What is a good bounce rate?
It depends on the page.And on who’s counting.
There is no single good bounce rate. In Google Analytics 4, where a bounce is a session that wasn’t engaged, Databox’s September 2024 benchmark put the median at 44% across 15 industries. Tools that count every one-page visit as a bounce report higher numbers, and an article that answers the question in one page can sit at 70% and be doing its job. This page explains both definitions, gives sourced ranges by site and page type, and reads your number for you.
By the Builderify team
What is a good
bounce rate?
Two questions first.
First: what does your tool call a bounce? Universal Analytics, and most simple analytics tools, count a visit that saw one page. GA4 counts a session that wasn’t engaged: it didn’t last longer than 10 seconds, had no key event and had fewer than two page views.Google The same traffic gives a lower bounce rate in GA4, so a GA4 number can’t be judged against a one-page benchmark, or the other way round.
Second: what is the page for? A blog post that answers a question in one page will bounce more than a shop’s category page, and that is fine. Google’s own help said so for Universal Analytics: if your success doesn’t depend on people viewing more than one page, a high bounce rate on a blog is “perfectly normal.”Google
With those two answers, rough guides do exist. For GA4, the median across 15 industries in Databox’s September 2024 data was 44.04%, and every industry median sat between 35.76% and 48.38%.Databox For one-page counting, Plausible’s compiled ranges run from 20–55% for online shops to 65–90% for content pages.Plausible Treat them as a place to start, then compare your own pages with each other: that is where the useful differences are.
Is my bounce rate good?
Read it against the right benchmark.
Pick your kind of site, the page, and the tool the number came from, then enter the rate. The reader compares it with the benchmark that uses the same definition, says what the band means and what to check first. It starts with a made-up accountant’s GA4 numbers. Nothing you enter leaves your browser.
In GA4, leave the bounce rate empty and enter the engagement rate if that’s the column you have: one is 100% minus the other.
Compared with
Databox GA4 median, Sept 2024, Consulting & Professional Services: 47.84%. Typical here means within 10 points of it.
What it means
Above the reference range. Not a crisis on its own, but worth finding which pages and sources carry it.
A site-wide average mixes pages with very different jobs. Split it by landing page before you change anything.
Check first
Split the rate by landing page and by source. One page or one channel usually carries the average.
How it reads. GA4 numbers are compared with Databox’s GA4 median for the closest industry; typical means within 10 points of it, which is our judgement. One-page numbers are compared with Plausible’s range for the site or page type, or Contentsquare’s averages where Plausible has none.
Rough guides, not targets. The benchmarks come from different samples, years and definitions, and they disagree. Under 10% is flagged whatever the page, because real traffic rarely gets there.
Your own trend matters more. A page that went from 40% to 60% after a change tells you more than any industry median. Compare a page with itself, before and after.
How does GA4
calculate bounce rate?
As the sessions
that weren’t engaged.
GA4’s bounce rate is the percentage of sessions that were not engaged. Google’s help defines an engaged session as one that lasts longer than 10 seconds, has a key event, or has two or more page or screen views.Google Engagement rate is the opposite share, so the two always add up to 100%.Google
The 10 seconds can be changed
Ten seconds is the default threshold for an engaged session, and Google says it can be adjusted in the Google tag settings of your web data stream (Admin, Data streams, your stream, then the tag settings), next to the 30-minute session timeout.Google Raise it and more short visits become bounces from that day on, so note the date before you compare months.
Universal Analytics counted one page
In Universal Analytics, a bounce was a single-page session, and bounce rate was single-page sessions divided by all sessions. Those sessions were recorded with a duration of 0 seconds, however long the person read.Google Standard Universal Analytics properties stopped processing data on 1 July 2023.Google If a benchmark or an old report predates GA4, it almost certainly uses the one-page definition.
What is a good engagement rate in GA4?
