“What’s a good bounce rate?” is one of the most searched analytics questions, and most answers give you a single number. That number is close to useless, because bounce rate depends heavily on what kind of page you’re looking at and on how your tool defines a bounce in the first place. This guide gives you rough ranges, explains why they vary so much, and shows what to do when yours looks worrying.
First: which bounce rate are you looking at?
Different tools mean different things by “bounce”:
- Classic definition (Universal Analytics, most privacy-friendly tools). A bounce is a visit that viewed only one page. Someone who reads your whole article for eight minutes and leaves counts as a bounce.
- Google Analytics 4. Bounce rate is the opposite of engagement rate. A session is “engaged” if it lasted at least 10 seconds, had a key event (conversion), or viewed two or more pages. Everything else is a bounce. The same eight-minute reader is not a bounce in GA4.
- Tools with custom events. Some tools let interaction events (a video play, a form start) count as engagement, which lowers bounce rate further.
That’s why GA4 bounce rates usually look much lower than the classic one for the same site. Neither is wrong. Just don’t compare a GA4 number with a benchmark built on the classic definition, or vice versa.
Rough ranges by page type (classic definition)
These are rough, commonly observed ranges for the one-page-visit definition, not hard benchmarks. Every site is different, and published “industry benchmarks” vary widely depending on who collected them. Use them to spot something unusual, not to grade yourself.
| Type of page | Rough typical range | Why |
|---|---|---|
| Blog posts, news, documentation | 65–90% | People come for one answer, get it and leave. That’s success. |
| Landing pages (single offer) | 60–90% | One page by design. Judge them by conversion rate instead. |
| SaaS or product marketing sites | 40–65% | Visitors usually check pricing, features or docs. |
| Ecommerce | 25–55% | Shoppers browse categories and products. |
| Homepages | 30–60% | The homepage’s job is to send people somewhere else. |
If you want to check your own number against these, the free bounce rate calculator does the arithmetic and tells you which range you’re in.
When a high bounce rate is fine
A bounce is only bad if the page was supposed to lead somewhere. It’s perfectly normal when:
- The page answers a question completely (a recipe, a how-to, a definition, opening hours).
- The visitor’s goal is off-site, like clicking a phone number, a download or an affiliate link. (Track those clicks as events, so you can see it.)
- The visitor is a returning reader checking one new post.
For content pages, time on page and scroll or read events tell you much more than bounce rate. A 90% bounce rate with three minutes of reading is a great article.
When a high bounce rate is a real problem
Pay attention when:
- A page that should lead to a next step (pricing, a product page, signup) has a high bounce rate and very short visits.
- Bounce rate jumped suddenly after a release. Something might be broken: a slow load, a layout shift, a missing button, a JavaScript error on one browser.
- One traffic source bounces far more than others. This usually means a mismatch between what the ad, post or search snippet promised and what the page delivers.
- Mobile bounces far more than desktop on the same page. Usually a layout or speed issue.
How to diagnose it, step by step
- Split by page. A site-wide bounce rate mixes your blog and your pricing page. Look at entry pages one at a time.
- Split by source. Compare Search, Social, Email, Paid and AI assistants for the same page. A bad source points to a promise problem; a page that’s bad from every source points to a page problem.
- Split by device. Then open the page on a real phone over a slow connection.
- Look at time on page. Bounces in under ten seconds mean “wrong page” or “broken page”. Long bounces mean “got what I needed”.
- Watch a few real journeys. Averages hide the story. Reading through ten individual visits (where they came from, what they saw, where they stopped) often explains a number in minutes.
Practical ways to lower a bad bounce rate
- Match the promise. The headline should echo the ad, search snippet or social post that brought people in.
- Speed up the first view. Compress the hero image, cut third-party scripts, and keep your analytics small. See how much analytics scripts weigh.
- Give one obvious next step. A clear button beats a menu with fourteen options.
- Link related content in context. “If you’re doing X, read Y next” inside the article works better than a generic “related posts” grid.
- Fix mobile layout. Sticky banners, pop-ups and cookie walls that cover half the screen send people straight back.
Don’t chase the number
It’s easy to lower bounce rate artificially: split articles over several pages, fire a fake interaction event after a few seconds, or count any scroll as engagement. Your dashboard will look better and nothing real will change. Bounce rate is a symptom metric. Use it to find pages worth a closer look, then measure the thing you actually care about: signups, purchases, leads, or readers who come back.
For that, set up a few custom events and check your conversion rate by page and by source. That’s the number that pays the bills.
How Cool Analytics shows bounce rate
Cool Analytics uses the classic definition: a visit that saw one page counts as a bounce. Next to it you get time on site, so you can tell a quick exit from a long read, and every visitor’s journey page by page. The plain-English insights flag it when a large share of visits leave right away, and suggest checking what your top entry page promises. That’s usually the right place to start.