metrics
Bounce rate
What is bounce rate, and is a high one always bad?
Bounce rate is the share of visits where someone lands on a page and leaves without any further interaction - one page, no second click. It's easy to measure and easy to misread: a high bounce rate can mean people couldn't find what they wanted, or that they found it instantly and left satisfied. The number means nothing until you know what the page was for.
Also known as: bounce rate, bounces, single-page sessions
Prefer to watch?
Watch the recap 0:48
The demo
Here's a page with an 80% bounce rate - 8 in 10 visitors leave after one page. Disaster, or a roaring success? Flip the page's purpose and decide.
What this demo shows (text version)
A page with an 80% bounce rate is shown against two different purposes. As a checkout step, a high bounce is alarming: people are arriving and leaving without continuing to buy, so something is driving them away. As a quick-answer article (say, "what time does the post office close?"), the same 80% bounce can mean success - visitors got their answer in one page and left satisfied, with no reason to click further. Sessions are shown bouncing or continuing, and the verdict flips with the page type.
The lesson is that bounce rate is meaningless without intent: it's the share of single-page, no-interaction visits, and whether that's good or bad depends entirely on what the page was meant to do. It's also technically fragile (how "interaction" is defined changes it a lot), so treat a surprising bounce rate as a question to investigate with other signals - scroll depth, time, what people do next - not a verdict. (Figures illustrative.)
Bounce rate isn't good or bad on its own - it depends entirely on the page's job. On a checkout funnel, a high bounce is alarming; on a blog post or a "what's your phone number" page, it can mean total success - they got the answer and left happy. So always read bounce against intent, and never treat one number across very different pages as comparable.
The same 80% bounce rate flips from disaster to triumph the moment you change what the page is for. On a product funnel it means people are leaving before buying; on a quick-answer article it means they got what they came for in one glance. Bounce rate without context isn't a metric, it's a Rorschach test.
A "bounce" is just a single-page session with no tracked interaction, so its meaning is entirely contextual. On a page meant to lead somewhere - a homepage, a category page, a checkout step - a high bounce signals a problem: people aren't continuing. On a page meant to answer and release - a blog post, a contact-details page, a definition - a high bounce can be the sign of a job well done.
It's also a fragile measure technically. How "interaction" is defined changes the number wildly (analytics tools differ, and engaged-session definitions have shifted), and a visitor who reads a long article for five minutes then leaves still counts as a bounce. So a raw bounce figure is noisy, easy to game, and easy to misattribute.
Treat it as a prompt, not a verdict. A surprising bounce rate is a question - "is this page doing its job?" - to answer by looking at the page's intent and pairing it with other signals: time on page, scroll depth, what people do next, and conversion. Comparing bounce across pages with different purposes, or chasing a lower number for its own sake, is how teams optimise the wrong thing.