The Databox median for September 2024 was 56.21% across the same 15 industries, from 52.91% for IT and services to 61.74% for health and wellness.Databox The medians for bounce and engagement don’t add up to exactly 100%, because each is a separate middle value across many accounts.
| The visit | GA4 | Universal Analytics | User Journey |
|---|---|---|---|
| Reads one page for 2 min 40 s, then leaves | Engaged: longer than 10 s | Bounce, recorded as 0 s | Bounce, 2:40 active time |
| Opens two pages in 6 seconds | Engaged: 2 page views | Not a bounce | Not a bounce |
| Opens one page, leaves after 4 seconds | Bounce: not engaged | Bounce | Bounce |
| Bounce rate | 33% | 67% | 67% |
Three made-up visits. The first is the difference: a long read of one page is engaged in GA4 and a bounce in a one-page count.
What is a good bounce rate
for a website?
Three sources, and why they disagree.
Benchmarks disagree because they measure different things. One uses GA4’s definition, two count one-page visits; they cover different sites and different years, and one doesn’t name its studies. Here is each, with what it measures.
Databox
GA4: not engagedMedian 44.04% across 15 industries; industry medians 35.76% to 48.38%.
September 2024, accounts connected to Databox, by industry. Sample size not published.Source
Contentsquare
One-page visitsAcross industries: 43% on desktop, 51% on mobile, 45% on tablet. E-commerce: 47%.
Contentsquare’s benchmark data, quoted in an article last updated October 2023. The data year isn’t stated.Source
Plausible
One-page sessionsFrom 20–55% for retail and e-commerce to 65–90% for content. Suggests 50% or lower site-wide.
Ranges compiled from reported averages by agencies and tools; the studies aren’t named. Updated August 2026.Source
GA4 by industry (Databox, September 2024)
| Industry | Bounce rate | Engagement rate | Avg. session |
|---|---|---|---|
| Apparel & Footwear | 35.76% | 60.03% | 2m 40s |
| Ecommerce & Marketplaces | 38.61% | 60.71% | 2m 38s |
| Travel & Leisure | 38.84% | 60.71% | 3m 1s |
| Food | 38.93% | 55.53% | 2m 37s |
| Health & Wellness | 39.41% | 61.74% | 2m 58s |
| Automotive | 40.10% | 61.48% | 2m 51s |
| Health Care | 40.94% | 59.10% | 2m 40s |
| Industrials & Manufacturing | 41.78% | 57.85% | 2m 38s |
| Real Estate | 42.14% | 59.57% | 2m 43s |
| Construction | 45.28% | 55.06% | 2m 22s |
| Education | 46.28% | 53.82% | 2m 53s |
| Consulting & Professional Services | 47.84% | 52.97% | 2m 33s |
| SaaS | 48.27% | 52.93% | 2m 34s |
| Technology | 48.28% | 53.34% | 2m 43s |
| Information Technology & Services | 48.38% | 52.91% | 2m 40s |
Medians per industry, from Databox’s benchmark page.Databox
One-page count by site and page type (Plausible)
| Type | Range |
|---|---|
| Content websites and pages | 65–90% |
| Landing pages | 60–90% |
| Service industry websites (legal, automotive…) | 50–70% |
| Travel and hospitality websites | 40–60% |
| B2B websites | 25–65% |
| Real estate websites | 30–55% |
| Retail and eCommerce websites | 20–55% |
Compiled by Plausible from agency and tool studies it doesn’t name.Plausible
Two patterns hold across the sources. Pages built to be read and left (articles, landing pages) bounce more than pages people browse from, like shops. And phones bounce more than desktops: 51% against 43% in Contentsquare’s averages.Contentsquare
Bounce rate
vs exit rate.
Who arrived here,
against who left here.
Bounce rate only looks at visits that started on a page; exit rate looks at every view of it. Google’s Universal Analytics help put it this way: for all sessions that start with a page, bounce rate is the share that were the only page of the session; for all page views of a page, exit rate is the share that were the last of the session.Google
Every visit exits somewhere, so a high exit rate on a thank-you page is expected. A high exit rate on a step in the middle of a sign-up or checkout is the one to look at: people got that far and stopped.
