GA4 engagement rate is the percentage of sessions that were engaged sessions: visits that lasted longer than 10 seconds, included a key event or had at least two page views. Bounce rate still exists in GA4, but it is now simply the inverse.
By Kristijan GjorgjioskiIndependent SEO consultantPublished 18 October 2026 · 6 min read
Bounce rate was the first metric most marketers learned.
Then GA4 arrived, bounce rate disappeared from the default reports, and a new number took its place.
This guide explains what GA4 engagement rate measures, how it is calculated, what happened to bounce rate and how I read both in an audit.
GA4 engagement rate is the percentage of sessions on your site or app that were engaged sessions.
Google's formula is engaged sessions divided by total sessions.
An engaged session is one that meets any of three conditions, according to Google's engagement rate guide:
Here is a simple worked example:
| Figure | Value |
|---|---|
| Total sessions | 1,000 |
| Engaged sessions | 620 |
| Engagement rate | 620 / 1,000 = 62% |
| Bounce rate | 380 / 1,000 = 38% |
Notice the last row. Bounce rate is not calculated separately any more. It is what is left.
Engagement rate is not a quality score, and it is not a measure of how much someone read.
A visitor who leaves the page open on screen for 11 seconds while answering the door is engaged.
A visitor who finds a phone number in 8 seconds and calls you is not, unless that click is a key event.
That second case is why I never judge a page on engagement rate alone.
Bounce rate is still in GA4. It just means something different.
In Universal Analytics, a bounce was a single-page session. Read one article for five minutes, leave, and it counted as a bounce.
In GA4, bounce rate is the percentage of sessions that were not engaged.
The same five-minute reader is now engaged, because the session lasted longer than 10 seconds.
| Universal Analytics | GA4 | |
|---|---|---|
| Bounce means | Single-page session | Session that was not engaged |
| Long read of one page | Bounce | Engaged, not a bounce |
| Two pages in 5 seconds | Not a bounce | Engaged, not a bounce |
| One page, 6 seconds, no key event | Bounce | Bounce |
| Relationship to engagement rate | None | Always adds up to 100% |
The practical result: GA4 bounce rates usually look lower than the old ones.
That is a change of definition, not an improvement in your site.
Never put a Universal Analytics bounce rate and a GA4 bounce rate on the same chart. They measure different things.
Engagement rate sits in a family of metrics. Google's dimensions and metrics reference defines them like this:
| Metric | What it measures |
|---|---|
| Engaged sessions | Sessions that met at least one engaged condition |
| Engagement rate | Engaged sessions / total sessions |
| Bounce rate | Not engaged sessions / total sessions |
| Engaged sessions per active user | Average engaged sessions per active user |
| Average engagement time | Total time the page was in focus or the app in the foreground, divided by active users |
| Session key event rate | Sessions with a key event / total sessions |
Average engagement time deserves a note.
It counts time the page was in focus in the browser.
A background tab does not add to it.
That makes it a better read on attention than the old "time on page", which had no idea whether anyone was looking.
For commercial pages, session key event rate is the one I care about most.
It ties visits to outcomes, provided your key events are set up properly.
A few details change how the number behaves.
Key events count, but not all of them. Google excludes first_visit, first_open and session_start from the engaged session calculation, even if someone marked them as key events.
Page views count per view, not per page. Two page_view events in one session make it engaged. That is useful, and it is also the source of a common fault.
The timer is adjustable. The 10-second threshold can be changed per web stream. Google's session timeout instructions cover the path:
Raising the timer makes more short visits count as bounces, from that day on.
Past data does not change, so note the date anywhere people read trend charts.
Note: I only change the timer when there is a clear reason, such as a site where 10 seconds is not enough to load and read anything. Changing it to make a dashboard look better just breaks comparisons with your own history.
Engagement rate is easy to inflate by accident. These are the patterns I see most in audits:
Each of these is a tracking question first. My GA4 audit checklist covers duplicate tags, key events and referrals in order.
I do not give a single target, and I would be wary of anyone who does.
Engagement rate moves with:
The useful comparison is internal. Group landing pages by type, then look for pages that sit well below their group.
A guide with a much lower engagement rate than its sibling guides is worth opening.
Maybe the intro buries the answer, the page loads slowly or the search query does not match the content.
That is where engagement data becomes an input to content strategy, not just a number on a dashboard.
Engagement rate appears in several default reports, including the acquisition reports. Bounce rate does not.
To add bounce rate to a detail report, an Editor or Administrator can:
You can only add metrics to detail reports, not overview reports.
For SEO work, I add engagement rate and session key event rate to the Landing page report, filtered to organic search.
That shows which entry pages hold attention and which ones produce enquiries.
Engagement rate is a useful filter, not a verdict.
It tells you where to look, and your key events and sales data tell you whether it matters.
If the numbers look too good, or swing for no reason, check the tracking before the content.
Why GA4 data does not match covers the normal gaps, and data retention settings explain why explorations sometimes stop short.
If you want a second pair of eyes, my GA4 audit service tests what counts as an engaged session on your site, from duplicate tags to key events.
It also tells you whether the property is worth repairing.
If it is not, a planned GA4 setup starts from what you need to measure.
If it is, the audit's ranked fix list shows what to fix first.
No. Universal Analytics treated a bounce as a single-page session. GA4 defines bounce rate as the percentage of sessions that were not engaged, so a visitor who reads one page for a minute and leaves is engaged in GA4 but would have bounced in Universal Analytics. Expect GA4 bounce rates to look lower, and do not compare the two directly.
A sudden jump usually means something changed in tracking, not in visitor behaviour. A duplicated GA4 tag sends two page views per page, which makes almost every session engaged. A new key event on a common action has a similar effect. Check for double tagging and recent key event changes before celebrating.
Most GA4 reports do not show bounce rate by default. An Editor or Administrator can add it to a detail report by customising the report, then choosing Metrics and Add metric. It cannot be added to overview reports. In explorations, bounce rate is available as a metric you can add directly.
There is no single figure that applies to every site. Engagement rate depends on the page type, the traffic source, the device and how long your pages take to answer the visitor's question. Set your own baseline per page type and channel, then watch for pages that fall well below their group.
About the author
Kristijan Gjorgjioski
Independent SEO consultant · 6+ years in SEO · 60+ client websites worked on
I'm an independent SEO consultant with 6+ years in SEO. I run audits, migrations and content programmes for businesses in Europe, the UK and the USA, and I build the automation behind my own research.
Book a free 30-minute call. I look at your site first and tell you plainly what I would fix, and in what order.