Bounce rate tells you what share of visits to your site were not engaged. In Google Analytics 4 a bounce is a session that lasted under 10 seconds, viewed one page and had no key event; you bring it down by giving visitors what they came for, faster: a quick page, the answer at the top and a clear next step. If speed is the problem, see website speed optimisation.
Bounce rate: what it means in GA4 and how to reduce it
In GA4, bounce rate is the share of sessions that were not engaged: under 10 seconds, one page, no key event. How to measure it and how to bring it down.
By
7 min read
Published on
Updated:
In short
- In GA4, bounce rate is the inverse of engagement rate: 100 % minus the engagement rate.
- The old Universal Analytics definition (single-page sessions) no longer applies: UA stopped processing data on 1 July 2023.
- Bounce rate is not a Google ranking factor. It is a symptom, not a target.
- The usual causes: a slow page, copy that does not match the search, mobile problems and pop-ups.
- Measure only with cookie consent, and watch the trend for each page rather than one number for the whole site.
What is bounce rate?
Bounce rate is the percentage of sessions that were not engaged. GA4 counts a session as engaged if at least one of these is true:
- it lasted longer than 10 seconds (the default, which you can raise to 60);
- it included at least one key event (called a “conversion” until March 2024), such as a form sent or a phone number tapped;
- it had at least two page or screen views.
Everything else is a bounce. The maths is simple: bounce rate = 100 % minus engagement rate. If 62 % of sessions are engaged, the bounce rate is 38 %.
The original version of this article used the Universal Analytics definition: the share of visits in which someone saw only one page. That is history now. Universal Analytics stopped processing data on 1 July 2023, and GA4 measures something more useful: whether the visitor did anything, not just whether they opened a second page.
| Universal Analytics (until 1 July 2023) | Google Analytics 4 (today) | |
|---|---|---|
| What counts as a bounce | a single-page session | a session that was not engaged: under 10 seconds, one page, no key event |
| How it is calculated | single-page sessions ÷ all sessions | 100 % minus engagement rate |
| Someone reads an article for 3 minutes and leaves | bounce | engaged session, not a bounce |
| Someone opens the page and leaves after 4 seconds | bounce | bounce |
| Someone taps your phone number within 5 seconds | bounce | engaged, if the tap is a key event |
| Where you see it | standard reports | once added to a report, or in Explore |
Why does bounce rate matter?
Because it points to where a page fails to keep its promise. If people arrive from search and leave within seconds, something between the Google result and the page does not match: the speed, the content or the first impression.
It also matters what bounce rate is not. It is not a ranking factor: Google has said many times that it does not use Google Analytics data to order its results. The old article claimed otherwise, and that was wrong. But the causes of bouncing often hurt the things Google does measure. A slow page has poor Core Web Vitals, and Google uses Core Web Vitals in its ranking systems. So do not chase the number; fix the causes.
How do you measure bounce rate in GA4?
Bounce rate is not in GA4’s standard reports, but it takes a minute to add:
- Open Reports › Engagement › Pages and screens.
- Click the pencil icon (Customize report) at the top right.
- Under Metrics, add Bounce rate (and Engagement rate, if you want both).
- Click Apply and save your changes to the report.
Three tips to make the number useful:
- Mark the actions that matter as key events: a form sent, a phone link (
tel:) tapped, an e-mail link clicked, an add to basket. A visit that ends in a phone call should not count as a bounce. - Look by page and by source. One sitewide number mixes your blog, your contact page and your ads. Compare a page with itself before and after a change.
- Leave the 10-second threshold alone unless you have a reason. You can raise it (Admin › Data streams › your stream › Configure tag settings › Adjust session timeout), but old and new data will no longer compare.
One rule is not optional: analytics loads only after consent. In the EU (the ePrivacy Directive and GDPR) and in the UK (PECR), analytics cookies generally need consent, so your banner needs equal “Accept” and “Reject” buttons and nothing loads before the choice. That means GA4 only sees visitors who said yes. Watch the trend, not the absolute figure.
What causes a high bounce rate, and how do you reduce it?
This is the checklist we run on a page with a high bounce rate. Start at the top: the first three causes are the most common.
| Cause | How you spot it | What helps |
|---|---|---|
| Slow page | LCP above 2.5 s in PageSpeed Insights, especially on mobile | WebP images at the size they are shown; less JavaScript; page caching |
| Copy does not match the search | Search Console shows the page for questions it does not answer | the answer in the first sentences; a title that says exactly what the page holds |
| Mobile problems | small text, buttons crowded together, the page shifts (CLS above 0.1) | responsive design tested on a real phone; image dimensions in the markup |
| Pop-ups | a box covers the content as soon as the page opens | a smaller cookie banner; no pop-ups before people have seen what they came for |
| No next step | people read but do not go further | related articles, a clear button, a phone number and form in view |
| Confusing navigation | people go back to search instead of opening the menu | a short menu in your customers’ words; breadcrumbs |
| Technical errors | 404 pages, broken forms, console errors | regular link and form checks; error monitoring |
| The wrong traffic | high bounce from one ad or one social network only | ads that promise what the page actually shows |
Speed is the cause that is easiest to measure and most often missed. On Tandem Export, pages are served from a cache, a product change refreshes only the pages it affects, and images are WebP at the size they are shown. For real visitors on phones, the home page’s main content appears in 0.9 seconds (Chrome UX Report, the 28 days to 1 November 2026).
Not every bounce needs fixing. The Bulgarian original of this article picked up Google impressions for “what is a coefficient”: people looking for maths. They leave at once, and that is not the page’s fault. Check in Search Console what people searched for before you rewrite anything.
When is a high bounce rate fine?
When the page did its job in one move. Someone looks up your phone number, sees it and calls. Someone wants an answer, reads it in 8 seconds and closes the tab. Both visits succeeded, yet both can count as bounces if the phone tap is not set up as a key event.
So set up your key events first, and judge the page after.
Frequently asked questions
What is a good bounce rate?
There is no single good number for every site. An article that answers a question and a contact page where people grab your phone number will always behave differently. Compare each page with itself over time, and with similar pages on your own site, rather than with someone else's averages.
Does bounce rate affect Google rankings?
Not directly. Google has said repeatedly that it does not use Google Analytics data for ranking. The causes of a high bounce rate, such as a slow page or copy that does not match the search, do hurt visitors and the signals Google does use, such as Core Web Vitals.
What is the difference between bounce rate and exit rate?
Bounce rate is about the whole visit: how many sessions were not engaged. Exit rate is about one page: how many of its views were the last in a session. A page with lots of exits can be perfectly fine if it is the final step, such as an order confirmation.
Why is my bounce rate lower in GA4 than it was in Universal Analytics?
Because the definition changed. In Universal Analytics every single-page visit was a bounce, even if someone read for five minutes. In GA4 that visit counts as engaged if it lasted more than 10 seconds or included a key event, so it is not a bounce.