- V1 Home Pricing Contact then left
- V2 Pricing then left
- V3 Pricing Contact then left
- V4 Home then left
| Page | Landed here | Bounce rate | Views | Exit rate |
|---|---|---|---|---|
| Home | 2 | 50% (1 of 2) | 2 | 50% (1 of 2) |
| Pricing | 2 | 50% (1 of 2) | 3 | 33% (1 of 3) |
| Contact | 0 | none | 2 | 100% (2 of 2) |
Counted the one-page way. Nobody landed on Contact, so it has no bounce rate, yet every view of it ended a visit.
What is a good
average session duration?
In Databox’s GA4 data for September 2024, the median average session duration was 2 minutes 38 seconds, with industries from 2m 22s (construction) to 3m 1s (travel and leisure).Databox That is a useful sense of scale, not a goal: a visitor who finds your phone number in 20 seconds and calls had a short, perfect session.
Check which time you are reading. GA4 defines average session duration as the average duration of users’ sessions, and average engagement time as the average time your site was in focus in the visitor’s browser.Google A tab left open behind others adds to the first and not to the second, so the two can be far apart. Universal Analytics recorded a bounced session as 0 seconds,Google which dragged its averages down; old and new figures don’t compare.
Bounce rate too high?
Check these, in this order.
Most high bounce rates are one page, one source, or not people at all. Work from the data to the page: make sure the number is real, find where it comes from, and only then change the page.
If people stay and browse but still don’t get in touch, the problem is elsewhere: the guide to a website that isn’t generating leads ranks the places inquiries leak.
-
01
Is it people?
Bots, uptime monitors and scrapers open one page and leave, which looks like a bounce in every tool. A sudden jump from one city, one network or one referrer is the usual sign. The guide to bot traffic in Google Analytics lists what to look for and how to filter it.
-
02
Did the counting change?
A second copy of the tracking code, a changed engaged-session timer in GA4, a cookie banner that now holds analytics back until people agree, or a switch of tools: each moves the number without any visitor behaving differently. Check the date it moved against your change log.
-
03
Which page, which source?
Split the rate by landing page and by source, and sort by the number of bounces rather than the rate. A large share from Direct often hides untagged links; read what direct traffic really is before blaming the page.
-
04
Does the page keep the promise?
Put the ad, email or search result that sent people next to the page they land on. If the headline doesn’t repeat what they clicked for, they leave.
-
05
Does the first screen work on a phone?
Phones bounce more than desktops in Contentsquare’s averages.Contentsquare Open the page on your own phone, on mobile data, and see what shows before you scroll.
-
06
Is a bounce bad here at all?
An article that answers the question, or a contact page that shows the phone number, can bounce and still work. Measure the action instead: a scroll, a form, a call.
How User Journey
counts a bounce.
And why it won’t match GA4.
In User Journey, a bounce is a visit that viewed one page. A visit ends after 30 minutes without activity, and the next page view starts a new one. Time on the page doesn’t change it, and neither does a goal reached on that page: goals are reported on their own. That is the one-page definition, so read Journey’s rate against the one-page benchmarks above. For the same visits it is usually higher than GA4’s, because GA4 lets a long read of one page count as engaged.
Bounce rate where it matters
Above the visitor list: visitors, page views, pages per visitor, bounce rate, returning visitors and average time on site. Insights break it down by source, campaign and landing page, and set it against the period before.
Bounce and exit rate per page
Paste any address into a page report to see its bounce rate, out of the visits that landed there, and its exit rate, out of all its views, with where people went next.
The landing page that loses most
Insights call out a landing page when 60% or more of the people who arrived on it left without opening a second page, once enough people have landed there.
