What is a Session in Google Analytics 4?

If you've recently moved over to Google Analytics 4, you've probably noticed that the term "session" is still central to many reports. Yet, how GA4 counts a session is fundamentally different from how its predecessor, Universal Analytics, did it. This change isn't just a small tweak, it reshapes how you analyze user engagement, traffic sources, and overall website performance.

This article will break down exactly what a session is in GA4, how it's measured, the key differences from the old way of doing things, and why this shift gives you a much clearer picture of the user journey.

How GA4 Defines and Calculates a Session

In Universal Analytics (UA), a session was a collection of "hits" - like pageviews, events, or transactions - that a user made on your site. In GA4, everything is an event, and a session is simply a group of events a user triggers within a specific timeframe.

Each session in GA4 is identified by two key parameters that are automatically collected:

• Session ID (ga_session_id): This is a unique timestamp that identifies the specific session. Every event that occurs during that session (like a page_view, scroll, or add_to_cart) is stamped with this same ID, grouping them together.
• Session Number (ga_session_number): This counts how many sessions a specific user has had. A new visitor's first session will be "1," their second will be "2," and so on.

A new session begins automatically when a user visits your site or app, and there isn't another active session for them. This triggers a special event called session_start. It’s this event that logs the session ID and number, effectively kicking off the measurement period.

When Does a GA4 Session End?

This is where one of the biggest changes from Universal Analytics comes into play. In GA4, a session ends only one way: when 30 minutes of inactivity have passed.

That's it. This period of inactivity is called the "session timeout." If a user is on your website, moves their mouse, scrolls down a page, or navigates to a new page, they are triggering events that reset the 30-minute timeout clock. If they walk away from their computer for 31 minutes and then come back to click on something in the same browser tab, GA4 will start a new session because the previous one has timed out.

How to Adjust the Session Timeout in GA4

The default 30-minute timeout is fine for most websites, but you may want to change it. For example, if you have a content-heavy media site where users might read long articles or a complex application form where they might spend longer than 30 minutes on a single page, you might want to increase the timeout period.

Here’s how to adjust it:

1. Navigate to the Admin section of your GA4 property (the gear icon on the bottom-left).
2. Under the Property column, click on Data Streams.
3. Select the data stream you want to modify (e.g., your website).
4. Scroll down and click on Configure tag settings.
5. On the settings screen, click Show more if needed.
6. Click on Adjust session timeout.
7. Set your desired timeout duration (e.g., 45 minutes, 1 hour, etc.) and save your changes.

Keep in mind that changing this will affect how your data is collected from that point forward, so it's a good idea to consider your user behavior carefully before making a change.

GA4 Sessions vs. Universal Analytics: The Key Differences

For anyone used to Universal Analytics, the changes in GA4 can make metrics feel inconsistent. If you’ve ever compared your session data between the two platforms, you’ve likely noticed that GA4 reports fewer sessions. This discrepancy is intentional and a direct result of two legacy session-ending rules that GA4 has abandoned.

1. No More Midnight Resets

In Universal Analytics, a session would automatically end and a new one would start at the stroke of midnight, based on the time zone set in your view settings. For example, if someone started browsing your ecommerce site at 11:50 PM on Tuesday and made a purchase at 12:05 AM on Wednesday, UA would record that as two separate sessions.

GA4 eliminates this rule completely. A session is now a continuous period of activity, regardless of the time of day. In that same scenario, GA4 would correctly see it as a single, uninterrupted session that started on Tuesday and finished on Wednesday. This gives you a more logical and accurate view of the user’s true path to purchase.

2. New Campaign Sources No Longer Start a New Session

This was another significant factor in why UA often reported inflated session counts. In Universal Analytics, if a user came to your site through one campaign (e.g., a Google Ad), left the tab open for a few minutes, then came back through a different campaign (e.g., an email marketing link), UA would end the first session and start a new one as soon as the campaign parameters changed.

GA4 does not do this. As long as the user returns within the determined timeout period (default: 30 minutes), it remains the same session. The engagement is considered continuous, even if the touchpoint is different.

Example:

• 10:00 AM: A user clicks a Facebook Ad and lands on your blog.
• 10:15 AM: The user leaves to check email without closing the browser tab. In an email newsletter, they see a link to a different product on your site and click it.
• Universal Analytics's interpretation: Two sessions. One from the social source, and a new one from the email source.
• GA4's interpretation: One session. The user’s interaction was continuous, just with two different touchpoints.

This provides a much more intuitive view of the customer journey, preventing a single browsing period from being artificially fragmented into multiple sessions.

Meet the New Session-Based Metrics in GA4

Alongside the new definition of a session, GA4 introduced a new family of metrics focused on engagement. These metrics are designed to help you understand session quality - not just session quantity.

Engaged Sessions

An "engaged session" is the foundation of engagement metrics in GA4. A session is counted as "engaged" if the user does any one of the following:

• Lasts longer than 10 seconds (you can adjust this timer).
• Triggers a conversion event (like purchase or form_submit).
• Has at least two pageviews or screenviews.

This serves as the primary way GA4 helps you filter out low-quality traffic. If someone lands on your site and leaves after a few seconds without taking any action, that session is not considered engaged.

Engagement Rate

This is arguably one of the most important metrics in GA4. Engagement Rate is the percentage of sessions that were engaged sessions: (Engaged Sessions / Total Sessions).

It’s essentially the opposite of Bounce Rate. Instead of measuring how many users leave immediately, Engagement Rate measures what proportion of your visitors actually stuck around and interacted. Higher is better.

Engaged Sessions per User

This metric shows the average number of engaged sessions per user. It aims to give you an idea of how frequently your users are interacting with your site or app.

Average Engagement Time

This is GA4’s modern replacement for Universal Analytics' Average Session Duration. It measures the period of time that your app was in the foreground, or your web page was in focus for users. Unlike session duration, it ignores periods of inactivity.

Final Thoughts

Understanding sessions in GA4 requires a fundamental shift in perspective compared to Universal Analytics. While the number of sessions may be fewer in GA4 due to the removal of traditional session-ending rules, the new approach emphasizes a more accurate and cohesive view of the user journey. As you begin to get comfortable with GA4, remember that it’s centered around the concept of engaged sessions, influenced by meaningful interactions. Leveraging these insights will help you create meaningful posts and gain deeper insights into your audience's behavior.

Once you're ready to analyze and manage your data in a more organized way, consider testing Graphed that connects all your analytics into a single dashboard, making reporting and data visualization easier and more comprehensive.