Bounce Rate
Bounce rate is the share of sessions on your site where a visitor views one page and leaves without triggering a second interaction, shown as a percentage. It doesn't tell you whether the visit succeeded or failed on its own — a high number can mean a page disappointed someone, or it can mean the page answered their question so completely that leaving was the right outcome.
Why bounce rate matters
Bounce rate is a diagnostic, not a verdict. It flags pages where visitors aren't doing anything after arrival, which is worth investigating even though the cause varies. One pattern search teams watch for is pogo-sticking: a visitor lands on your page, doesn't find what they need, and clicks straight back to try a different result. Google has never confirmed that its own systems compute a "bounce rate" as a ranking input, but behavior patterns correlated with dissatisfaction can still feed into how a page performs over time. Treat a rising bounce rate on a specific page as a prompt to check load time, message match, and whether the content actually resolves the query that brought people there — not as a scoreboard number to chase down in isolation.
How bounce rate is measured
In Universal Analytics, a bounce was any session that generated only a single request to the server: one pageview, nothing else, session ends. Bounce rate was single-page sessions divided by all sessions. Google Analytics 4 replaced that logic with "engaged sessions": a session counts as engaged if it lasts at least ten seconds, includes a conversion event, or spans two or more pageviews or screen views. GA4's bounce rate is defined as the inverse of that — the percentage of sessions that were not engaged. Because the underlying definition changed, GA4 bounce rate numbers are not directly comparable to Universal Analytics numbers from the same site; a jump or drop when you migrated tools likely reflects the definition change, not real user behavior change.
Common mistakes in interpretation
The most common error is treating every bounce as a failure. A blog post that fully answers a specific question, or a contact page that just needs to display a phone number, can do its job in a single pageview — the visitor got what they came for and left satisfied. Judge bounce rate against the page's job, not a flat target. A second error is broken tracking: a duplicated analytics snippet can double-count pageviews and drag the reported bounce rate toward zero, making a page look healthier than it is. A third is comparing a site-wide average across pages that serve completely different intents. Segment by landing page, channel, and device instead, so you're comparing similar pages to each other rather than blending a reference page with a checkout page.
Relevance to AI visibility and GEO
Zero-click behavior has been rising independent of AI, and generative answers are accelerating it. SparkToro and Similarweb found 68.01% of US Google searches between January and April 2026 ended with no click to any website, up from 60.45% in an earlier study. Pew Research Center, working from real browsing data across roughly 68,879 Google searches, found people clicked through to a traditional result in only 8% of visits where an AI summary appeared, versus 15% when it didn't — roughly half the click-through rate. Ahrefs, analyzing 300,000 keywords from December 2023 to December 2025, measured a 58% lower average click-through rate for the #1 organic result once an AI Overview appeared on that query. None of this is captured by a page's bounce rate, since the visit that would have happened never occurs. For GEO, the more useful signals are whether your brand gets cited or mentioned inside the AI answer itself; bounce rate still matters for the traffic that does click through, but it says nothing about the traffic that doesn't.
Example
Imagine an online shop for hiking boots. A user searches Google for "waterproof hiking boots review" and lands on a product category page. The page takes five seconds to load, has no reviews or trust signals visible, and the layout is cluttered. The visitor clicks straight back to the results list — a bounce. If the shop cuts load time, adds a short comparison of top models, and puts a clear buy button above the fold, more visitors stay, click into product pages, and the bounce rate on that category page drops. The same logic applies to a small tax firm in Leipzig checking whether visitors to its services page actually reach the contact form, or a neighborhood cafe checking whether people who land on its online menu scroll before leaving.
Common questions
What counts as a good bounce rate?
There's no universal target. Content and reference pages often sit in the 40-60% range; blog articles can run higher without it being a problem; shops, landing pages, and conversion-focused pages should generally sit lower. Compare a page against its own history and against similar pages on your own site, not against a number from a different industry.
Does bounce rate directly affect my Google ranking?
Google has not confirmed that it uses your analytics tool's bounce rate as a direct ranking signal, and it can't — Google doesn't have access to your Google Analytics data. Behavioral patterns like pogo-sticking may factor into ranking systems indirectly, but treat bounce rate itself as a diagnostic for your own use, not a lever you're pulling for Google's benefit.