A good bounce rate for most websites is roughly 25–55%, but "good" depends heavily on the type of page: online stores often sit around 20–45%, B2B and lead-generation sites around 25–55%, while blogs and single-purpose landing pages commonly run 60–90% without anything being wrong. Before comparing yourself to any benchmark, check which definition your tool uses — Google Analytics 4 changed what "bounce" means, and the same visit can count as a bounce in one tool and not in another. This guide explains both definitions, gives realistic benchmarks by site type, shows how to tell a healthy bounce from a harmful one, and walks through the fixes that actually lower it.
Key takeaways
- A good bounce rate for most business sites is roughly 25–55%; blogs and landing pages often run 60–90% with nothing wrong.
- GA4 counts a visit as a bounce only if it wasn’t engaged (under 10 seconds, no key event, one page), so GA4 rates are usually lower than classic single-page bounce rates.
- Always split bounce rate by source and device — in our demo, direct visitors bounced 27% and Instagram visitors 57%.
- High bounce is fine when the page answered the question; it’s a problem on pages meant to lead somewhere.
- Watch recordings of bouncing visits before redesigning anything.
What is bounce rate? Two definitions you need to know
| Classic definition (single-page sessions) | Google Analytics 4 | |
|---|---|---|
| A bounce is… | A session that viewed only one page | A session that was not engaged |
| "Engaged" means | — | Lasted more than 10 seconds, OR had a key event (conversion), OR had 2+ page views |
| Formula | Single-page sessions ÷ all sessions | 100% − engagement rate |
| Effect | A visitor who reads one article for 5 minutes is a bounce | That same visitor is not a bounce |
The practical consequence: GA4 bounce rates are usually lower than classic bounce rates for content sites, because long single-page reads now count as engaged. Most published benchmarks (including the ranges in this article) were built on the classic single-page definition, so they line up best with tools that still use it. Clicktics uses the classic definition — a bounce is a single-page session — and reports engaged time separately, so you can see both "did they leave after one page?" and "did they actually spend time?" side by side.
What is a good bounce rate? Benchmarks by site type
These ranges are widely cited across analytics studies and are best treated as orientation, not targets — your own trend matters more:
| Type of site or page | Typical bounce rate | Why |
|---|---|---|
| Ecommerce & retail | 20% – 45% | Shoppers browse several products and categories |
| B2B websites | 25% – 55% | Visitors research across solution, pricing and case-study pages |
| Lead generation / services | 30% – 55% | Visitors compare services before contacting |
| Content sites (non-ecommerce) | 35% – 60% | Mixed reading and browsing |
| Landing pages | 60% – 90% | Single-page by design — one offer, one action |
| Blogs, news, dictionaries | 65% – 90% | Visitors get their answer and leave satisfied |
A useful rough scale for most business sites: under ~40% is strong, 40–55% is average, 55–70% is worth investigating, and over 70% deserves a closer look — unless the page is a blog post, a landing page, or a page whose whole job is answered in one view (a contact page, an opening-hours page). Very low bounce rates (under ~15–20%) are often a tracking problem — for example, the analytics script loading twice and counting a second pageview.
Read bounce rate by source, not as one number
A sitewide bounce rate hides the useful information. Here's a demo store's overview — a perfectly ordinary 41.7%:
Break the same traffic down by source and the picture changes completely:
Direct and Google visitors bounce at healthy rates and drive most of the conversions. Instagram visitors bounce at 57% and convert at zero. That's not a homepage problem — it's a message-match problem: the social post promises something the landing page doesn't immediately deliver. The fix is a dedicated landing page or a different destination for that campaign, not a redesign that risks the pages already working for search visitors.
Device is the other split worth making every time. Mobile visitors routinely bounce more than desktop visitors on the same page, and a sitewide average that's drifting upward is often just mobile traffic growing — or a mobile-only layout problem.
Bounce rate vs exit rate: what's the difference?
The two are often confused, but they measure different things:
| Bounce rate | Exit rate | |
|---|---|---|
| Question | Of visits that started on this page, how many viewed only this page? | Of all views of this page, how many were the last page of the visit? |
| Includes | Only sessions that landed here | Every visit that saw this page, wherever it started |
| Best for | Judging landing pages and first impressions | Finding where multi-page journeys end |
Every page has an exit rate, because every visit ends somewhere. A high exit rate on an order-confirmation page is perfect; a high exit rate on a cart or pricing page is a warning sign. Bounce rate, by contrast, only tells you about visits that began on that page. Use bounce rate to judge whether a page makes a good first impression, and exit rate to find where journeys break down.
A worked example: lowering bounce on one page
Suppose a product page gets 2,000 landing visits a month with a 68% bounce rate, and 3% of non-bouncing visitors go on to buy at a $70 average order:
- Non-bouncing visits: 2,000 × 32% = 640 → about 19 orders → about $1,340 a month.
- After fixing a slow-loading hero image and moving the size guide above the fold, bounce falls to 55%.
- Non-bouncing visits: 2,000 × 45% = 900 → about 27 orders → about $1,890 a month.
That's roughly an extra $550 a month from one page, with no new traffic — and the change was found by watching ten recordings of visitors who left within five seconds. Bounce rate is worth improving when you can tie it to what happens next, the way this example does.
