GA4 – The new Google Analytics, should you upgrade?

Google Analytics 4 is here! Released mid-October GA4 has been built to provide a more holistic and insight driven analytics platform- and one that is more suited to evolving privacy regulations and considerations. 

GA4 can’t be compared to Universal Analytics but anyone who has used Firebase and/or the previous App + Web property released back in summer 2019 will spot quite a few similarities. GA4 has been built to be a more powerful version of App + Web with the previous name providing an indication of the influence App tracking methods will have going forward.

Those who have been using Google Analytics to track only websites to date, will find the biggest difference in navigating the new GA reporting interface – but it isn’t just reports that look and feel different, the fundamental data model that GA4 uses is different.

A short recap of Google Analytics

Looking back at previous versions of Google Analytics- there have been 4 big shifts in the product to date, starting when Google Analytics purchased Urchin back in 2005 and launched Google Analytics.

The main features included:

• Unique visitor tracking
• Visitor segment reporting
• Marketing campaign result reporting

Four years later in 2009, they launched the Asynchronous version with additional tracking capabilities including:

• Event tracking
• Multi-channel funnels
• Realtime reporting

Then 5 years later in 2014 came Universal Analytics that brought with it:

• Custom dimensions and metrics
• Attribution reporting
• Measurement protocol

And now, skipping forward to 2020 we have Google Analytics 4, bringing with it the following features:

• Machine learning driven
• Privacy centric tracking
• Automatic and Web + App tracking

Sure, there have been tweaks in the original codes in-between these releases, but these are the 4 main Google Analytics product changes.

So what has changed?

Quite a bit has changed but from a GA user point of view the biggest change is in how user interactions with the site is tracked and reported on, when compared to Universal Analytics.

Most of the main navigation and sub reports have changed- from the naming conventions to how the reports are displayed, and some reports, metrics and dimensions have been removed. Tracking and configurations are different- the fundamental way Google Analytics thinks, and works is changing, the data model is moving to an event-based model, everything is considered an event- even a pageview is now considered an event!

Events are also not what you’d traditionally think of- there is no such thing as Category, Action or Label as you’d commonly see in Universal Analytics. The most comparable to these traditional events are called enhanced measurement events. Google have tried to simplify in-page tracking by offering enhanced measurement- automatic tracking of outbound clicks, video views, downloads and scrolls etc. These can just be enabled or disabled and configuration of how these work cannot be changed (apart from site search and pageviews). Not all these events have parameters linked to them so parameters on some events aren’t directly comparable, but manual custom parameters can be sent – although won’t show in most standard reports but will be available via BigQuery.

Custom dimensions are sent via new User properties. User properties are any user attribute that you want to use to describe user types. You have up to 25 uniquely named, user properties- outside those that are automatically set by GA4  (age, country, device, gender) and custom user properties don’t need to be configured in the interface the way custom dimensions currently are required to be.

GA4 has to be set-up under a new property and uses a measurement ID that ties the data to a ‘data stream’. If you select to Upgrade your existing property- this will automatically create a new property- there is no way to ‘migrate’ Google Analytics Universal data to GA4.

Any events that are feeding into Universal Analytics- whether this be by manually tagging or tags set-up in GTM- need to be created from scratch under GA4.

If you use Google Analytics 360- GA4 for 360 is still being worked on but it is thought that more details on this will be announced in Q2 2021.

When will Universal Analytics be deprecated?

What we know historically is that support for each previous version of Google Analytics has been slowly removed– in general within 1-2 years and during a recent Q and A held by Cardinal Path a GA4 Project Lead estimated that within 18 months they expect all clients to have GA4 implemented BEFORE a deprecationtimeline is then announced.

Again, other iterations of Google Analytics have maintained a similar look and feel but reports in GA4 are modelled on the previous Firebase/Web and App reports and are built for web and app reporting and analysis- an event driven model. So while Universal Analytics will continue to be supported for some time, it is inevitable that support will be withdrawn at some stage and as GA4 is quite a shift change from the current version- what this means is we need to accept and get used to this new version of GA!

In summary- should we change to GA4?

What are the advantages of using GA4?

This product is built to be smarter through data-driven predictions and insights- churn probability, trend alerts, revenue forecasting– all key predictive modelling selling points of GA4 that are fuelled by Google Analytics machine learning models. We’ve had intelligence alerts, analytics intelligence, custom alerts and annotations- all these could be considered current background noise or add-ons rather than a more obvious and integral part of what is to come with GA4 intelligence.

Universal Analytics required additional tags to track common user behaviour interactions on pages, via GA4, as explained above- you can now opt to track key page events by default, without worrying about tracking – from outbound clicks to downloads. The import wizard means adding data layers is made easier through Data Import and the removal of quotas when exporting data to BigQuery (for free) also makes GA4 a more open and powerful tool.

One of the main driving forces behind this new product is that is it privacy centric- meaning that where there was potential before for gaps in data due to cookie restrictions and privacy laws, GA4 has been built with less dependency on cookies. It also allows easier control over data collection including data deletion, allowing user data to be deleted based on specific events or user properties, within specific date ranges:

Next steps to take with GA4

Because data can’t be retrospectively updated (data starts to build once the tracking is installed) what we are recommending all clients to do is to create a new GA4 property, start the stream of data feeding into the new set of reports and this will allow you to test the set-up, configuration and reporting UI– helping you visualise what the new measurement plan might look like, while still keeping your current (UA) data in the meantime.

We’ve experienced no issues in running both GA and GA4 concurrently- in fact, the advice coming from Google Analytics is for anyone currently running on Universal Analytics is to dual tag so this approach can be used to build familiarity with GA4 while keeping Universal GA as the primary platform until you are ready to switch. GA4 is lacking a lot of the features that are available in Universal Google Analytics-  There are still lots of improvements in the GA4 roadmap- especially around goals, segmentation, ecommerce tracking and data studio integration but each week we are excited to see more updates being released..

In short, this isn’t as simple or straight-forward as previous GA product updates were. It requires a thoughtful and considered future migration to a very different property. The Google Analytics team know this and is providing us with the chance to run both old and new properties alongside each other to allow us time to become familiar with GA4. This is the Google Analytics we’ll be using in years to come, so dual tag – get your new data stream set-up and feeding data into a GA4 property, navigate the interface and then start to plan a measurement framework around it.

This is our first in a series of blogs on GA4. Keep an eye out for our upcoming blogs on GA4:

• Tag installation options for GA4
• Creating a measurement plan for GA4
• Navigating the GA4 User Interface