After many years of loyal service, Google Analytics, as we've known it for years, will disappear in 2023. Google Analytics 4 (GA4) will become the standard as of July 1 next year. For everyone who was wondering: “When should I do something with GA4?” - You now have a date to work towards!
So the decision has finally been made: July 1, 2023, Google Analytics as we know it will “disappear”. From that date, it is no longer possible to create a Universal Analytics (UA) property. There will also be no more data collection in the existing UA properties. Of course, your data will not disappear on the night of June 30 to July 1: Google has indicated that the data will remain available in the UA properties for (at least) 6 months.
There is no doubt that the transition to GA4 will have consequences for the collection of data and insights. However, that does not mean that it will necessarily decrease. Otherwise, it will certainly be.
We could now dive into history and discuss the differences between UA and GA4. But then we still don't really know what you'll get from the new analytics as a business owner. That is why in this blog, I will mainly look at what the new Google Analytics 4 data model entails and what the possible consequences may be for your current implementation of Google Analytics. The implications will vary by webshop, but I especially want to give you more insight into the setup of Google Analytics 4.
While the name suggests that Google Analytics 4 is the same as the previous Google Analytics (or an upgrade from), this is far from the truth. This is because this has to do with the fact that GA4 is based on a completely different data model than UA.
Universal Analytics has a model based on sessions and web data. The web data is measured on four different scopes: hit, session, user and product. As a result, in many cases, it is not easy to make certain analyses. This is because not every type of data can be properly combined with each other.
Within GA4, it was decided to make all data “equal” to each other. That means: everything is a hit (event). Of course, this offers opportunities. In theory, you can compare everything, because the data is no longer written in different scopes.
This ensures that you have much more flexibility within the package to connect data as you see fit. This can lead to very good insights, but of course there is also a challenge. How do you know which event to connect with what to still get the right insight?
In Universal Analytics, you were still pretty much taken by Google. That's different in Google Analytics 4: you'll have to set up more yourself. It is therefore important to carefully review your current structure of events. First, the events will need to be clearly described: What is being measured? Secondly, the measurement must also work properly: Is what should be measured also measured?
In addition to the fact that GA4 is therefore fully event based, you also lose the hierarchy within these events. Event Category, Action, and Label, as the events are built in UA, are not available in GA4. The structure is completely different in GA4, so events will also need to be adapted for use in GA4.
To get a better idea of this, I outline an event below that measures which filter is clicked:
Universal Analytics
Google Analytics 4
Thanks to the layered structure in UA, you can quickly and easily see which filters are being used and which filter option is chosen. Unfortunately, this setup is therefore no longer possible within GA4, which is a challenge.
It is tempting to write out a standard GA4 setup that everyone can use. Only with a copy-paste setup, there is a chance that you will not take enough account of what the data in GA4 should now provide for insights. For you. For marketers. For everyone who works with the data from your Analytics environment. This brings me to the following question: what does your web data really say?
For example, does it make sense to continuously measure the use of your website navigation? Based on this data, can you tell if your menu works properly? It's nice that you can see how many times someone in the menu is on Sale clicks, but what's in it for you? Of course, it doesn't mean anything if you just look at how many people have viewed this page. Because if it turns out that people who went to the Sale via the menu convert 25% worse than the rest: are you going to remove that menu item? Of course not: after all, it's a logical place to navigate to a Sale category.
I want to make it clear that you often see that (over the years) - unnecessarily - many measuring points are being set up. Purely because it's possible. That doesn't mean everything that adds to your insights. In addition to using Google Analytics' hit limit (10 million hits per month) with many measuring points, it also creates a growing amount of data that you're not really going to do anything with. By the way, that hit limit will not disappear with the arrival of GA4. Too bad.
The transition to GA4 therefore offers the perfect opportunity to take a good look at your measurements in UA. After all, you can't just take over (most) part 1 on 1. By the way, this does not apply to (almost all) Ecommerce events (add to cart, purchase, etc.): these are almost the same in GA4.
It is especially important to have a good overview of all the events that are taking place. What do these events add? And they can be transferred to GA4 without reducing the value of the event. By value, I don't mean a number or amount that you can assign to an event in UA (Event Value), but what does it add to your insights? A valuable event makes it possible to obtain the desired insights, for example to provide insight into customer behavior or to steer accordingly. This includes the number of clicks on certain product filters, a visitor's checkout steps and, of course, a completed transaction.
In practice, this may mean that, in addition to the 'standard' Enhanced Ecommerce events, you have no other valuable events. GA4 also uses the well-known DataLayer, but the DataLayer will of course have to be complete and properly set up. Just like for UA. So it pays to invest time in the right DataLayer setup, then do half the work to measure as many side issues as possible. Without a good basis, those other measurements don't mean much anyway.
If you haven't clicked around in a GA4 property before: the look and feel of the reports is definitely different from what you're used to from UA. But it is precisely because of the new data model that you actually have much more freedom. After all, it offers the possibility to involve web analytics much more within your current data infrastructure. After all, you no longer have to comply with the, somewhat rigid, Google Analytics model. This means that, with the new form of events, it is much easier to link your data from Analytics to other systems or a data warehouse, for example with Google BigQuery. Indeed, GA4 already has a connection to BigQuery by default. The beauty of data, although it may not mean much on its own, is that when you combine it, it can provide golden insights!
Finally: we assume that the door Google said date of July 1, 2023 will no longer change. It may seem a bit far away, but before you know it, the time has come. We therefore recommend starting the migration to GA4's new data model as soon as possible with regard to all measurement points that are currently set up for UA. To limit data loss, it is smart to set up GA4 now. If you haven't set up GA4 next to the existing UA property yet. After all, both properties can function side by side (until July 1, 2023).
Do you want to know what the arrival of GA4 means for your website's data collection? Or could you use help devising your new data model? Don't hesitate to contact us!