When a high bounce rate is fine
- The page answered the question. A visitor searched "store opening hours", found them, and left. That bounce is a success.
- It's a single-action landing page. If the conversion happens on the page (a form or a click to call), bounce rate is the wrong metric — use conversion rate.
- It's a blog post. Readers arriving from search for one answer often leave after reading. Look at engaged time and scroll depth instead: a 4-minute average read with 80% scroll depth is a very good page even at a 75% bounce rate.
The test is simple: did the page do its job? Pair bounce rate with engaged time, scroll depth and conversions before deciding it's a problem.
When a high bounce rate is a problem — and why it happens
A high bounce rate matters on pages that are supposed to lead somewhere: homepages, category pages, product pages, pricing and service pages. The common causes:
- Intent mismatch. The ad, post or search result promised something the page doesn't show above the fold.
- Slow loading. Visitors leave before the page renders — especially on mobile connections.
- Unclear next step. No obvious call to action, or several competing ones.
- Mobile layout problems. Pop-ups covering content, tiny buttons, text that needs zooming.
- Intrusive interruptions. Newsletter pop-ups or chat greetings firing in the first seconds.
- Broken or confusing elements. Buttons that don't work, images that look clickable but aren't — the friction patterns covered in our guide to rage clicks and dead clicks.
- Weak internal links. Nothing on the page invites the visitor to a second page.
How to lower your bounce rate: the fixes that work
1. Watch the visitors who bounce
Numbers tell you a page has a problem; recordings tell you what the problem is. Filter session recordings to single-page visits on the page in question and watch ten of them. Do visitors scroll and leave? Leave instantly? Hunt for something? The pattern usually appears within minutes. Our guide on how to see what visitors do on your website shows how to set this up.
2. Match the first screen to the promise
Whatever brought the visitor — the search query, the ad headline, the social post — should be visibly answered in the first screen of the page. Repeat the key phrase in the headline, show the product or offer immediately, and move anything generic below the fold.
3. Check the heatmap for where attention goes
A click heatmap shows whether visitors engage with your call to action or ignore it. If attention clusters on navigation and images but not on your primary button, move or redesign the button. (More on reading heatmaps in our landing page heatmap guide.)
4. Speed up the page
Compress images, remove unused scripts and plugins, and test on a real mid-range phone. Every analytics and marketing tag adds weight — one reason we keep the Clicktics script to about 24 KB.
5. Give visitors a natural second step
Add relevant internal links within the content, a "related products" or "related articles" block, and a clear call to action. For blogs, link to the next logical question your reader will have.
6. Delay or remove interruptions
Move pop-ups to exit intent or after meaningful scroll. Let visitors read before you ask for anything.
What to track alongside bounce rate
| Metric | Answers |
|---|---|
| Engaged time | Did single-page visitors actually read? |
| Scroll depth | How far down the page did attention reach? |
| Conversion rate | Did the page achieve its goal? |
| Exit pages | Where do multi-page journeys end? |
| Bounce by source & device | Which audience is mismatched? |
Bounce rate is a smoke alarm, not a diagnosis. These five numbers, plus the recordings behind them, tell you whether there's a fire. For how bounce fits into the bigger picture, see our guide to sales funnel metrics and the conversion rate benchmarks by industry.
Frequently asked questions
What is a good bounce rate?
For most business websites, roughly 25–55%. Ecommerce often runs 20–45%, while blogs and landing pages commonly reach 60–90% without a problem. Compare like with like — the same page type, the same traffic source, and your own history.
Is a 70% bounce rate bad?
For a blog post or a single-purpose landing page, not necessarily. For a homepage, category page or pricing page, it's worth investigating — start by checking bounce rate by traffic source and device, then watch recordings of the bouncing visits.
Why is my bounce rate lower in GA4 than before?
GA4 counts a visit as engaged (not a bounce) if it lasts more than 10 seconds, includes a key event, or views two or more pages. Long single-page reads that used to count as bounces no longer do, so GA4 bounce rates are usually lower.
Does bounce rate affect SEO?
Google has said it doesn't use Google Analytics bounce rate as a ranking factor. But the causes of a high bounce rate — slow pages, poor intent match, weak content — are things that also affect how well a page performs in search.
Where do I find bounce rate in GA4?
GA4 doesn't show bounce rate in most standard reports by default. You can add it by customizing a report (Customize report → Metrics → add "Bounce rate") or by building an Exploration. Remember it uses GA4's engagement-based definition, so it won't match tools that count single-page sessions.
Why is my mobile bounce rate higher than desktop?
Mobile visitors are more often interrupted, on slower connections, and more sensitive to pop-ups, small text and slow pages. Some gap is normal, but a large one usually points to a mobile-specific problem — watch a few mobile recordings of the page to find it.
What causes a very low bounce rate, like 5%?
Usually a tracking issue — the analytics tag firing twice, or an event being counted as a second pageview. Check your installation before celebrating.
See why visitors bounce, not just how many
Clicktics shows bounce rate by source and device, engaged time, scroll depth, and the session recordings and heatmaps behind every number — so a high bounce rate turns into a specific fix instead of a guess. Set it up in five minutes and try it free for two weeks.
Start your 14-day free trial → or see how engagement analytics works.
Tomás García
Tomás García writes for the Clicktics blog about session replay, analytics engineering, and building privacy-first products that agencies love. Reach the team at [email protected].