What is a good bounce rate?
There is no universal good bounce rate. A blog post at 80% can be healthy while a homepage at 55% is failing, because a bounce means the visitor got what they came for on one kind of page and left frustrated on another. Judge it per page type, per traffic source, and against your own trend.
Worked example
The same site, one number, split three ways.
| Site-wide bounce rate | 68% |
|---|---|
| Blog, 6,100 sessions | 82% |
| Pricing page, 900 sessions | 44% |
| Homepage, 1,400 sessions | 51% |
The 68% site-wide figure describes no page on the site. The blog is doing its job. The homepage at 51% is the actual problem: half of the people who land on a page whose only purpose is routing never route anywhere.
Why the question has no single answer
Every published benchmark for bounce rate averages across page types that have opposite goals. A reference page succeeds when someone reads it and leaves. A category page succeeds when someone clicks deeper. Averaging the two produces a number that describes neither.
The second problem is that bounce rate is defined by whatever your tool counts as an interaction, and that varies by tool and by how carefully your tracker is installed. A site that fires a scroll event will report a bounce rate twenty or thirty points lower than the identical site that does not.
So the honest version of the question is narrower: what is a good bounce rate for this page type, from this source, on this device, compared to what it was last month?
Compared
Bounce rate benchmarks by page type Directional ranges for pageview-only bounce tracking, which is what most privacy-first analytics does by default. If your tool auto-fires engagement events, expect every number here to read lower.
| Page type | Healthy range | Investigate above |
|---|---|---|
| Blog post | 65% to 85% | 90%, and only if engagement time is also low |
| Documentation page | 70% to 90% | Rarely a problem. Look at search refinements instead. |
| Landing page, paid traffic | 45% to 70% | 75%, which usually means an ad-to-page message mismatch |
| Homepage | 30% to 50% | 55% |
| Pricing page | 35% to 55% | 60% |
| Product or category page | 20% to 45% | 50% |
| Logged-in app screen | Under 20% | Anything higher: check single-page-app route tracking |
Directional patterns from published industry reporting, not measured mrkr data. Your own month-over-month movement is a better signal than any external benchmark.
Live demo
See it on real data. The live demo is a fully populated dashboard on demo traffic, with no signup and no card.
Open the live demoFour things that change what good looks like
Traffic source
Organic search visitors arrive with intent and bounce least. Paid social visitors were interrupted mid-scroll and bounce most, often twenty points higher on the identical page. Compare a campaign to the page average, not to the site average.
Device
Mobile bounce rates run higher than desktop almost everywhere, partly through genuine impatience and partly through slower loads. A ten-point mobile gap is normal. A thirty-point gap is a performance problem.
Intent match
A page ranking for an informational query will bounce hard and should. A page ranking for a commercial query that bounces hard is losing money, because those people were ready to act.
Bot traffic
Crawlers load one page and never come back, so unfiltered bot traffic pushes bounce rate straight up. Measured over 30 days on a live site running mrkr, AI and search crawler hits ran at 8.7 times the human pageviews.
How to tell a healthy bounce from a broken page
Bounce rate alone cannot distinguish them, which is the whole limitation of the metric. Two other numbers can:
- Engagement time. A bounce with two minutes of engaged time is a satisfied reader. A bounce at four seconds is a bad landing or a slow page.
- Conversion events. A bounced session can still complete a meaningful action, so track the outcome the page exists to produce.
Read both alongside the page’s purpose. mrkr reports engagement time per page next to the bounce rate for exactly this reason, so you are never left interpreting a percentage in isolation. To check a single page's number against its type before you dig in, the free bounce rate calculator is the quickest route.
One last check before you act on any of it: confirm you are tracking client-side navigation. A large share of alarming bounce rates are single-page apps counting six screens as one pageview.
FAQ
Is a 70% bounce rate bad?
On a blog or documentation site, 70% is normal and often healthy. On a homepage or an ecommerce category page it is a real problem, because those pages exist to send people somewhere else. The page type decides, not the number.
What is the average bounce rate across all websites?
Published averages usually land somewhere between 40% and 60%, but the figure is close to useless because it blends page types with opposite goals and tools with different definitions of an interaction. Compare against your own trend and against the same page type instead.
How do I lower my bounce rate?
First check that it is real: track client-side route changes, and exclude bots. Then work on load speed, match the page's promise to whatever the visitor clicked, and give a clear next step above the fold. Never add an artificial timer event just to move the number, because that throws away the only signal it carries.
Start with one line of HTML.Your first visitors show up within a minute.