Open the people who bounced
Filter the visitor list to bounced visitors and open any of them to replay what they did: where they came from, which page, how long and how far they scrolled.
People, not machines
Each visitor is marked a real person, VPN or proxy, or datacenter. One switch leaves VPN, proxy and datacenter traffic out of every figure, bounce rate included.
Readers who never click
Scroll and time-on-page goals show who read an article to the end, so a one-page visit that did its job is counted as a success, not just a bounce.
Visitors who say no to cookies still count. With the cookieless option, people who haven’t allowed analytics in your cookie banner show up in the totals without a cookie or stored ID, though they don’t get a journey page of their own. A GA4 setup that waits for consent may not count them at all, which is one more reason the two numbers differ. Average time on site counts active time on the page, so compare it with GA4’s engagement time rather than its session duration.
Then fix the page, and measure the fix. A/B Testing compares two versions of a landing page against the same goals Journey uses. For people already on their way out, an exit-intent popup from Popups can offer a smaller next step. If you’re weighing up GA4 itself, the Google Analytics alternative guide compares the options.
A tracked journey costs 0.1 credits, and only once the visitor opens a second page or stays ten seconds; quick one-page visits cost nothing. Every account gets 250 free credits a month, and one-time packs start at $9 with no monthly fee. See pricing.
Start freeBounce rate, before you ask.
It depends on the page and on how your tool counts a bounce. In GA4, Databox's September 2024 benchmark put the median bounce rate across 15 industries at 44.04%, with industry medians from 35.76% to 48.38%. Tools that count one-page visits report higher numbers: Plausible's compiled ranges run from 20–55% for online shops to 65–90% for content pages.
GA4's engagement rate is the share of sessions that lasted longer than 10 seconds, had a key event or had two or more page views, and it is 100% minus the bounce rate. Databox's September 2024 benchmark put the median at 56.21% across 15 industries. Use it as a reference point, then compare your own pages and sources with each other.
Bounce rate looks at the visits that started on a page and counts those that saw only that page. Exit rate looks at every view of a page and counts the views that were the last of the visit. A checkout step can have a high exit rate and almost no bounces, because hardly anyone lands on it.
Databox's GA4 benchmark for September 2024 put the median average session duration at 2 minutes 38 seconds across 15 industries. In GA4, average session duration covers the whole session, while average engagement time counts only the time your page was in focus, so compare like with like.
User Journey counts a bounce as a visit that viewed one page, however long it lasted; a visit ends after 30 minutes without activity. GA4 counts a session as engaged, not a bounce, once it lasts longer than 10 seconds. So someone who reads one page for three minutes is a bounce in Journey and engaged in GA4, and Journey's rate is usually the higher of the two.
Sources
- Google [GA4] Engagement rate and bounce rate, Analytics Help
- Google [GA4] Analytics dimensions and metrics, Analytics Help
- Google Configure your Google tag settings (session timeout and engaged-session timer), Analytics Help
- Google Comparing metrics: Google Analytics vs. Universal Analytics, Analytics Help
- Google [UA] Bounce rate [Legacy], Analytics Help
- Google [UA] Exit Rate vs. Bounce Rate [Legacy], Analytics Help
- Google Universal Analytics has been replaced by Google Analytics 4, Analytics Help
- Databox Content Marketing Benchmarks by Industry for 2025 (GA4 data for September 2024)
- Contentsquare What Bounce Rate Says About Your Customer Experience (updated 4 October 2023)
- Plausible What is a good bounce rate for a website? (and industry benchmarks), updated 3 August 2026
Google’s help pages and the three benchmarks were read on 29 October 2026. How User Journey counts a bounce comes from Builderify’s own code; the other Builderify details come from the module pages and the rate card on this site. The reader’s bands are rough guides built on those benchmarks, not targets.
Stop losing the visitors
you already paid for.
More leads from the traffic you already have, and follow-up that sends itself. Free to start. No card, no monthly fee.