Bounce Rate vs Engagement Rate in GA4: What Changed

Quick answer

In GA4, engagement rate is the percentage of sessions that were engaged: they lasted longer than 10 seconds, included a key event, or included at least 2 page views. Bounce rate is simply the opposite, the percentage of sessions that were not engaged. It is not the same metric as Universal Analytics bounce rate, which counted single-page sessions. A session where someone reads one page for 3 minutes was a bounce in UA but counts as engaged in GA4.

If you migrated from Universal Analytics and your bounce rate suddenly looked amazing, GA4 didn’t fix your website. It changed the definition.

Bounce rate in GA4 is a different metric that happens to share a name with the old one, and comparing the two is one of the most common reporting mistakes marketers make. This guide covers what an engaged session actually is, how GA4 calculates both metrics, how to get bounce rate back into your reports, and which number to put in front of your boss.

The short version

Engagement rate = engaged sessions divided by total sessions.
Bounce rate = the percentage of sessions that were not engaged. The two always add up to 100 percent.
An engaged session is any session that clears at least one of three bars: longer than 10 seconds, at least one key event, or 2 or more page views.

That third bullet is the whole story. Everything else follows from it.

What counts as an engaged session

GA4 checks every session against three conditions:

It lasted longer than 10 seconds. Not 10 seconds on one page, 10 seconds of the session being in the foreground.
It included a key event. A key event is an event you’ve marked as important, like a form submission or purchase. (Google renamed conversions to key events in 2024; here’s our guide to setting up key events in GA4.)
It included 2 or more page views (or screen views, for apps).

This is OR logic, not AND. A visitor who lands on your pricing page, reads for 15 seconds, and leaves is engaged (condition 1). A visitor who submits your contact form 8 seconds after arriving is engaged (condition 2). A visitor who clicks straight from your homepage to your blog and closes the tab is engaged (condition 3).

The only sessions that bounce are the ones that fail all three: under 10 seconds, no key event, one page. Think of a bounce in GA4 as “arrived and immediately left without doing anything.”

Why Google changed it

Universal Analytics defined a bounce as a single-page session with no interaction events. That definition punished exactly the sites that were doing content well.

Picture a searcher who lands on your recipe, reads the whole thing, cooks dinner happy, and closes the tab. In UA that was a bounce, indistinguishable from someone who clicked in by accident and left in one second. Blogs lived with 70 to 80 percent bounce rates that said almost nothing about content quality.

GA4’s definition measures attention instead of navigation. Ten seconds isn’t a high bar, but it reliably separates “actually looked at the page” from “left immediately,” which is what most marketers wanted bounce rate to mean all along.

The flip side: GA4 bounce rates are much lower than UA bounce rates for the same site and the same traffic. If you keep a year-over-year report that spans the UA-to-GA4 migration, annotate it. The drop is definitional, not a win. For more of these renamed-and-redefined traps, see GA4 vs Universal Analytics: what changed.

Where to find engagement rate in GA4

Engagement rate shows up by default in the acquisition reports:

Go to Reports > Acquisition > Traffic acquisition.
Scroll the table to the right. You’ll see Engaged sessions, Engagement rate, and Engaged sessions per active user alongside your session counts.

The same metrics appear in User acquisition and can be added to most other reports. GA4 also leans on engagement in related metrics you’ll see nearby:

Engaged sessions per active user: how many engaged sessions the average active user had.
Average engagement time: how long your site was in the foreground, averaged per active user. This replaced UA’s “average session duration,” and like bounce rate, it’s calculated differently, so don’t compare the two.

How to add bounce rate back to your reports

Bounce rate exists in GA4, it’s just not on the default reports. Two ways to see it:

Customize a standard report. Open the report, click the pencil icon in the top right (you need Editor or Administrator access on the property), choose Metrics, add Bounce rate, and save. It becomes a permanent column for everyone on the property.

Use an Exploration. In Explore, bounce rate is available as a metric in free-form tables without touching the shared reports.

Since bounce rate is always 100 percent minus engagement rate, adding it is purely cosmetic. Some teams find “26 percent bounced” easier to react to than “74 percent engaged.” Use whichever framing lands, but don’t report both; it’s the same number wearing two outfits.

