Bounce rate

Bounce rate is a web-analytics metric: the percentage of sessions in which a visitor lands on a page and leaves without any interaction — no click to another page, no form submission, no tracked event. A high bounce rate usually signals a mismatch between what the visitor expected and what the page delivered, though the "right" value depends heavily on page type and traffic source.

How bounce rate is measured

Definitions differ between tools, which trips up many comparisons. Classic Universal Analytics counted any single-pageview session as a bounce — even a visitor who read an article for five minutes. GA4 (Google Analytics) inverted the logic: it first defines an engaged session (lasting 10+ seconds, containing a conversion, or including at least two pageviews) and reports bounce rate simply as the share of sessions that were not engaged. The same site therefore shows a much lower bounce rate in GA4 than it did in UA.

Common causes of high bounce rates include:

• slow loading — visitors abandon before the page renders,
• content that does not match search intent or the ad that brought the click,
• intrusive pop-ups, ad clutter or broken mobile layouts,
• no obvious next step — weak calls to action and poor internal linking,
• misleading titles that overpromise.

Practical use

Treat bounce rate as a diagnostic, never a KPI in isolation. Segment it by landing page, channel and device: organic traffic bouncing at 85% on a comparison article may be fine, while paid traffic bouncing at 70% on a product page is burning budget. Pair it with engagement time and conversion data before drawing conclusions.

Reducing bounces starts with performance — compressed images, caching, fast hosting with low TTFB — then moves to relevance: headlines that answer the query on the first screen, visible CTAs, related-content links that invite a second pageview. A/B testing headline and layout variants turns bounce-rate reduction from guesswork into measurement.