GA4 redefined what counts as a bounce compared to older analytics tools — it now measures 'engaged sessions' (a session lasting 10+ seconds, with a conversion event, or with 2+ page views) and reports bounce rate as the inverse, which means bounce rates reported in GA4 are frequently lower than the same site would have shown in Universal Analytics, purely from the definition change, not from any real improvement.
A high bounce rate is not automatically bad. A blog post that fully answers a reader's question in one page, or a landing page with a single clear call-to-action that visitors click without a second page view, can show what looks like a 'high' bounce rate while performing exactly as intended — the single-page visit was the entire successful interaction.
The metric is only diagnostic when segmented by page type and intent. A high bounce rate on a checkout or pricing page is a real problem worth investigating; the same rate on a single-answer FAQ page that converts through non-pageview clicks or offline follow-up is often a non-issue entirely.
Formula
Bounce Rate = Single-Page Sessions with No Further Interaction ÷ Total Sessions × 100
In GA4, bounce rate is calculated as 100% minus the engagement rate — check which definition a given report is using before comparing it across tools or against historical Universal Analytics data.
Why Bounce Rate matters
Treating bounce rate as a universal quality signal, rather than a metric that means different things on different page types, leads teams to 'fix' pages that were never actually broken and to overlook real problems on pages where a high bounce rate genuinely does indicate failure.
Same metric, opposite meaning
A recipe blog post shows an 85% bounce rate — most readers get the recipe and leave, which is exactly the successful outcome the page was built for. A product page in the same site shows the same 85% bounce rate — visitors arrive and leave without viewing pricing, reviews, or any other page, which signals a real problem with the page's ability to move a genuinely interested visitor forward.
Common mistakes
Treating bounce rate as a universal health metric
The same percentage means something completely different on a single-answer content page than on a product or pricing page — judge it against what success looks like for that specific page.
Comparing GA4 bounce rate to old Universal Analytics data
The two tools define the metric differently. A drop in reported bounce rate after switching platforms can be entirely a definitional artifact, not a real improvement.
Trying to reduce bounce rate without asking why visitors are leaving
Adding unrelated internal links or auto-playing content to artificially lower the number doesn't address whatever actually made the visitor not want to continue, and can hurt the experience for visitors who were already satisfied.
Where we work on this