Analytics guide
Average bounce rate: why your own is the only one that counts
The average bounce rate you find online mixes old Universal Analytics data, different tools and every kind of page. This guide shows how Google Analytics 4 measures bounce rate, how to work out your own average by page type, device and source, and how to tell a bounce rate that needs fixing from one that does not.
Short answer
There is no single average bounce rate worth aiming for, because published averages mix years, tools, page types and traffic sources, and many use the definition from before Google Analytics 4. The useful average bounce rate is your own: each page type's rate over recent whole weeks, split by device and traffic source, kept as a baseline.
- In GA4, bounce rate is the share of sessions that were not engaged, the opposite of engagement rate.
- Never compare a GA4 bounce rate with a Universal Analytics one, because they measure different things.
- Read a high bounce rate against the page's job: a page that answers one question can bounce and still succeed.
- A bounce rate near zero, or one that falls overnight with no change to the site, usually means broken tracking.
- Start with the busiest pages, and judge changes on conversions, not bounce rate alone.
What is the average bounce rate?
There is no single average bounce rate that tells you whether your site is doing well. The averages published online combine different years, different analytics tools, different page types and different traffic sources, and many were measured under the old Universal Analytics definition. The average bounce rate that matters is your own, worked out page type by page type.
That is not a dodge. Bounce rate is a useful number, but only when you compare like with like: the same kind of page, on the same site, measured the same way, over time. This guide explains how Google Analytics 4 measures it, how to build your own baseline, and how to read what you find.
What is bounce rate in Google Analytics 4?
In Google Analytics 4, bounce rate is the percentage of sessions that were not engaged. It is the opposite of engagement rate: if 60% of sessions were engaged, the bounce rate is 40%. So to understand bounce rate in GA4, you need Google's definition of an engaged session.
According to Google's Analytics Help page on engagement rate and bounce rate, a session is engaged if it meets any one of these conditions:
- It lasts longer than 10 seconds.
- It has a key event (what GA4 now calls an action you mark as important, such as a purchase or a lead form).
- It has two or more page or screen views.
A visitor who reads a page for eight seconds and leaves without clicking anything is a bounce. A visitor who reads the same page for two minutes and leaves is not, even though they never saw a second page.
How the definition changed in 2023
Universal Analytics, the previous version of Google Analytics, counted a bounce as a session with a single page view, however long the visitor stayed. Google says standard Universal Analytics properties stopped processing data starting July 1, 2023. Any bounce rate average measured before then describes a different number from the one in your GA4 reports, which is the first reason old benchmarks mislead.
Why published bounce rate averages mislead
A published bounce rate average is a blend of sites, pages and visitors that are not yours. Even a careful benchmark can only tell you what happened on other sites, measured their way. The same caution applies to other conversion rate statistics. These are the differences that make one site's bounce rate meaningless next to another's:
- Definition. GA4 counts time, key events and page views. Universal Analytics counted only page views. Other analytics tools use their own rules.
- Tracking setup. Which events you mark as key events changes your engagement rate, and so your bounce rate.
- Page type. A blog post, a landing page and a product page have different jobs, so they bounce differently.
- Traffic source. Visitors from a paid social ad, a newsletter and a branded search arrive with very different intent.
- Device. Phone visitors see less of the page at first and often load it on slower connections.
- The mix of pages. A sitewide average moves when your traffic mix moves, even if no single page changed.
The last point is the one that catches most site owners. If a blog post suddenly gets a lot of search traffic, your sitewide bounce rate can rise while every page on the site performs exactly as before. That is why the rest of this guide works page type by page type.
How to find your own average bounce rate in GA4
To find your own average bounce rate, add the bounce rate metric to your GA4 pages report, choose a recent period long enough to smooth out odd days, then split the result by page type, device and traffic source. Record each figure as your baseline and compare future periods against it, not against other sites.
-
Add bounce rate to your report
Most GA4 reports do not show bounce rate by default. Google's instructions: in Reports, open a detail report such as Pages and screens, click Customize report, open Metrics and add Bounce rate. You need an Editor or Administrator role to do this.
-
Pick a fair period
Use a recent stretch of whole weeks, and avoid unusual days such as a big sale or an outage. Never compare a GA4 period with a Universal Analytics period: they measure different things.
-
Group pages by type
Put pages with the same job together: articles, landing pages, product pages, category pages, the homepage. In GA4 you can do this with content groups or with a filter on the page path.
-
Split by device and source
Add device category and session source or medium as secondary dimensions. A landing page with a fine desktop bounce rate and a poor mobile one has a mobile problem, not a page problem.
