What bounce rate is
Bounce rate is the share of visits in which the user viewed one page and left with no interaction of any kind. A bounce means: did not move to another page, did not press a button, did not add anything to the cart.
The difference between GA4 and Universal Analytics
In Universal Analytics a bounce was defined strictly: 1 pageview = a bounce. GA4 introduced the engaged session:
Engaged session = lasting ≥ 10 seconds
OR containing a conversion event
OR containing 2+ page views
Bounce Rate (GA4) = 1 − Engagement Rate
Because the methodologies differ, the same site can report a 65% bounce rate in Universal Analytics and 45% in GA4 — the figures are not comparable.
Interpretation by page type
| Page type | Normal range | Warning level |
|---|---|---|
| Homepage | 40–60% | >70% |
| Category / listing (PLP) | 30–50% | >65% |
| Product page (PDP) | 50–70% | >80% |
| On-site search results | 30–50% | >65% |
| Blog / articles | 60–80% | >90% |
Important: bounce rate is a relative metric. A 60% bounce rate on high-quality traffic — the user found the answer and left satisfied — is better than 40% on irrelevant traffic that will never convert.
How to lower bounce rate in e-commerce
A high bounce rate on category pages is most often a mismatch between what the traffic expected and what the first screen of the page delivers.
Technical causes:
– Slow loading (over 3 seconds) — the user leaves before the page appears
– Broken rendering on mobile
– Irrelevant title and description → attracted the wrong user
Content causes:
– An empty or badly filtered category
– A first screen that does not match the search query
– No faceted navigation → no way to narrow the results
Personalization fixes:
– Personalized PLP sorting — put relevant products in front of the user first
– Recommendation blocks on the page — create a reason to go further
– A personalized homepage — relevant categories and promotions above the fold
Bounce rate and behavioural signals in SEO
For a search engine, a high bounce rate followed by an immediate return to the results page is a clear negative signal. The algorithm reads it as “the page did not answer the user’s query” and may lower the ranking.
Diagnostic approach: in your analytics platform, look at bounce rate broken down by landing page and by the query that brought the visit. Find the queries with abnormally high bounce rates and rework the pages behind them.