How to actually use these metrics

The absolute number matters less than the comparisons. Three that pay off:

By channel. In Traffic acquisition, compare engagement rate across organic, paid, social, and email. A paid campaign with a dramatically lower engagement rate than your other channels is buying clicks that don’t stick, which is a targeting or landing-page problem you can fix. Our Traffic acquisition walkthrough covers reading this report in more depth.

By landing page. Add landing page as a dimension and look for pages well below your site average. Those are your candidates for rewriting intros, fixing page speed, or matching the page to the search intent that brings people in.

Over time. A slow slide in engagement rate on stable traffic usually means your traffic mix is shifting toward lower-intent sources, or a slow site got slower. It’s an early-warning metric more than a KPI.

One caution: engagement rate is easy to game and easy to fool. Consent banners that delay tracking, single-page checkouts, and embedded content can all move the number without any change in real visitor behavior. When a shift looks too sharp to be real, check your setup before your content. (Our guide to how GA4 events work helps when you need to dig into what’s firing.)

The 10-second timer is adjustable (but probably leave it)

If 10 seconds feels too generous for your content, GA4 lets you raise the bar: Admin > Data streams > your web stream > Configure tag settings > Show all > Adjust session timeout, then set Adjust timer for engaged sessions anywhere up to 60 seconds.

Most sites shouldn’t touch it. The moment you change the threshold, every engagement metric breaks comparability with your own history, and 10 seconds does the job of filtering accidental visits. The setting earns its keep mainly on sites where the first screen loads something people watch passively, and you want “engaged” to mean more than “the video autoplayed.”

What to remember

Engagement rate and bounce rate are mirror images; pick one and be consistent.
Never compare GA4 bounce rate with Universal Analytics bounce rate. Same name, different metric.
An engaged session needs only one of: 10+ seconds, a key event, or 2+ page views.
Use the metrics comparatively (channels, landing pages, trends), not against generic benchmarks.

If you’re still getting oriented in GA4, the beginner’s guide to Google Analytics 4 covers the full reporting model these metrics live inside.

Frequently asked questions

What is an engaged session in GA4?

A session counts as engaged when any one of three things happens: it lasts longer than 10 seconds, it includes at least one key event (conversion), or it includes 2 or more page views (or screen views in an app). Only one of the three needs to be true. Every session that meets none of them is a bounce.

Is bounce rate still in GA4?

Yes, but it's hidden by default. GA4 reports show engagement rate instead. You can add bounce rate to most standard reports by customizing the report (the pencil icon, which requires Editor access on the property) or by using it in an Exploration. It's the same data either way: bounce rate is 100 percent minus engagement rate.

Why is my GA4 bounce rate so much lower than it was in Universal Analytics?

Because it's a different metric with the same name. UA counted any single-page session with no interaction event as a bounce, even if the visitor read the page for 10 minutes. GA4 only counts sessions that fail all three engagement checks: under 10 seconds, no key event, fewer than 2 page views. Blogs and content sites routinely dropped from 70 to 80 percent bounce rates in UA to much lower numbers in GA4 without anything changing on the site.

What is a good engagement rate in GA4?

It depends heavily on your traffic mix and site type, so treat published benchmarks with suspicion. Content sites and blogs tend to sit lower than sites people visit with intent, like SaaS apps or stores. The more useful habit is comparing your own segments against each other: engagement rate by channel, by landing page, and by campaign. A landing page well below your own site average is a real signal; a number below some industry average may not be.

Can I change the 10-second threshold for engaged sessions?

Yes. In Admin, open your web data stream, then Configure tag settings, then Show all, then Adjust session timeout. The setting called Adjust timer for engaged sessions can be raised from 10 seconds up to 60 seconds. Most sites should leave it at 10; changing it makes your engagement rate incomparable with earlier data.

Which metric should I report, bounce rate or engagement rate?

They contain identical information, since one is the mirror of the other. Engagement rate is usually the better choice for reporting because it frames the number positively and because GA4 shows it by default. Use bounce rate when your audience is used to the old vocabulary, but label it clearly as GA4 bounce rate so nobody compares it with old UA numbers.