The Bounce Rate in Google Analytics 4 is Vastly Different from Universal Analytics

If we look closely, we can see a big difference in the bounce rate between GA4 and the UA version of Google Analytics, with GA4 often having a significantly lower bounce rate than the UA version. Do we understand why there is such a significant gap between the two? We are prone to make wrong decisions if we fail to appreciate the differences between them and continue to make decisions based on our previous understanding. As a result, let’s look closely at what distinguishes the bounce rates in these two versions today.

The Difference in Sessions between UA and GA4

Before delving into bounce rate variances, let’s first analyze session disparities, as sessions are directly tied to bounce rate.

A session is an important indicator in website analytics that is used to track user activity on a website or application. Different analytics platforms, such as Universal Analytics (UA) and Google Analytics 4 (GA4), define and monitor sessions in different ways.

In Universal Analytics (UA), a session expires after 30 minutes of inactivity, at midnight, or when a new campaign parameter is encountered, and then restarts. However, in Google Analytics 4 (GA4), a session ends after more than 30 minutes of inactivity and does not automatically restart at midnight or when new campaign parameters are encountered. Furthermore, GA4 uses statistical estimation to quantify the number of sessions to improve accuracy and reduce mistake rates, a process lacking in UA.

FeaturesUniversal Analytics (UA)Google Analytics 4 (GA4)
session definitionThe period when a user actively interacts with a website or appThe period when a user actively interacts with a website or app
session endedEnds when there is no activity for 30 minutes, at midnight, or new activity parameters are encountered.It will end when there is no activity for more than 30 minutes. It will not end automatically at midnight or when new activity parameters are encountered.
Session startsStart a new session after session timeout, at midnight, or when new activity parameters are encounteredOnly start a new session after session timeout
Tracking methodDetermine the start and end of the session using defined parametersGenerate a session ID from the session_start event and associate this ID with subsequent events in the session

If you want to know more about the difference between GA4 and UA Google Analytics definitions, you can Click here.

The factors impacting session counts differ between Universal Analytics (UA) and Google Analytics 4 (GA4), which has a direct impact on data analysis accuracy and interpretation. Here are some in-depth explanations of some significant factors:

Geographical location and time zone

For global businesses, users in different regions may cross midnight, causing session resets in UA. In GA4, as sessions do not automatically reset at midnight, the impact of users’ time zones on session counts is less significant.

Use of UTM Parameters

If UTM (Urchin Tracking Module) tags are used on a website or app in UA, the session is reset with each new UTM tag, potentially resulting in falsely boosted session numbers. GA4 is more adaptable in this aspect, as it does not immediately reset sessions due to updated UTM tags.

Data filtering and configuration

Both UA and GA4 support data filtering, but the configuration of filtering rules can result in different session counts. Certain sorts of filters, for example, may exclude specific user groups or traffic categories, influencing session counting. Because the influence of filter settings on session counts differs between GA4 and UA, even the same filtering rules can result in different results.

Estimation Method for Session Counts

GA4 calculates session counts using a statistical estimation method. By calculating the amount of unique session IDs, this approach seeks to deliver more accurate session counts. UA does not employ an estimating approach and instead counts sessions depending on user behavior, which can result in under- or over-counting in various circumstances.

User behavior and interaction patterns

Session counts is also affected by how users engage with a website or app. Frequent page refreshes, navigation patterns between different pages, and so on can all result in varying session counts between the two platforms.

Understanding these elements is essential for accurately understanding data on various analytics platforms. Each of these factors can result in significant changes in session counting between UA and GA4, influencing data analysis and business decisions.

Sessions with Participants

Engaged session

Let’s propose another phrase after understanding sessions: Engaged Sessions.

“Engaged Session” is a new concept introduced in Google Analytics 4 (GA4) to better assess user activity with a website or application. This concept is exclusive to GA4 and is not present in Universal Analytics (UA). Here’s an overview of Engaged Sessions:

An “Engaged Session” is one in which the user participates and interacts significantly when browsing a website or app. Such sessions often reflect a high level of user interest in the website’s or application’s content.

Criteria for Making a Decision

Engaged Sessions in GA4 are often determined by the following criteria:

Session Duration: When a user’s session surpasses a specified threshold (for example, 10 seconds).

