Everything Marketers Need to Know About GA4

GA4-blog

Google Analytics 4 (GA4) has dominated the digital marketing headlines, with issues ranging from the illegality of the solution to emerging information on the true shape of the newest Analytics solution from Google.

And with more marketers getting set up to make the switch from Universal Analytics to its new iteration we thought it was as good a time as any to dig into GA4 and consider:

So, what exactly is GA4?

GA4 is the upgraded version of the existing Google Analytics offering (or Google Universal to give it its proper name).

Google recently announced that it will be sunsetting Google Universal on 1st July 2024. So, all new Google accounts are now set up under GA4 which is the default for new users going forward.

What’s the difference between GA4 and Universal Analytics?

In reality, there are a LOT of differences between GA4 and Universal Analytics. And, as you might expect, a lot of the devil is in the detail.

But here are some of the headline differences between the two:

GA4 provides a platform for both website and mobile app

We will return to this in the next section when we consider the pros of GA4. However, at an outline level, it is important to point out that GA4 improves your capability to track cross-device journeys across web and mobile.

Something that wasn’t possible previously.

GA4 uses an improved event-based data model

One of the biggest changes in GA4 is a move to an event-based data model. This replaces the approach taken by Universal Analytics which uses a “cookie-based” approach to tracking users based on data from a cookie placed on the user’s browser.

As we know, from work with our own clients, this previous session-based data model results in data that is 80% correctly attributed due to the poor job that cookies do in ‘stitching’ together user journeys.

Take the above example, where only 20% of data was being correctly attributed using Google Analytics for marketing attribution.

However, by introducing the use of Machine Learning and AI in our Corvidae platform, we were able to significantly improve on this and achieve an incredible 85% accuracy – at that point in time. A figure that is up closer to 95% at this time.

The event-based model in GA4 is an improvement on this approach. Instead of tracking sessions it opens up the possibility to measure across different platforms and devices, which includes the use of data from Google Signals and first party data.

Improved support for user privacy

In line with the previous decision by Google to move away from 3rd party cookies for reasons of privacy, GA4 provides additional capability to disable ads personalisation. With extended functionality that allows this to be done within specific geographies.

It is also possible to exclude specific events that could be sensitive for the end user, to ensure you are not able to build audiences around those types of data points.

GA4 looks different

GA4 has moved away from what was a table heavy and complex layout. Removing many silos of tabs to combine data into a single overview with widgets – that you can drill down into for more detailed information to help you identify trends in your data.

GA4 user interface

What are the pros and cons of GA4?

So, those are the main differences between Google Universal and GA4 but what are the actual pros and cons of GA4 itself?

We take a look at both below.

Pros of GA4

GA4 provides a platform for both website and mobile app data

As its’ original name suggests – GA4 was launched in Beta as Google App and Web – the platform now provides you with the capability to combine web based and mobile app data into one, single property.

In a lot of ways this is just the kind of starter for ten you would expect. But this wasn’t something you could do in the Universal Analytics version of GA.

So, bringing these two data properties together means you have the ability to better track audiences across web and app platforms. To analyse user behaviour and carry out more advanced cross-platform attribution analysis.

Deeper analytics capabilities

GA4 closed funnel analysis report

GA4 has a new reporting section which is called Analysis. The focus of this is to enable you to drill down and create funnels on the fly for any page or event you need to see the background data on.

A toolset and functionality that has been ported across from the paid for GA360 platform.

Better integration with other Google products

GA4 also allows for better integration with other Google applications like Google Ads in particular. Which opens up the possibility of a deeper level of insight than was available within Universal Analytics.

Cons of GA4

So those are the pros for GA4 but what about the flipside? Are there any drawbacks that we can see in the solution.

Based on what we know to date we think there are number worth noting – as you explore whether the great new world promised by GA4 is right for you or not:

Both Google Universal and GA4 are illegal right now

This one is fairly fundamental and something that we continue to keep an eye on.

As things stand both Google Universal and GA4 are deemed to be illegal under data compliance laws in the EU.

This is on the back of an initial finding in Austria which found in favour of the Schrems II legal finding that Google Analytics was illegal under GDPR.

This was the result of work by privacy advocate Max Schrems who was had key court cases passed which are the foundation for privacy in the EU and UK – and is a decision which swiftly saw France following suit.