-
Weigh pages by traffic and value
Sort by sessions and by revenue or leads. A high bounce rate on a page with a handful of visits is noise. A high bounce rate on a page that receives your paid traffic is money.
-
Write the baseline down
Record each page type's bounce rate by device and source, with the dates. That table is your average bounce rate. Every change you make later is judged against it.
How to read a high bounce rate, page by page
A high bounce rate is a question, not a verdict. Visitors who bounce had a reason, and the page type tells you which reasons are likely. Read it against the page's job: a page meant to answer one question can bounce and still succeed, while a page meant to move people deeper into the site cannot.
| Page type | Its job | High bounce rate usually means | Check first |
|---|---|---|---|
| Blog post or guide | Answer a question | Often fine: the visitor got the answer. A problem only if a next step was meant to be taken | Time on page, scroll depth, clicks on related links |
| Paid landing page | Turn an ad click into a lead or sale | The page does not match the ad's promise, or loads slowly | Ad copy against headline, speed on mobile |
| Homepage | Send people to the right section | Visitors cannot tell what you offer or where to go | First screen, navigation labels, main call to action |
| Category page | Help people choose a product | Thin or irrelevant results, poor filters, products hidden below the fold | Filters, sort order, what shows first on a phone |
| Product page | Lead to add to cart | Price, delivery or returns unclear, wrong traffic, out of stock | Price and delivery near the button, stock status, traffic source |
| Affiliate review page | Send visitors to a merchant | Possibly fine, if outbound clicks are not tracked as events | Whether outbound clicks are measured |
| Contact or thank-you page | Confirm or complete a task | Usually fine: there is nothing else to do | Nothing, unless leads have dropped |
Two more checks apply to every page. First, the traffic source: visitors from social feeds tend to arrive curious rather than ready, so a page fed mostly by social traffic will bounce more than the same page fed by search. Second, the query or ad that brought people in: if the page ranks for a search it does not really answer, the fix is the page's content or its title, not its design.
When a low bounce rate is a warning sign
A bounce rate that drops close to zero, or falls sharply overnight with no change to the site, almost always points to broken tracking rather than happier visitors. Fix the measurement before you celebrate, because every other number in the report is affected too.
In GA4 the usual culprits follow from the engaged session rules above:
- The tag is installed twice. A theme and a plugin, or a tag manager and a hard-coded snippet, both send a page view. Every visit then has two page views and counts as engaged.
- An automatic event is marked as a key event. If an event that fires on every page load is marked as a key event, every session has a key event and none can bounce.
- A redirect or interstitial records an extra page. A language or cookie step that loads as its own page can add a second page view to every visit.
The quick test: open your site in a private window, visit one page, leave after a few seconds, and check GA4's real-time report. One visit should show one page view.
Why is bounce rate higher on mobile?
Bounce rate is usually higher on mobile because a phone shows less of the page on the first screen, loads it over a slower connection, and is used in short bursts with less patience. Pop-ups and cookie banners cover more of a small screen, and forms are harder to fill in with a thumb.
To check whether mobile is your problem, compare the same landing pages on mobile and desktop, from the same traffic source. If the gap is large, open those pages on a real phone on mobile data and look at what appears before you scroll. Google's guidance on Core Web Vitals gives the loading, responsiveness and layout stability targets to measure against.
The full method is in Convertica's guide to mobile conversion optimization.
Average bounce rate for ecommerce sites
For an online store, a single ecommerce bounce rate is close to useless, because the pages that matter behave so differently. A product page reached from a shopping ad, a category page reached from search and a homepage reached from a brand search each need their own baseline, split by device.
The pattern to look for is a product or category page with plenty of traffic, a bounce rate well above other pages of the same type, and few add to cart events. That combination usually means the page is not answering a buying question: the price with delivery, when it will arrive, whether it is in stock in the right size, and what happens if it needs to go back. Bounce rate shows you where to look; your ecommerce metrics and a funnel report show what it costs.
How to reduce bounce rate in 8 steps
To reduce bounce rate, first confirm the tracking is right, then work on the busiest pages whose bounce rate is high for their type: match each page to the search or ad that brings people in, make the first screen clear, speed it up on mobile, remove what covers it, and give visitors one obvious next step. The same steps improve website engagement across the whole site.
-
Fix the measurement
Run the private window test above. If a single visit records two page views, or every session has a key event, fix that first.
-
Start with pages that carry traffic and money
Pick the five to ten pages with the most sessions or revenue whose bounce rate is high for their page type. Ignore the rest for now.
