How to use Bounce Rate in GA4
GA4 redefined bounce rate as the inverse of engagement rate — sessions under 10 seconds with no conversion and no second pageview. Here's how to surface it, read it, and act on it.
Bounce Rate in GA4
In GA4, bounce rate is the percentage of sessions that were NOT engaged — the inverse of engagement rate.
Google Analytics 4 redefined bounce rate. It is no longer "single-page sessions" the way Universal Analytics measured it. In GA4, a session is engaged if it lasts at least 10 seconds, fires a conversion event, or includes two or more pageviews/screenviews. Bounce rate is simply 100% minus engagement rate — the share of sessions that failed all three of those tests.
That shift matters for online stores. A shopper who lands on a product page, watches a 15-second video, and leaves is no longer a "bounce" in GA4 — even though they viewed only one page. The metric now leans closer to a quality signal than a navigation signal.
If you migrated from Universal Analytics in 2023, your bounce rate probably dropped by 30-60 percentage points overnight. That's not your site improving — it's the definition changing. Comparing pre-July-2023 bounce rate to your current GA4 number is apples to oranges.
Treat GA4 bounce rate as a diagnostic, not a headline KPI. It tells you what share of visits failed to engage at all — useful for spotting broken landing pages, bad paid traffic, or slow mobile experiences. For deeper context, pair it with its inverse, engagement rate, which GA4 surfaces by default.
How GA4 actually calculates bounce rate
The formal definition: bounce rate = 1 − engagement rate. Engagement rate is the percentage of sessions that meet at least one of three criteria — duration ≥ 10 seconds, at least one conversion event, or at least two pageviews/screenviews.
The 10-second threshold is configurable. In Admin → Data Streams → Configure tag settings → Adjust session timeout, you can change "Adjust timer for engaged sessions" between 10 and 60 seconds. Most stores leave it at the default 10s; raising it makes engagement stricter and pushes bounce rate up.
Note one quirk: GA4 only counts the foreground time of a tab as session duration. A visitor who opens your page, switches tabs for 30 seconds, then closes is counted as engaged only if the tab was active for 10+ seconds. This is why GA4 bounce rates skew lower than gut feel.
The conversion-event trap
Because firing a conversion event counts as engagement, marking too many events as conversions in GA4 will artificially deflate your bounce rate. If you tagged scroll_50 or view_item as conversions "to track them better", every qualifying session becomes engaged regardless of real intent. Keep conversions to genuine business outcomes — add_to_cart, begin_checkout, purchase.
Where to surface it: Explorations and Looker Studio
Bounce rate is hidden in standard GA4 reports — by default you see engagement rate. To add it, open any report, click "Customize report" (pencil icon, top right), go to Metrics → Add metric → Bounce rate, then save. It will now appear in tables alongside views, sessions, and engagement rate.
In Explorations, build a Free-form report, drag "Bounce rate" from the Metrics panel into the Values shelf, and pair it with a dimension like Landing page, Session source/medium, or Device category. This is where most landing-page diagnostics live. In Looker Studio, the metric is exposed as "Bounce rate" on any GA4 data source — no custom calculation needed.
Typical GA4 bounce rate by device — Shopify apparel store
The pattern above is typical: mobile bounces higher than desktop, and paid-social mobile traffic bounces highest of all — those visitors are pulled from a feed rather than actively searching, so intent is shallower. If your mobile bounce gap is wider than ~15 points, your mobile PDP is probably the bottleneck.
What's a "normal" bounce rate in GA4?
There is no universal benchmark — bounce rate depends heavily on traffic mix, vertical, and how long your average page takes to engage. The table below gives realistic ballpark ranges for stores in the €1M-€15M revenue band, segmented by vertical.
Read these as goalposts, not targets. A beauty store doing 45% bounce on cold paid traffic is healthy; the same number on branded search would be a flag. Always segment by source/medium before judging.
