Bounce Rate vs Engagement Rate: What's the Difference and Which to Track
Bounce rate is gone from GA4 and replaced with engagement rate. Here's what each metric actually measures, when they conflict, and what to use for cookie-free analytics.
If you moved from Universal Analytics to GA4, you noticed bounce rate is no longer the headline metric. GA4 replaced it with "engagement rate" — the inverse of the new bounce rate. The change has confused a lot of analysts. Here's what's actually different and what it means for how you interpret traffic quality.
The old bounce rate (Universal Analytics)
In Universal Analytics, a bounce was a session with a single hit — the user loaded one page and left without any other interaction. No second page view, no event, no form submission.
The classic bounce rate was simple: sessions with one hit ÷ total sessions. A bounce rate of 70% meant 70% of sessions had only one page view.
The problem: modern analytics fires dozens of events on every page — scroll depth, video plays, button clicks, time on page. Even a user who reads your entire article and leaves satisfied fires many events. In UA, that's still a bounce.
GA4's engagement rate
GA4 defines an "engaged session" as one that meets at least one of:
- Lasted at least 10 seconds
- Had a conversion event
- Had at least 2 page views
The engagement rate is the percentage of sessions that meet this threshold. GA4's new "bounce rate" is simply 100% minus the engagement rate — it's the share of sessions that were NOT engaged.
The implication: a user who reads your 3,000-word article for 8 minutes and leaves is an engaged session in GA4. In UA, they were a bounce.
Which is more accurate?
Neither is "truth" — they answer different questions.
Old bounce rate answered: "Did the user do anything beyond the first page load?"
Engagement rate answers: "Did the user spend meaningful time or take an action?"
For content sites (blogs, documentation, editorial), engagement rate is more useful. A reader who spends 5 minutes on an article and leaves is a success, not a bounce.
For e-commerce or lead generation, old bounce rate was sometimes more useful as a proxy for intent: a 1-second visit with no action is a genuine bounce regardless of how it's labeled.
Cookie-free tools
Cookie-free analytics typically measures a simpler version of bounce rate: sessions where only one page was viewed. This is closer to the UA definition than GA4's.
Without JavaScript scroll events or time-on-page calculations tied to a persistent ID, you can't reliably measure engagement rate the GA4 way. You need at minimum two signals from the same user — but in cookie-free tools, "same user" within a session is defined differently (by IP hash within a time window).
In Gauss, a "bounced" session is a session with a single page view. Sessions with two or more page views are considered engaged. It's a proxy — not as nuanced as time-on-page — but it works across all traffic types without requiring cookies.
What each metric tells you by page type
Blog posts: a high single-page-view rate is expected and normal. Users often come from search, read an article, and leave. Low "bounce rate" on a blog post often means the content didn't satisfy intent (user hit back to search results). High bounce with long time-on-page (engagement rate) = content is working.
Landing pages: high bounce rate is concerning if you're running paid traffic. It means the ad promised one thing and the page delivered another. Low engagement rate on a landing page = weak conversion potential.
Homepages: moderate bounce rate is fine. Many users land on homepage to navigate somewhere else. A "bounce" that navigates internally in a single-page app may not register as a second page view in all tools.
Checkout / sign-up flows: a bounce here is almost always bad. If users enter the flow and leave at step 1, that's an abandonment problem, not a content problem.
How to use these metrics together
Track both where possible. They complement each other:
- High single-page-view rate + short time = genuine disinterest (traffic quality problem)
- High single-page-view rate + long time = content works but no next step (navigation / UX problem)
- Low bounce, low conversions = users are exploring but not converting (offer / CTA problem)
- Low bounce, high conversions = healthy funnel
If you're migrating from UA to GA4 or to a cookie-free tool, don't compare bounce rates directly. Recalibrate: set a new baseline for the current definition, then measure change over time.
The metric that actually matters
Neither bounce rate nor engagement rate tells you whether users achieved what they came to do. That requires goal tracking — did the user sign up, purchase, or request a demo?
Bounce rate and engagement rate are proxies for attention, not outcomes. Use them as health indicators for traffic quality and page experience, not as primary success metrics for your business.