views and blocking IP addresses the way you did it in UA are no longer in the same place. The most important thing to remember: if a GA4 figure differs from what you know from your UA memory, that is usually not an error. The definition has simply changed.

That is why it is pointless to compare today’s GA4 numbers with UA numbers from two years ago. Rebuild your reference points inside GA4 itself.

What can GA4 do that Universal Analytics could not?

The event model opens a few doors that were closed in UA:

• Predictive metrics. GA4 uses machine learning to estimate things like purchase probability, churn probability and expected revenue. Useful for spotting your best contacts, provided you have enough data.
• Deeper integration with Google Ads and BigQuery. You can build audiences around high-value visitors and export your raw data to BigQuery for your own analysis. That BigQuery integration used to be reserved for paying GA360 customers.
• More flexible reports. You adapt dashboards and follow the customer journey from acquisition to conversion in one line.

It sounds impressive, and it is. But more features and more data points do not automatically lead to better decisions. That is where the real pitfall sits.

How do you use GA4 without drowning in numbers?

We are a small team and we keep seeing the same pattern at B2B companies: GA4 is switched on, every event is collected, and then everyone stares at dozens of charts without knowing what to do. Honest advice: that is the wrong order.

Start with the question that matters: how many qualified enquiries does this deliver? For a B2B company, which is not a webshop with thousands of transactions a day, the number of hard conversions matters more than every micro-interaction GA4 can measure. A quote request, a demo booking, a completed contact form: those are the events you really need to set up and monitor.

In practice that means:

• Define your conversions first. Establish which events bring a customer closer before you measure anything else. Our guide to conversion tracking shows how to set that up.
• Monitor your conversion rate per channel. Not as a vanity metric, but to see which channel brings in customers and which channel only costs money.
• Connect source to result. Only with a solid marketing attribution model do you know which effort caused the enquiry.

Do not ignore the rest, but do not let it steer you either. A hundred data points nobody looks at are worth less than three numbers you decide on every week.

Frequently asked questions about GA4 vs Universal Analytics

Can I still go back to Universal Analytics? No. According to Google, UA stopped measuring in July 2023 and the old data was deleted afterwards. GA4 is the only version still running. The choice between the two therefore no longer exists; the only question is how well you set up GA4.

Why do my GA4 numbers look so different from my old UA reports? Because the definitions differ. UA counted sessions and pageviews, GA4 counts individual events and looks at engagement instead of bounce rate. The same reality, a different yardstick. So do not compare old UA numbers with new GA4 numbers.

Is GA4 free? Yes, just like Universal Analytics, the standard version of GA4 is free to use.

What is the most important difference between GA4 and Universal Analytics? The measurement model. UA was session-based, GA4 is event-based. As a result GA4 measures more interactions and works better across web and app, but its numbers are not comparable one to one with those of the past.

Do I need to set up every event in GA4? No. Set up your hard conversions properly first, because those predict revenue. You only measure the rest if it helps you decide, not because you can.

Ready to steer on the right numbers?

GA4 gives you more data than UA ever did, but data only becomes valuable once it tells you which channel brings in customers. We help you set up your tracking so that you steer on enquiries and revenue, not on charts that look good but say nothing. No noise, just knowing what works.

Book your free intake