07 May 2023
10 major differences between UA and GA4
- Louder’s Google Analytics experts have joined forces to share the 10 most important things to know about GA4 for new users and users transitioning over from Universal Analytics (UA)
- Google Analytics 4 (GA4) was developed with consideration for rising privacy concerns, with a view to utilising machine learning to fill known measurement gaps.
- With the imminent sunset of Universal Analytics there’s a lot to gain from GA4, and it will only continue to get better as the product evolves going forward.
- Its important to understand the some fundamental differences when moving from UA to GA4
Lets get into it.
1. Data structure
In GA4, every interaction tracked on the page is considered an event. Aside from the ecommerce events, GA4 offers you the flexibility to use any custom event name and add in any custom dimension. This customisation makes GA4 reports easier to read and more tailored to your business specific nomenclature.
UA, event structuring on the other hand, was more rigid. Custom events had four parameters: category, action, label and value, with a nested tree like structure and specific governing rules around their use.
For example, event category and event action are required parameters for a UA custom event hit. If you wanted to associate an event with a custom dimension, you could but these dimensions and their values always felt a bit like a third wheel when it came to reporting.
However the good news is things are far different in GA4, there is no need to send a minimal number of parameters (you can send no parameters and just an event if you want), which eliminates the struggle to map your business specific context within this event model.
2. Metrics differences
Since the GA4 event structure is so different, some of the metrics and ways user engagement is measured has also changed when compared to UA. This makes apples to apples comparisons between UA and GA4 datasets challenging (more of an oranges and apples comparison). There are other structural changes at the account, property and views levels.
GA4 no longer has a concept of views and filters, instead GA4 uses data streams and whilst some data filter concepts remain in GA4, most UA filter types have been deprecated. As a result many common metric calculations have also been updated to reflect this change in GA4.
For example, filters in UA could stop or transform data from being collected in a view, whereas filters in GA4 are used to label data after collection. GA4 may also report a lower user count than UA, since it uses more accurate user identification methods and is able to de-duplicate same users across different environments.
Finally UA also had a concept of ‘interaction events’ which impacted bounce rate and time on site/page calculations. GA4 no longer has a concept of ‘interaction events’, but instead refers to engaged visitors.
You can read more about the Metric differences between UA and GA4.
3. User privacy
GA4 has a privacy first approach and is is designed to allow businesses to comply with privacy laws, including CCPA and GDPR. Consent mode in GA4 respects users’ data sharing choices via cookies and adjusts tag and tracking behaviour accordingly.
GA4 also offers data modelling, by using machine learning to fill in data gaps introduced by browsers blocking cookies (eg: via Apple’s ITP updates) or simply opting out of tracking via consent mode. This technology respects users’ privacy while still providing valuable insights for your business.
In contrast to this, UA was created at a time where privacy concerns were less prevalent and Apple ITP and Firefox ETP didn’t exist. As a result UA is not fit for the rapidly changing regulatory landscape as compliance with privacy laws and the shifting measurement landscape mean if you stick with UA going forward you will get less reliable data and less accurate insights.
4. Modelled data
GA4 uses machine learning to fill in data gaps where the data can’t be directly observed. This can be due to the tightening cookie policies and privacy related changes.
Modelled conversions can provide more accurate reporting when data is restricted so businesses aren’t left to guess where conversions came from. By modelling the behaviour of users who decline cookies based on the behaviour of similar users who accept cookies, you can better optimise your marketing campaigns.
5. Attribution modelling
GA4 now offers data driven attribution (DDA) as the standard for all users of the product whereas previously, DDA was only available to 360 (paying) customers. With DDA, Google is using AI-based algorithms to help you understand which marketing touch points are the most effective leading up to a conversion.
GA4’s event-based structure also gives users the flexibility to switch between attribution models as many times as they want without changing the underlying collected data. Changes will be reflected across all reports allowing a more meaningful comparison.
Contrary to this, UA only ever offered a last click model across the majority of it’s core reports, the only exception to this was the isolated Model Comparison Report (tool) which was an isolated report with limited ability to mix and match different reporting dimensions.
6. Product integrations
GA4 offers product linking with the Google Marketing Platforms which was not offered to users on the free version before. This means businesses that link GA4 with their GMP accounts can share data across platforms, unlocking additional insights and data sharing opportunities. Since GA4 will have a more holistic view of audiences, it will facilitate the creation of more granular audience lists for retargeting.
Currently available for linking to GA4 is Google Ads, SA360 and DV360 and CM360 is on the roadmap.
7. Ecommerce reporting
GA4 offers a detailed ecommerce reporting framework that will exceed UA’s offering, allowing for more advanced insights.
Currently GA4’s core ecommerce reports will look at bit sparse, especially if you’re used to the detailed Enhanced Ecommerce reports provided in UA. Therefore you’ll want to get familiar with GA4’s exploration reports which are similar to custom reports in UA but offer more advanced configuration options. These reports allow you drill down into specific ecommerce dimensions and metrics, offering improved funnel analysis and path explorations.
8. App+Web Reporting in a singular GA4 Property
GA4 was previously known as App+Web, however, this was changed to focus on unifying measurement across multiple platforms/environments (via streams) into a singular property type for your business.
If your business had a website, an Android App or an iOS App, environment tracking was split into it’s own silo in the older UA products. This meant that users who engage with your business via different platforms were counted distinctly in each environment.
GA4 seeks to unify user engagement (de-duplicate distinct user engagements) with your business in multiple contexts, devices and environments, giving you more meaningful user lifetime value insights and cross-device campaign attribution abilities.
9. Enhanced conversions
The last few years has seen a degradation in campaign attribution abilities in the measurement industry due to factors such as Safari ITP and Firefox ETP. Recently the Google’s Ads product launched Enhanced Conversions to help restore campaign attribution and look back windows, but you were forced to use the native Google Ads pixel in order to leverage this new measurement ability.
However in the future GA4 will also support enhanced conversions, and with product integrations such as CM360, DV360, SA360 and Google Ads, you’ll be able to import GA4 based enhanced conversions into these respective products for a more complete picture of campaign attribution and effectiveness.
This enhanced form of measurement will not be available to UA.
10. Measurement protocol
GA4, like UA, has a measurement protocol (MP), which is designed to accept hit data from contexts outside of a web-page. An example may be an offline sales system or in person sale.
GA4’s MP has changed slightly to support more use cases outside of web tracking. Some of these changes include significant improvements to an open session window to accept additional hits, which has increased from 30 minutes up to 72 hours.
The new GA4 MP has a new schema meaning you’ll need to create a new measurement stream, an API key and token to use it. Changes to how GA4’s MP works may impact your existing product integrations and functionality which is likely to require some adjustments from your UA set up, and third party tools that send data back to GA4 via the MP.
Need help with GA? Get in touch with Louder to understand how these benefits or changes will impact your business.