GA4 bounce rate ranges by vertical and traffic source (online stores, €1M-€15M)
| Vertical | Branded search | Non-branded search | Paid social | |
|---|---|---|---|---|
| Apparel & accessories | 30-40% | 45-55% | 55-65% | 35-45% |
| Beauty & personal care | 28-38% | 42-52% | 50-60% | 32-42% |
| Home & furniture | 35-45% | 50-60% | 60-70% | 40-50% |
| Consumer electronics | 32-42% | 48-58% | 58-68% | 38-48% |
| Food & beverage (DTC) | 30-40% | 45-55% | 55-65% | 35-45% |
If your numbers fall outside these ranges by more than 10 points in either direction, look at the cause before celebrating or panicking. Suspiciously low bounce rates usually trace back to over-tagged conversion events, as covered in the engagement-rate comparison.
Acting on the metric: what to investigate
Start with the landing pages with the highest bounce rate AND the most sessions — that's where the biggest revenue is leaking. In Explorations, sort by sessions descending, then filter to pages with bounce rate above your site average. The top 10 will usually account for 60-70% of bounce volume.
Then segment those pages by device and source. A PDP that bounces at 40% on desktop but 70% on mobile is a mobile UX problem — slow LCP, tap-target issues, or above-the-fold price shock. A homepage that bounces at 60% from paid social but 25% from email is a message-match problem between ad creative and landing copy.
Faster diagnosis with historical context
Bounce rate by itself rarely tells you why. Pair it with scroll depth, Core Web Vitals, and the next-event funnel for the same landing pages. Metricuno imports your historical GA4 sessions on day one and overlays them onto session recordings, so high-bounce pages come pre-annotated with the friction patterns — no waiting four weeks to collect data.
Frequently asked questions
Because the definition changed. UA bounce rate counted any single-page session; GA4 only counts non-engaged sessions, where engagement means 10+ seconds, a conversion event, or 2+ pageviews. Most sites see GA4 bounce rate land 30-60 points lower than UA.
They are exact inverses: bounce rate = 100% − engagement rate. GA4 surfaces engagement rate by default and treats it as the primary metric; bounce rate is available but you have to add it manually to most reports. See our bounce rate vs engagement rate comparison for when to use each.
Yes. Go to Admin → Data Streams → your stream → Configure tag settings → Show all → Adjust session timeout. You can set the engaged-session timer between 10 and 60 seconds. Raising it makes the bar stricter and your bounce rate will go up.
No. A session that fires a purchase event (or any conversion event) is automatically engaged, so it cannot be a bounce — even if the visitor only saw one page and stayed under 10 seconds.
Open the report, click the pencil "Customize report" icon, choose Metrics → Add metric → Bounce rate, save, and publish. It will now appear alongside views and sessions. The metric is also available natively in Looker Studio on any GA4 data source.
Not as a primary KPI. Treat it as a diagnostic for landing-page quality and traffic-source fit. Optimizing for engagement rate (or conversion rate directly) is more actionable, because bounce rate can be gamed by adding low-value pageviews or loose conversion events.
Heatmap and session-recording tools usually count any meaningful interaction (mousemove, scroll, click) as engagement, while GA4 requires 10 seconds of foreground time, a conversion, or a second pageview. Different definitions, different numbers — neither is wrong.
Only if you've enabled the scroll event as a conversion, which is not recommended. By default, scroll events are tracked but do not count toward the engagement definition. Time on page (10s+) and pageview count are what move the needle.
In Explorations, build a Free-form report with Session source/medium as the row dimension and Bounce rate as the value. Add Sessions as a second value so you can weigh the bounce rate against volume — high bounce on a tiny channel is noise.
No. Because it is computed as 1 − engagement rate, and engagement rate is bounded between 0 and 1, bounce rate is always between 0% and 100%. If you see odd values, it is almost always a data-source or formula issue in a downstream dashboard.
Get an AI expert review of your site
Paste your URL — Metricuno's AI runs the same heuristic checks a senior CRO consultant would, scoring your page and prioritising the fixes that'll move conversion fastest.