-
Match the page to the promise
Look at the search queries in Google Search Console, or the ad copy, that send people to the page. If the headline does not answer that query or repeat that offer, rewrite it so visitors know at once they are in the right place.
-
Make the first screen do its job
On a phone, before scrolling, a visitor should see what this is, who it is for and what to do next. Move anything that does not help with that further down.
-
Speed the page up on mobile
Test the page in PageSpeed Insights and on a real phone. Heavy images, sliders and third-party scripts are the usual causes.
-
Remove what gets in the way
Pop-ups that appear on arrival, chat widgets covering buttons and oversized cookie banners all push people back to the search results.
-
Give one clear next step
Every page should make the next action obvious: a related guide, a product, a form. Link to related pages inside the text where they help, not only at the end.
-
Test the bigger changes
Fix obvious breakages straight away. For changes you are unsure about, run an A/B test and judge it on conversions, not bounce rate alone. Use the A/B test sample size calculator to check your traffic is enough, and the statistical significance calculator to read the result.
A lower bounce rate is only good news if conversions follow. If a change keeps people on the page but fewer of them buy or enquire, it has not helped. Track your conversion rate alongside bounce rate for every page you change.
Bounce rate vs exit rate vs engagement rate
Bounce rate, exit rate and engagement rate are often confused because they all describe people leaving. Bounce rate and engagement rate describe whole sessions and always add up to 100%. Exit rate describes the last page someone saw, however engaged they were before that.
| Metric | What it measures | Use it to |
|---|---|---|
| Bounce rate | Share of sessions that were not engaged | Find pages that lose people on arrival |
| Engagement rate | Share of sessions that were engaged (the opposite of bounce rate) | Report the same thing as a positive number |
| Exit rate | Share of a page's views that were the last in the session | Find where people leave a multi-step path, such as checkout |
For a multi-step path such as a checkout or a lead form, exit rate and a funnel report are more useful than bounce rate. The guide to the conversion funnel shows how to measure each step.
Bounce rate questions
What is the average bounce rate?
There is no single average bounce rate worth aiming for. Published averages mix data from different years, tools, page types and traffic sources, and many were measured before Google Analytics 4 changed the definition in 2023. The useful average is your own: the bounce rate of each page type over a recent period, split by device and traffic source.
How does Google Analytics 4 calculate bounce rate?
In Google Analytics 4, bounce rate is the percentage of sessions that were not engaged. Google counts a session as engaged if it lasts longer than 10 seconds, has a key event, or has two or more page or screen views. Bounce rate is the opposite of engagement rate, so the two always add up to 100%.
Is a 70% bounce rate bad?
It depends on the page. On a blog post that answers one question, many visitors leave satisfied, so a high bounce rate can be fine. On a homepage, category page or product page, whose job is to send people further into the site, the same rate deserves a closer look. Compare it with the same page type on your own site, not with a number from another site.
What is a good bounce rate for an ecommerce site?
A good ecommerce bounce rate is one that is falling, or stable while sales grow, on your own product, category and landing pages. Compare each page type with itself over time and split by traffic source, because paid shopping traffic, email and organic search behave differently. A sitewide ecommerce average hides the pages that are actually losing buyers.
Why is my bounce rate so high on mobile?
Bounce rate is usually higher on phones because pages load more slowly on mobile data, the first screen shows less, pop-ups cover more of the page, and people browse on the move with less patience. Check your mobile landing pages on a real phone on mobile data, then compare their bounce rate with the same pages on desktop.
Why did my bounce rate suddenly drop to almost zero?
A bounce rate close to zero almost always means a tracking problem, not a better site. Common causes are the analytics tag installed twice, which records two page views per visit, or an event that fires on every page load being marked as a key event. Either one makes every session count as engaged.
What is the difference between bounce rate and exit rate?
Bounce rate is the share of sessions on a page that were not engaged. Exit rate is the share of page views on a page that were the last in the session, however engaged the visit was. A checkout confirmation page has a high exit rate and should; a high bounce rate there would be strange.
Does bounce rate affect SEO?
Bounce rate is a metric in your analytics tool, not a ranking report. What matters for search is whether the page answers the query quickly and works well on a phone. Those are also the most common causes of a high bounce rate, so fixing them tends to help both, but the bounce rate number itself is a symptom to investigate.
Where bounce rate fits in a CRO audit
Bounce rate is one of the first places a CRO audit looks, because it points to pages that lose people before they have done anything. It never answers why. That comes from watching sessions, asking visitors, and reading the page the way a first-time visitor would. Convertica's free CRO audit app checks a page for people and for AI agents and gives you three fixes you can make now.