8 Key Differences between GA360 and GA4

by Ricardo Cristofolini

There’s been a lot of talk about the new Google Analytics Property – GA4 (former App + Web) since its first release in Jul. 31st, 2019 and turned officially into GA4 later on. Tutorials, videos, webinars are all over the place with, at least, something new that was not discussed previously. That’s not surprising as Google is heavily working on this property and rolling out updates all the time and we know that it can be hard to keep up with all the changes and updates.

With that in mind, we, (I wrote this article with fellow Napkynian, Rob English) put together this article to explain in detail 8 of what we see as the key differences between GA360 and GA4 (future articles will feature more differences but we have to start somewhere!) This information is the result of the knowledge gained playing with the tool, implementing both GA4 and GA360  for clients, having internal discussions, testing, and endless research. 

1 New Data-Model

2 Scope Differences

Custom dimensions and metrics in Universal Analytics are used to add orthogonal pieces of information to collected data. In GA4, events or event parameters and user properties can serve this purpose. You can map your custom dimensions and metrics according to their scope:

3 Measurement and Session Tracking

Universal Analytics properties are based on Session-based data-model. That means Analytics groups data into sessions, and these sessions are the foundation of all reporting. During a session, Analytics collects and stores user interactions, such as pageviews, events, and eCommerce transactions, as hits. A single session can contain multiple hits, depending on how a user interacts with your website.

GA4 on the other hand uses a flexible event-based data model. Here, you can still see session data, but Analytics collects and stores user interactions with your website or app as events. Events can collect and send pieces of information that more fully specify the action the user took or add further context to the event or user. This information could include things like the value of purchase, the title of the page a user visited, or the geographic location of the user.

 For example:

  • Universal Analytics
    • Active user calculation
      • GUA relies on manual instrumentation (firing of an interactive event)
    • Session counting
      • A new campaign will start a new session regardless of activity
      • Hits are processed if they arrive within 4 hours of the close of the preceding day
  • GA4
    • Active user calculation
      • User activity is detected automatically in GA4. A user can launch an app and be considered an active user. This may lead to higher active user counts for this new property.
    • Session counting
      • A new campaign does not begin a new session. This may lead to lower session counts in the reports
      • Events are processed upon arrival up to 72 hours late

Although the Data-Model is different and session tracking is not as important as before, that still exists and a default inactivity of 30 minutes will start a “new session”. However, that will be translated into events.

When a user opens your site or app, an event called session_start will fire. After 30 minutes of inactivity in your site (or if a user sends the app to the background), another event called user_engagement will fire. Once the user resumes the usage, further events will start firing again.

4 Conversions Events (i.e. GA goal in GA360)

GA4, the most important events are called conversions when you enable it as a conversion. It’s a similar concept to goals in UA. It’s possible to enable up to 30 events per property as conversions, in addition to the five that Analytics defines by default:

  • first_open
  • in_app_purchase
  • app_store_subscription_convert
  • app_store_subscription_renew
  • purchase

Attribution information associated with an event is collected from the time of enablement forward, so start thinking about that as soon as possible.

There are ways to mitigate this using GA4s Create Event and Modify Event functionality in the interim, though optimally setting the data consistently at the source would be recommended, particularly for ensuring consistency not just in your GA4 data for both App and Web. That is really interesting as you don’t need a developer to implement everything in your website, but for consistency in any third party data integrations that might be occurring pre-GA in both App + Web.

You can find the following menu in the All Events area.

5 Modify Event

Modify events can be used to, as the name says, modify an event that has already been implemented. For example, if you have implemented in your website an event called ProductClick but in your app you have Product_Click, in GA4 you can fix that by modifying the event for a single event name.

6 Create Event

Create Event is another interesting thing GA4 is bringing to the table. The idea behind this is to give Marketers and Analysts the possibility to create their own events based on something already implemented. Goals are a great example of how to use this option.

Because GA4 is based on events (event-model) everything needs to be based on that. A Goal won’t be different. In that case, if we want Purchase as a Goal, all you have to do is create an event here that would, for example, check the Transaction_id.

Unfortunately, based on our testings, it can take a few hours for this event to show in the Realtime report and in your All Events area, but once it happens (assuming that all the configuration was done properly) you should see a new hit in the Network tab in your browser DevTools. The good side of this is that, once you have the event working, parameters configurations will be added to the event immediately, so no need to wait hours for it.

7 Reporting

Universal Analytics has limited cross-drive and cross-platform reporting. So, in order to track users that started on a mobile device, but ended up buying your product in your website, you would need a separate property to understand that those “two” different users are actually one.

Furthermore, in GUA properties, most reporting relies heavily on Device ID, although a few reports and features can also use the Google Signals identify space. When the user-ID feature is enabled, its data is reported separately from the rest of your data and doesn’t integrate with other identity spaces. Because those identity spaces work separately, it’s difficult to measure user journey across devices and de-duplicate users in UA properties.

GA4 on the other hand, process the data using all available identity spaces. First, Analytics looks for user-ID because this feature uses the data you collect. Next, it tries Google Signals, and finally, if there isn’t any match for either, it relies on the device ID (for apps) or client ID (for web). From there, Analytics creates a single user journey from all the data associated with the same identity and, because these identity spaces are used in all reports, they allow you to de-duplicate users and tell a more unified, holistic story about their relationship with your business.

8 Data Management & Automation

GA4 brings some new features related to Data management as well compared to Universal Analytics. Furthermore, it also has machine learning throughout to improve and simplify insights and discovery, different from GUA where automation was limited.

We’re going to end it here. In our next post, we’ll tackle Events and Parameters. 

Stay Tuned!

Ricardo Cristofolini

Implementation Specialist

I’m passionate about what I do. If you meet my manager or co-workers, they would say I’m a team player, engaged and always excited to learn something new. Like everyone else I have some flaws. However I’m not afraid to work around those to bring the best in myself and for the company.

See more posts from Ricardo