With the announcement that Universal Analytics will be discontinued in July 2023 and that GA4 will be the only web analytics tool from that point on, questions and developments have accelerated. This article reviews the biggest differences between the two products to help you prepare for working with the new Google Analytics.

Insights into Your Data

Many marketers are familiar with the Google Analytics setup of properties and views. In Universal Analytics, different products of a company are separated into properties: do you have an app and a website? Then you probably have one property for your app and one for your website. These properties are then subdivided into views; your master view with filters and goals, one unfiltered view as a backup of raw data, and extra views for test environments or acceptance stages.

Google Analytics 4 no longer includes views. You can also merge different properties by measuring apps and websites in a single property. This change stems from the new data model, which operates based on events (see below). These events can be sent from apps and websites using the same framework to GA4.

Data Views and Filters

Since views no longer exist, settings that you previously applied at the view level have shifted to the property level. This includes rules to exclude internal traffic, filter bots, or set up cross-domain tracking.

Cookies, Users, and Privacy

Universal Analytics collects data about website usage via cookie-based tracking. When you install Universal Analytics by embedding the JavaScript into the website, cookies are placed in the visitor’s browser. These cookies enable the platform to monitor user activity during a session. Universal Analytics’ session-based data model is built on this foundation: all interactions (the hits) are summarized in a session.

According to Google, GA4 allows businesses to track users “across different platforms and devices through various forms of identity.” This includes Google Signals, a platform used by users who have opted in for ad personalization. During Google I/O on May 11th, Google announced that they would give consumers even more control over the ads they see, promoting consumer empowerment and encouraging opt-in. In GA4, sessions are no longer measured; instead, it uses an event-based data model. More about this in the next section.

GA4 still uses cookies for tracking where possible, but these are expected to become less important in the coming years. The platform is promoted as privacy-focused and designed to work with or without cookies. Google is building on its knowledge of Machine Learning and data modeling techniques such as FLoC. These techniques for GA4 are still under development and will receive numerous updates in the coming years.

It is clear that we are moving toward a cookieless environment, especially with developments from Apple with iOS 14. Google aims to align GA4 with these changes to keep marketers engaged with Google services. For web visitors, these developments mean more control and privacy-centered solutions. Despite the challenges this presents for marketers, many opportunities will be developed over the coming years.

New Data Model

As much as Google would like to present GA4 as a new version of Google Analytics, the fact remains that Google Analytics 4 is fundamentally a new product. While Universal Analytics is an evolution of Urchin Analytics (acquired by Google in 2005), the new Google Analytics 4 is based on the Firebase platform, purchased by Google in 2014.

The biggest difference between the two platforms is the data model they use. Universal Analytics uses a session- and pageview-based data model. This means that hits are grouped into sessions, which form the foundation of the data model. Sessions are a collection of interactions (hits) on a website that occur within a specific time frame and can include different pageviews, events, or e-commerce transactions.

Google Analytics 4 uses a data model that aligns more closely with standard practices in web analytics: all interactions are measured as events. Events provide insights into what happens on the website, based on the event name and the parameters sent along with them.

Event Categories

Universal Analytics also offers the option to track events in addition to pageviews or e-commerce hits. Events in Universal Analytics are a unique hit type and always include a category, action, and label (sometimes with a value). In GA4, categories, actions, and labels no longer exist. Every hit is an event and can include parameters to describe the event (although this is not mandatory).

Example: In GA4, there is a standard event called page_view. This event includes parameters such as page_location (the page URL), page_referrer (URL of the previous page), and page_title (title of the page).

Event Categories in GA4

Events in GA4 can be divided into four categories. The first two categories can only be configured via GA4, as they are built into the platform and are part of the JavaScript that loads when you install GA4 on your website. You can find a complete list of automatically loaded events here. Recommended and custom events (categories 3 and 4) are best set up via Google Tag Manager and sent to GA4.

1. Automatic Events: These are automatically collected by GA4 once the code is installed on the website. These include events such as page_view, first_visit, and session_start.

2. Enhanced Measurement Events: These events are also automatically collected by GA4 but can be turned on or off based on the website's functionalities. These include events such as scroll depth, outbound clicks, search queries, and video engagement.

3. Recommended Events: Google suggests these events based on the industry the website operates in. In addition to general events for all websites (such as login, content sharing, account creation), there are specific events for e-commerce websites and games. The primary reason these are categorized as "recommended" is that Google has already established standards for these events. Although you may use a different standard, it is advisable to follow Google's guidelines. Doing so allows the machine learning algorithms running in the background to automatically recognize certain events, saving time in the future.

4. Custom Events: These events are fully customizable and can track unique interactions specific to your website. It is important to note that there is a maximum of 500 unique named events per account, which is usually more than enough for most accounts. However, you cannot delete events once they are set up, so it's essential to plan carefully which events you expect to track in your account.

No More Hit Limits

Due to the setup of Universal Analytics, most websites using the free version are limited to 10 million hits per month, which includes both pageviews and events. This limit has been removed in GA4.

As mentioned above, the limits in GA4 are now based on the number of different events that can be collected (500). At the time of writing, there are no limits set on the volume of events, meaning you can collect more than 10 million hits per month. Each event can have up to 25 parameters.

BigQuery Integration

One of the major differences between Universal Analytics and GA4 is the change Google made regarding BigQuery integration. In GA4, this is available to all users, whereas it was only accessible to 360 customers in Universal Analytics.

Not familiar with BigQuery? Then you're likely familiar with sampling issues in Universal Analytics. When trying to view trends over a long period or working with large datasets, load times can become extremely long. To keep the product functional, data is sampled so that the entire dataset does not need to be loaded. The downside is that trends in sampled data often don't match the actual data.

BigQuery runs on the Google Cloud Platform and pulls data from GA4. The product is designed to process very large and complex datasets, allowing for better analysis and insights. Google understands the importance of datasets and is eager to offer the capabilities of its Cloud platform and BigQuery. While GA4 and BigQuery integration are free, running BigQuery on the Google Cloud Platform is a paid service.

Ready for the Future

It's clear that there are significant differences between the two products. This means that both marketers and data specialists need to critically assess their current web analytics setup. Creating a detailed measurement plan before starting the migration is a crucial first step.

The transition provides marketers and organizations the opportunity to assess how web analytics is being used: Who is using the analytics, and for what purposes? Is everything being tracked correctly, or is too much data being stored without a clear purpose? Although there will be challenges in the coming years, understanding and explaining the differences between Universal Analytics and GA4 is the first step.

If you have any further questions about using your web analytics, contact the Gracious marketing team, and we’ll be happy to assist.

About the Author

Frank is a CRO specialist at full-service marketing agency Gracious. With a background in IT, he has delved into CRO and online personalization in recent years. He guides clients in conversion optimization, online personalization, and all the technical challenges that come with tracking and data management.