Engagement rate (GA4)(GA4)

Long definition

GA4 dropped Universal Analytics' bounce rate (single-page sessions divided by total sessions) and shipped engagement rate as the primary quality signal. The definition has three OR conditions: a session counts as engaged if any one is true.

• >10 seconds active duration. The browser tab must be in focus; background tabs do not accrue time.
• At least 2 page or screen views. A second page view triggers engagement regardless of time on page.
• At least 1 conversion event. Form submission, purchase, sign-up, or any custom event marked as a conversion.

Engaged sessions ÷ total sessions × 100 = engagement rate. Bounce rate in GA4 is now 1 - engagement rate and shows in some reports for backward compatibility, but its meaning has shifted. UA bounce rate counted any single-page-view session as a bounce. GA4 bounce rate counts only non-engaged sessions, so a single-page-view session lasting 30 seconds is an engaged session in GA4 — and would have been a bounce in UA.

For SEO interpretation, engagement rate is a better signal than UA bounce rate but still noisy:

• Single-page intent pages legitimately have lower engagement. A glossary entry that answers in one paragraph or a contact page may have 30% engagement and be doing its job.
• The 10-second threshold is short. Pages that load slowly can fail engagement purely on TTFB plus reading-start delay. Bot traffic that loads and immediately leaves drags the metric down.
• Custom events change the denominator. Adding a scroll-depth event marked as a conversion will inflate engagement rate without changing user behavior.

Compare engagement rate within content types and within channels, not across. Organic landing pages, paid landing pages, and blog posts have different baselines. The engagement metrics documentation lists the exact event-level rules.

Common misconceptions

• "Engagement rate = 100% - bounce rate from UA." No. UA bounce rate fired on any single-page session; GA4 engagement requires duration OR multi-page OR conversion. A 12-second session with one page view is engaged in GA4 and would have bounced in UA.
• "Higher engagement rate is always better." Pages designed to answer quickly (FAQ, definitions, contact info) deliver value with low engagement. The metric measures one definition of stickiness, not user satisfaction.
• "Engagement rate is a Google ranking signal." It is your analytics property's metric. Google Search does not have access to it. Don't optimize for engagement rate and expect SEO returns — optimize for content fit, and engagement rate often follows.
• "You can fix engagement rate by adding a scroll event as a conversion." Mathematically yes, but you've deceived yourself. The metric only helps when the trigger reflects real value. Tying it to a junk event makes it useless for diagnostics.