Bounce Rate and Engaged Sessions in GA4, Explained

Olamide Sule, founder of Dolphin Analytics: a digital analytics expert based in London working with agency and in-house teams.

Bounce rate confuses people in GA4 because Google kept the name and changed the meaning. We set up and audit GA4 for agencies and marketing teams most weeks, and the engagement metrics come up in almost every account. This is the plain explanation we give clients: what bounce rate now measures, how engaged sessions decide it, and why the number on its own tells you very little.

What is bounce rate in GA4?

Bounce rate in GA4 is the percentage of sessions that were not engaged, which makes it the exact inverse of engagement rate. If engagement rate is high, bounce rate is low, and the two always add up to the whole. That single fact clears up most of the confusion, because the number is defined entirely by what GA4 counts as an engaged session.

This is not what bounce rate meant in Universal Analytics. There, a bounce was a single-page session with no interaction event, so a visitor who read one long article and left counted against you. GA4 threw that definition out. A one-page visit can now count as engaged, and a session with several pageviews can still bounce if it ends quickly and does nothing.

What is an engaged session in GA4?

An engaged session in GA4 is any session that meets at least one of three conditions: it lasts longer than 10 seconds, it fires a key event, or it records two or more pageviews or screenviews. Meet one and the session is engaged. Meet none and it is a bounce. Google’s Analytics Help documentation (checked September 2026) sets the default engagement timer at 10 seconds and lets you raise it up to 60 seconds.

Each of the three conditions catches a different kind of real visit:

• Time on the page. A session that stays open past the timer counts, because someone reading or watching is engaged even on a single page.
• A key event. If the session fires an event you have marked as a key event (GA4’s name for a conversion), it counts, whatever else happens.
• Depth. Two or more pageviews or screenviews counts, because a visitor moving through the site is engaged by definition.

The timer is the part teams forget they can change. To adjust it, open Admin, go to Data streams, select your web stream, click Configure tag settings, show all settings, then Adjust session timeout. Raising the timer from 10 seconds makes engagement harder to earn, so the same traffic reports a higher bounce rate overnight. Change it deliberately, and write down when you did, or a later trend line will look like a problem that never happened.

What counts as a high bounce rate in GA4?

There is no universal benchmark for a high bounce rate in GA4, and any guide that gives you one number is guessing. Bounce rate depends on the job of the page. A blog article that someone reads for two minutes and leaves is a good visit that still bounces if it has one pageview, no key event, and the reader was quick. A checkout page with the same bounce rate is a fire to put out.

The useful question is not “is this number high” but “which of the three engagement conditions is this page failing”. If a content page bounces, the timer or a scroll event is usually the lever. If a commercial page bounces, a missing key event is the first thing to check, because a page that never fires a conversion event cannot register the one signal that matters most. That is where the number stops being a vanity metric and starts pointing at a fix.

Where do you find bounce rate in GA4?

GA4 hides bounce rate by default, which is why so many people think it was removed. It is still there; you have to add it. In a standard report, open Reports, choose the report you want, click the pencil icon in the top right to Customise report, select Metrics, then Add metric, and pick Bounce rate from the list. Save the report and the column appears.

In an exploration, the path is shorter. Open Explore, start or open a report, find Bounce rate in the Metrics panel on the left, and drag it into the Values box. Engagement rate sits right next to it in the same list, so add both and watch them mirror each other. If either number looks wrong the moment it appears, the cause is nearly always the engagement setup underneath, not the metric itself. See our GA4 audit checklist for the order we check that in.

Bounce rate or engagement rate: which should you report?

Report engagement rate. It holds exactly the same information as bounce rate, built from the same engaged-session definition, but it points the right way: a rising engagement rate is good news, where a rising bounce rate reads as a problem. Dashboards that lead with engagement rate get read correctly more often, because the direction of the number matches the direction of success.

Bounce rate earns its keep in one situation: explaining GA4 to someone who learned analytics on Universal Analytics and still thinks in bounces. For that person, “your bounce rate is the inverse of this new engagement number” is the sentence that makes it click. Once the team has crossed over, engagement rate is the cleaner metric to standardise on. Neither is more accurate; they are the same measurement wearing different labels.

If bounce and engagement rate still feel disconnected from the rest of GA4, the missing piece is usually how the tool counts sessions and events in the first place. Our guide to GA4 sessions and events covers that layer, and it is worth reading alongside this one.

How we check engagement metrics in a GA4 audit

When we audit a GA4 account, a misread engagement rate is one of the faults we find most. The number is not broken; the setup underneath it is. Key events are not marked, so commercial pages bounce when they convert. The engagement timer was raised months ago and nobody noted it, so a bounce-rate trend looks alarming for no reason. The metric is only ever as trustworthy as the three conditions feeding it, and those conditions are configuration, not fate.

Reporting accuracy is the other half of the job. We check every number in a client report against the raw GA4 data before it goes out; on one report that check caught a figure reported as 23% when the real number was 21.4%. A bounce rate quoted from a report nobody reconciled is a number waiting to be wrong. See how we approach tracking for the checks we run before trusting any metric.

If your engagement numbers look off and you want someone in the account to find out why, that is the paid GA4 audit: scoped per property, it works through the setup with access to GA4 and Google Tag Manager and returns a written findings report naming each fault and its fix. Start with the audit checklist if you would rather run the first pass yourself.