User Interaction: If the user engages in particular interactive behaviors like as clicking, surfing, filling out forms, and so on during the session.

Page or Screen Views: When a user views many pages or screens in the same session.

Importance

Measuring User involvement: Engaged Sessions are a more accurate indicator for determining a website’s or app’s true amount of user involvement.

Improving Data Quality: By focusing on Engaged Sessions, it is simpler to discover which content or features attract consumers, allowing for better user experience.

Marketing and Business Decisions: Data from Engaged Sessions can assist marketers and business owners develop more effective marketing strategies and business decisions.

In comparison to UA

In UA, a session is defined primarily by the user’s visit time and page views, with little distinction made for the degree of user interaction. In comparison, GA4’s Engaged Sessions approach provides a more in-depth degree of measurement for user involvement.

In conclusion, Engaged Sessions are an important feature in GA4, assisting website owners in more correctly understanding and analyzing user behavior, ultimately enhancing the performance and user experience of websites and applications.

Differences Between Bounce Rates in UA and GA4

Understanding the concepts of sessions and engaged sessions simplifies the concept of bounce rate.

The definition and computation of Bounce Rate varies significantly between Universal Analytics (UA) and Google Analytics 4 (GA4). These distinctions are critical for comprehending and assessing user involvement on websites. The following table compares the bounce rates on both platforms:

Universal Analytics (UA) Bounce Rate

The bounce rate is the percentage of visitors who depart a website after just seeing one page. It is a proportion of all sessions to single-page sessions.

Bounce Rate = Number of Single-Page Sessions / Total Number of Sessions.

The bounce rate is an important measure in UA for analyzing a website’s initial attraction and relevance.

Google Analytics 4 (GA4) Bounce Rate

The idea of bounce rate has been replaced in GA4 by the new statistic “Engagement Rate.”

The engagement rate is defined as the proportion of sessions in which users interact with the website. This covers sessions that last more than 10 seconds, sessions that include event interactions, and sessions that have two or more page visits.

Engagement Rate = Number of Engaged Sessions / Total Number of Sessions.

The engagement rate, rather than the number of pages visited, focuses on overall user involvement and interaction with the website.

The Primary Distinctions

Different Metrics: UA measures single-page visits with the bounce rate, whereas GA4 measures overall user involvement with the engagement rate.

Measurement Focus: UA’s bounce rate focuses on the page’s initial attractiveness, whereas GA4’s engagement rate focuses on overall user involvement and interaction.

User Behavior Reflection: GA4’s engagement rate more accurately reflects user behavior on the website, not only whether they visited many pages.

I think everyone now knows the variations in bounce rates between UA and GA4. Finally, let us use an example to demonstrate these distinctions and their implications for business.

MetricsUniversal Analytics (UA)Google Analytics 4 (GA4)
user11
session11
Bounce/interested conversationsBounce = 1Interested sessions = 1
Bounce Rate100%0%
session duration0 (no other interactions)>0 (user stayed for more than 10 seconds)

As an illustration, consider Universal Analytics (UA)

A user enters a shopping website and views a special promotion product but does not click any other links or interact with the page before closing it. Because no further interactions occur, this visit is classified as a bounce in UA.

Bounce Rate = 1/1 = 100% 

Users = 1 

Sessions = 1

Bounces = 1 

Bounce Rate = 100%

Session Duration = 0 (since no other interactions occur)

In Google Analytics 4 (GA4), for example

In GA4, the same visiting behavior will be logged as an engaged session if the user stays on the product page for more than 10 seconds without hitting any other links.

Active Users = 1 

Sessions = 1 

Engaged Sessions = 1 (if the user spends more than 10 seconds on the product page)

Engagement rate =100%

Engagement Time > 0 (time spent on the product page by the user)

Because the user does not engage in any activity other than viewing the site, the visit is recorded as a bounce with a 100% bounce rate in UA. In GA4, however, the user’s time spent is counted engagement even if they do not engage in any more page views or click activities, resulting in a 100% engagement rate and a 0% bounce rate. This demonstrates GA4’s granularity and flexibility in evaluating user interaction.

Scroll to Top