Google’s own response to the situation has been to:

  • declare that they are not going to attempt to make Google Universal not illegal and they have taken the decision to sunset the solution on July 1st 2023
  • they are, however, planning to make GA4 compliant by the end of this month (April 2022). So watch this space on that one.

The upshot of all of this is not only that both solutions are currently not compliant, but also that GA4 is the only ‘Google related’ analytics game in town after July next year.

Which means getting to know what it can and can’t do is crucial.

There is no data portability between Universal Analytics and GA4

A second potential drawback with GA4 is that Google have confirmed that they can’t unify data between the solutions. Which means there is no portability between the solutions as you move from one to the other.

For those of you who rely heavily on analytics views like year-on-year comparisons there is an element of ticking timebomb around ensuring that you are proactive in your approach to migration – or potentially a move away from GA if that is your preferred choice.

This is a topic that came up on a recent webinar where we were asked ‘what steps should I take now to get ahead of the GA4 change?’. Our Head of Paid Media, Malcolm Gibb, advised the following:

  • set up a Google sandbox at the moment which will allow you to play around with GA4 in an safe environment
  • if you don’t have a GA4 profile get used to how it operates. How it’s tracking. What it’s telling you. As well as understanding the fundamental changes to both the user interface and the measurement protocols
  • take note of the fact that the hierarchy is totally different. So, for example, we don’t have separate views in GA4. And the current hierarchies that we’re used to of filtered data and different profiles have gone away
  • you also need to take account of how you are going to measure like for like, and comparisons against historical Google Analytics Universal data

As a tangential issue here – it is also the ideal time to review your approach to making more use of under-utilized first party data. With cookies going away, cost per click is very likely to go up.

So, it’s a time to consider how you make the most of the first party data in your platforms and how you build strategies around that. You may already use first party data in terms of finding lookalikes on Facebook, for instance.

There’s a lot more that can be done there in terms of thinking about the logical segmentation of data and how you can apply it to Google Ads. In terms of search audiences, for example. And, also, on other platforms like Facebook and programmatic.

Retargeting as we know it is going away in GA4

Another of the big changes in GA4 is the move away from cookie-based retargeting.

So instead of creating user segments to build audiences like you would have done in Google Universal, GA4 relies heavily on Google Signals to make this happen.

Be sure to be ahead of issues around setting this functionality up properly – and revisiting your consent models to ensure they comply. If you are using server side-tagging then it is worth also noting this is not supported in GA4.

What you will be relying on for retargeting is the combination of Google Signals (essentially pulling on data from logged-in Chrome users) and also something called Google Topics which is Google’s partial replacement for cookies.

At this point in time Google is proposing a total of 350 topics which seems narrow.

To give a practical use case example here – let’s assume we are a specialist coffee company.

The closest topic that we would be able to associate our bid against in Chrome is not single origin coffee or speciality coffee. It’s not even coffee. It’s food and drink.

Not a close match we think you will agree.

Google Topics API example

It’s a steep learning curve ahead – and GA4 is in its infancy

One final point to make here is that there are a LOT of changes in the way that GA4 operates.

From the infrastructure and set up of the solution, to the way it captures data. To the way it enables you to analyze campaign performance. And it is a solution that is very much in its infancy with elements of the solution emerging incrementally. With many grey areas right now.

Which makes it an ideal time to look at potential alternatives to GA4 – alongside your assessment of GA4 itself.

What are the alternatives?

So, what are the alternatives to GA4?

Corvidae attribution tool

We know from painful experience with clients that the use of cookies or deterministic solutions, like Google Signals, delivers analytics data that ultimately is only about 20% accurate when you’re looking at a complete customer journey (which means that 80% of the data that’s being used by these more sophisticated systems offered in GA4 is incorrect).

And that will cause performance issues with GA4. Just in the same way that performance issues are being caused with third party cookie data today.

It was the realisation around this, and limitations with the previous version of Google Analytics, which set us out on our journey to develop a more accurate way to attribute across the user journey. Using AI and Machine Learning for effectively ‘stitching’ together customer journeys.

All of that learning is wrapped up and available within Corvidae – which is worth a look as you revaluate your options around analytics solutions.

Getting Started With Corvidae