Published on February 01, 2021 Last modified on April 28, 2022
Estimated reading time: 9 minutes
The new Google Analytics 4 (often shortened to GA4) is ‘new’ in several ways. Not only is it an upgraded version of the old Analytics (called Universal Analytics or UA) with a fairly new dashboard, but it also comes with new metrics, a new system of reporting, and an altogether new perspective on website analytics and optimization.
GA4 was officially launched in October 2020, but it has sort of been around since 2019 in the form of “App+Web”, a beta property in the old Analytics. It was then rebranded and introduced as the ‘next generation’ of Google Analytics to cater to all businesses having either a website or an app only, and not just those that have both.
Key differences between Universal Analytics and Google Analytics 4
If you already have a Universal Analytics implementation, you’ll find that there are quite a lot of differences between it and GA4.
GA4 introduces a fundamental shift in how you measure website activity—to gain a better understanding of how exactly this impacts your business, it’s worth exploring what changed in this new version from the traditional Analytics model that we’ve been used to.
UA is for websites only, GA4 is for both websites and apps
UA and GA4 are both property types in the overall Google Analytics product. A ‘property’ is basically a sub-component of an Analytics account used to organize reports and data. Since a UA property only supports websites, businesses with apps previously used Google Analytics for Firebase for a separate tracking.
GA4, on the other hand, supports websites, apps, or an integrated report for both. This makes way for a more streamlined view of multi-channel data, as well as a more holistic understanding of user behavior in different platforms.
Integrated ‘identity spaces’
Speaking of a more holistic understanding of user behavior, one of the most relatively interesting features of GA4 is the integration of audience data from different sources.
GA4 introduces the concept of “identity spaces”, a group of identifiers used to gain more understanding about your users. Identity spaces can vary from basic information such as:
the user’s device ID;
your own provided user ID (For example, if your site or app has a login feature, then you can better track logged in users as they browse and use your website or app; make sure, of course, that you have their consent to share their login data with Analytics for tracking.); and
Google Signals, or data from users who are signed in to their Google accounts and have turned on Ads Personalization in their account settings.
Previously in UA, only the device ID was primarily used in gathering user information. While you have the option to upload first-party user ID data, the report for this audience set is not merged with the rest of the Analytics reports, which may have resulted in unknowingly duplicate user data.
GA4, however, utilizes all available identity spaces when processing data—first checking if a user ID is available, then turning to Google Signals, and as a last resort checks the device ID, in order to ensure that similar users are deduplicated, i.e., only recorded as one across all Analytics reports. This means that a user who browses your website then uses your app later will be treated as one despite being in different devices (provided that there are enough identity spaces available to actually identify them).
This also benefits your Audience lists; for instance, let’s say that you exclude people who have already converted or purchased something from your retargeting audience; if a user purchases something through the app, he will be taken off from your retargeting list. Before GA4, there’s a chance that this same user may be in the web audience as well and continue to be retargeted with ads for products that he already bought. With GA4’s integrated identity spaces, the user will be taken off the list completely, whatever channel or device he uses in interacting with your business.
Session-based vs event-based models
In the last article about Google Consent Mode, we already talked about how measurement and conversion attribution has suffered from gaps in data due to strict regulations on third-party cookies. Without cookies, Universal Analytics’ session-based tracking, where website actions are grouped together within a certain time-frame, is just not as effective anymore.
In contrast, GA4 operates on an event-based data model, where the focus is placed on individual events on a website (i.e., page views, button clicks, scrolls, submissions, etc.) so you can better track the user actions that matter the most for your business.
As we’re moving away from reliance on cookies for tracking, Google is putting emphasis on higher-level analysis by utilizing AI to deliver actionable insights.
Some of the useful reports that you can find in GA4 are trends in users’ buying behavior and conversions (e.g., increased demand for certain products, fluctuations in online inquiries, etc.), potential revenue by using past data, which types of audiences are more likely to convert, and more. These insights are based on aggregate and non-identifiable data that they collect, paired with machine learning for modelling to fill in the gaps in traditional Analytics because of blocked cookies.
Ease of use
GA4 is almost an overhaul of the reports that marketers and analysts are familiar with in UA, so like any new technology, there is still a learning curve involved.
However, it seems that GA4 is ultimately the easier version to use. First, it enables codeless event tracking, which means that Analytics itself automatically tracks basic events without the need for manual setup on your end. It also presents a simplified style of reporting, providing business owners with smart insights that matter instead of a long list of different reports corresponding to different metrics.
Emphasis on privacy
In GA4, it’s easy to have granular control over data, allowing for easier compliance with relevant data privacy laws. There’s also an intuitive data deletion request in case a data subject requests for their data to be deleted in all of Google’s Analytics records.
With the shifts in how GA4 measures data, it’s also better to shift our perspective from the traditional data collection and reporting we’ve been used to in Universal Analytics. Marketing is definitely changing and heading for a cookie-less future, and GA4 is another nudge for marketers and business owners alike to adapt a perspective that doesn’t rely so much on user analysis based on personal information.
Google anticipates that “data sparsity will become the new norm”, in that there will inevitably be gaps in the actual data we collect as data privacy continues to be at the forefront of the online ecosystem. We’ll then have to utilize machine learning and AI to fill in the gaps in the reporting, and maybe even provide us smarter insights in the process than we ourselves can come up with in the traditional analytics model.
How to implement GA4
You don’t have to completely abandon your existing Universal Analytics setup to start using GA4. The good news is, Google lets you run parallel Analytics properties by setting up both. (Note that GA4 is the default property type when setting up new websites. However, you still have the option to set up a UA property.)