What you can do with Google Analytics 4

As many of you may already know, Google Analytics is a free web analytics tool provided by Google.

By using Google Analytics, those in charge at companies can obtain a wide range of insights about users who visit their websites—such as where they came from, what kind of users they are, and what actions they took. These insights can then be analyzed and applied to various business strategies and improvements.

GA4 and the previous Universal Analytics differ in how they collect user behavior data. Universal Analytics measures data on a “session” basis—tracking what users do from the moment they connect to the server until they leave. This allows you to understand which pages users visited and how long they stayed on them.

In contrast, GA4 measures data on an “event” basis, focusing on user interactions and experiences. For example, it can track how far a user scrolls on a page or how many seconds they watch an embedded video, enabling more detailed insights into user behavior within both web and app environments.

Additionally, GA4 was released as the official version of the App + Web property announced in 2019. Previously, web data was collected by Universal Analytics, while app data was handled separately through Google Analytics for Firebase. With GA4, these have been unified into a single platform.

As a result of these significant changes in both the scope and method of data measurement, it is now possible not only to capture more granular user behavior within a site, but also to track user activity across apps and between apps and websites.

GA4 enables the use of predictive metrics powered by Google’s machine learning technology. This makes it possible not only to analyze past and current data, but also to forecast future user behavior.

For example, it can predict metrics such as “purchase probability,” “churn probability,” and “revenue prediction.” These insights allow businesses to identify users who are more likely to convert, leading to more effective advertising and marketing strategies.

In line with the growing emphasis on privacy protection, GA4 is designed to handle user data more carefully. When storing user data, it is automatically deleted from servers after a certain period, which is fixed at either 2 months or 14 months. Compared to previous versions, this reflects a more cautious approach to the measurement and management of personal information.

Additionally, as collecting user data through cookies has come under scrutiny from a privacy perspective, GA4 enables a new approach to user behavior analysis by leveraging machine learning–based predictive features suited to the evolving data environment.

GA4 introduces many changes compared to Universal Analytics, meaning that users need to relearn its specifications from the ground up. However, as it is still a relatively new tool, there are not yet many comprehensive guides or resources available, which may make it more time-consuming to become proficient.

In addition, some features may still be incomplete, and new functionalities or changes may continue to be introduced. As a result, it is necessary to operate GA4 while adapting to an evolving set of features and information.

Google Search Console is a tool that collects data prior to a site visit—such as how often a site appears in search results and how frequently it is clicked—whereas Universal Analytics focuses on collecting user behavior data after a site visit.

In Universal Analytics, it was possible to link Search Console data and generate reports within the platform. However, at present, GA4 does not support this integration. Therefore, if you need reports that incorporate data collected via Search Console, you will need to use both GA4 and Search Console in parallel.

To utilize the machine learning–based predictive features mentioned earlier, the following conditions must be taken into account.

Source: Analytics Help “[GA4] Predictive Metrics”

To properly train predictive models, the following requirements must be met:

If you collect the purchase event, it is also necessary to collect the value and currency parameters, as these provide a more detailed understanding of the monetary value of the event.

Predictive metrics for each applicable model are generated once per day for each active user. If the model quality for a property falls below a minimum threshold, updates to the corresponding predictions may be automatically suspended, and the predictions may become unavailable in Analytics.

You can check the requirement status for each prediction in the [Predictive] section within the audience templates of the Audience Builder.

GA4 represents a significant shift from Universal Analytics in terms of how data is measured. This change enables the collection of more detailed and comprehensive user behavior data, making it better suited to the needs of the evolving data landscape.

However, as it is still a relatively new tool, some features available in Universal Analytics are not yet supported in GA4. In addition, ongoing updates and changes are expected, and there is still a limited amount of available information. As a result, some organizations may feel hesitant about adopting it.

That said, in order to take advantage of GA4’s advanced data collection and predictive capabilities as early as possible, it is advisable not to wait for the platform to be fully matured. Instead, a gradual transition—using GA4 alongside Universal Analytics—can be considered an effective approach.