Analytics 4

Google's new Analytics 4 is set to fully replace Universal Analytics (Analytics 3) on July 1st 2023. We have about a year left to transfer from the old system to the new.
Analytics 4

The end of an era

Google’s new Analytics 4 is set to fully replace Universal Analytics (Analytics 3) on July 1st 2023. We have about a year left to transfer from the old system to the new. We’re looking at many changes ahead in the dashboard, implications on GDPR and the website user experience, and all that using machine learning. 

Universal Analytics

Analytics 4 was released October 14th 2020, and has been a great tool to gain insights about the traffic coming into websites, and what they do when they’re there. But there surely was plenty of room for improvement. The tool could be hard for new users to understand, its interface isn’t always easy to navigate and some of the terminology doesn’t always make sense.

Universal Analytics
Universal Analytics

Privacy Implications & GDPR

The biggest change, that I want to address first, is what is happening to both the front end, for the user, and behind the scenes, with the data itself. Simply put, tracking users is going to change, and will build towards a cookie-less future, in which you won’t have to notify your users of using cookies anymore. No more cookie notices!

Okay, you still have to have to add a snippet of code to your HTML, but this has been made a lot easier in recent year, using Google’s Site Kit plugin for WordPress.

With the enactment of the EU GDPR bill, cookies have been made more visible to the user, however all the notices of who is tracking what and why, hasn’t lead to an improved user experience. Google is trying to find a Privacy-First solution to this by going away from cookies and turning to machine learning instead.

This approach is combined with a more convenient way to export data to BigQuery. And giving you the ability to choose where you want the data to be stored.

Machine Learning

Analytics 4 is build on Firebase and has machine learning at its core. It attempts to learn from the data to automatically surface insights it thinks are useful, about your data.

In practise this means that at its foundation Google tries to use machine learning to replace cookies. Rather than track individual users all across the internet, it bases its core knowledge on how users behave on individual websites and individual industries.

If the system functions as some hope it might, this will provide the ability to predict things such as purchase probability, churn probability and revenue prediction. With that knowledge, customer journeys can be tailored to high value customers in a completely different way.

Updates to the interface

In one sense the interface has been simplified, which at its face seems as something which limits the data which can be gathered from the dashboard. However, the way to display has changed, and I would argue, it makes it easier to get the information you need.

Not every website, business or even locality is the same. The dashboard in Universal Analytics is very set in stone; you can find a lot of data through pre-made lists and charts, but you’re bound by the limitations set in the original design.

The new dashboard is all about tailoring individual dashboards to the data which is most relevant to you. It will take some time to configure, but in essence what it does, is merge some of the functionality of Google’s Data Studio directly into Analytics. Bringing the customization options of Data Studio to Analytics, and removing a lot of the need to capture data in one platform, and display it in another.


In an attempt to make Analytics easier to understand, Google has made some major changes to some commonly used terminology. An important example of this is “bounce rate”. It’s often been hard for me to explain to customers that a high bounce rate is a negative, it felt counter-intuitive. By changing the term into “Engagement rate”, it now becomes a more action-oriented metric.

AdWords integrations

Way major improvement that Analytics 4 will bring for your website, is its integration with AdWords. In short it will be able to communicate with AdWords. When making a new campaign, you can link your website, and AdWords will scan the website for the Analytics Tag, making it easier to track conversions without as much manual labour.

This does mean that the current “page view” and “custom events” go away, but in its place turns everything into events. This benefit allows more accurate tracking across web and mobile apps, by not being bound by URLs and specified interactions.

Furthermore, by having a more intuitive connection between AdWords and Analytics, these platforms will be able to learn from each other, and create more relevant experiences for potential customers.

Google Analytics
Google Analytics

Tracking "Events"

There are 4 catagory groups of “events” tracked in Analytics 4:

  1. Automatically collected events.
  2. Enhanced measurement events.
  3. Recommended events.
  4. Custom events.

Automatically collected events

The automatically collected events, such as page_view, session_start and first_visit, are set automatically after installing the GA4 base code.

Enhanced measurement events

The enhanced measurement events are the types of events which speak more to the interactions of the users with the website, these include things such as scroll, site search and video engagement.

Recommended events

The recommended events are where the AI behind Analytics really comes in to assist your insights. These are events that Google recommends based on the data it has gathered on your industry. This way you benefit from your competitors insights, alongside your own, for the system to recommend insightful events that it thinks are relevant to your business. These can include add_to_cart, purchase and begin_checkout.

Custom events

The custom events are events that can be customized individually. These can be useful in A/B testing, or when the other events don’t fit a specific event relevant to you.

What happens now?

We have until July 1st 2023 to transition to Analytics 4, before Universal Analytics becomes deprecated. What you can do is to setup the new, and run it alongside. Giving you the ability to transition whilst retaining that foundational understanding of your data.

Read more