There is no single “good” bounce rate for every site. A good bounce rate is one that fits the job of the page: an explainer that fully answers a query may sit at 60–75% and still be healthy; a paid landing page at the same number with falling form fills is not. Before you compare yourself to a blog infographic, ask whether a one-page visit was a failure (wrong audience, slow load, broken promise) or a success (question answered, call made, form sent).
For agency support with this, see our local SEO across Scotland and the UK and about Smarter Digital Marketing.
For professional support with this, see our PPC management.
Decision lens: For this URL, did the session do what the page was built for?
I compare each URL to its own last 28 days and to sibling pages on the same template—not to a site-wide average that mixes blog posts, contact pages, and campaign landings.
The short answer by page type
Ranges only help when the page type matches. Use the table as a sanity check, then confirm with conversions and GA4 segmentation—not the number alone.
| Page type | Typical range (indicative) | High bounce often OK when… |
|---|---|---|
| Blog / explainer | ~50–80% | Query fully answered; strong SERP snippet |
| Service / product | ~30–55% | — |
| E-commerce product | ~20–45% | — |
| Landing page (paid) | ~30–60% | Single-step offer; form above fold |
| Contact / directions | ~60–90% | User called, opened maps, or emailed |
Claims that “26–40% is optimal” everywhere are misleading. Tag page type in GA4 (custom dimension or consistent URL pattern) so you are not benchmarking a contact page against a long guide.
What the metric actually is (UA, GA4, exit rate)
Bounce rate is a symptom metric: it suggests a gap between what the title, ad, or snippet promised and what the visitor got. Google has not listed it as a direct ranking factor. I use it with Google Analytics setup and website goals—not as a substitute for either.
In Universal Analytics, bounce was roughly single-page sessions divided by all sessions. GA4 centres on engagement rate; many reports show “bounce” as the inverse of engaged sessions. Thresholds (time on site, conversion, second pageview) can vary by property, so I cross-check Google’s engaged session definition and bounce vs engagement in GA4 before I present numbers to a client. Comparing this year’s GA4 figure to a 2019 UA dashboard is a common false alarm—the definition moved even when UX did not.
Exit rate is a separate question: where sessions ended, however many pages came before. A reader who opens three articles and leaves from this guide did not “bounce” here, but this URL can still show a high exit rate. I use bounce (or engaged-session inverse) for first-touch landings and paid pages; I use exit rate when a strong page needs a clearer “what next,” not when I am judging a one-shot answer page.
When high bounce is fine—and when it is not
High bounce is not a crisis by itself. Informational content often should end after one page. Calls, visits, and app opens never appear as page two. Missing engagement events on SPAs or embedded tools make GA4 overstate bounce, so I fix measurement before I rewrite headlines.
I always segment by channel, device, and landing URL. A mobile-only spike after a theme change often traces to LCP or layout (Core Web Vitals), not weak copy. A site-wide spike with flat conversions usually means tracking (duplicate tags, consent mode, broken events)—not a sudden hatred of your brand.
When I worry: bounce rises together with fewer leads, weaker scroll on money pages, or paid traffic where CPC climbs and time on page collapses. When I do not: a ranking guide holds ~72% bounce but assisted conversions and return visitors stay steady—the page did its job; forcing extra clicks would distort both UX and the metric.
Scenario I see often: LinkedIn sends traffic to a service page; bounce jumps from ~48% to ~79% and forms drop. The ad promised a fast audit; the H1 opens with a generic agency story. I align the first screen to the ad before any redesign. Another: after launch, only mobile bounce spikes—I check CrUX/LCP for that template and fix render before touching body copy. Slow servers still send people away; I tackle page speed before another hero rewrite.
How I diagnose before changing copy
I start by proving the data is trustworthy: engaged sessions and key conversions in GA4 should match Tag Assistant, realtime, and what the business sees in leads. Then I segment (channel, device, landing page, 14–28 day windows) and read the page like a cold visitor—does the first screen deliver what the title and ad promised?
Checklist (only after the above):
- Match intent across title, meta, and H1.
- Fix speed on templates that drive revenue.
- Add one clear next step on money pages—not five equal buttons.
I note in analytics whether the page is “acceptable informational bounce” or “fix required” so the next reviewer does not reopen a closed call. When many templates fail together, bounce alone will not explain it—a technical SEO audit covers crawl, indexation, and measurement; ongoing strategy sits under SEO services.
FAQ
Is 70% bounce bad? On a long blog post, often no. On checkout or a paid landing page with falling conversions, treat it as urgent.
Is bounce rate a ranking factor? Not as a stated direct factor. Fix tracking, UX, and ad alignment—not a mythical SEO threshold.
How do I see it in GA4? Use Engagement reports (Pages and screens) and confirm reporting identity and filters before you share a number. Use engagement rate for trends; do not mix UA and GA4 eras in one chart.
Closing point
A good bounce rate matches page intent and business outcomes after verified GA4 setup and honest segmentation. I trust goals and template-level history over universal “good” bands—and I fix promise, speed, and tracking before I chase a percentage for its own sake.
