Short answer
Use GA4 bounce rate by landing page and traffic source, alongside product views, add-to-cart activity or revenue. Check the settings before comparing.
The answer in detail
Yes, provided you know what it counts. GA4 bounce rate is the share of sessions that were not engaged: the exact inverse of engagement rate. For a store, it helps prioritize landing pages when read alongside a purchase-related measure.
The benchmark
In Universal Analytics, a one-page visit without an interaction was a bounce, even after several minutes of reading. Google's metric comparison explains that GA4 retained the name but changed the calculation. Engagement rate and bounce rate always add to 100%.
A session is engaged if it satisfies any one of these conditions:
| Criterion | Default threshold |
|---|---|
| Engagement time | More than ten seconds |
| Key events | At least one |
| Page or screen views | At least two |
Engagement time counts time when the page is visible. An abandoned background tab does not keep accumulating it.
The ten-second threshold is configurable by data stream. In the source's October 2026 interface, use Admin > Data streams > your stream > Google tag settings > Show all > Adjust session timeout. Check the engaged-session timer before comparing properties. Matching the metric name does not establish that its settings match.
The rule
Read bounce rate by landing page and traffic source, alongside product views, add-to-cart activity, or revenue. A sitewide average mixes pages with different purposes and identifies no specific page to improve.
Low bounce can be reassuring for the wrong reason. In a January 2026 Shopify audit described by the contributors, collection pages had 28.6% bounce, but just 16.3% of sessions landing there subsequently viewed a product. The best-sellers page had 31.4% bounce and only 4.3% product-view progression. The audit identified a merchandising problem: people stayed but did not reach products.
| Bounce | Purchase progression | Possible explanation | Inspect |
|---|---|---|---|
| Low | Weak | Page holds attention without guiding shoppers | Product prominence, filters, collection order |
| High | Reasonable among those who stay | Ad or search promise mismatches the page | Source targeting and message |
| High | Weak | Weak page, unsuitable traffic, or both | Start with the largest-volume source |
| Almost zero | None | Bots or an overly frequent key event | Traffic filtering and key-event settings |
A late-2025 jewelry-store audit found one referring site with 0.4% bounce and no product views. The team classified it as probable bot traffic; use the Shopify bot-filter guide when assessing such sources. Marking a common event such as a product view as key can similarly make almost every affected session engaged.
The steps
- Add the columns. The Landing page report does not include either rate by default. In Reports > Engagement > Landing page, choose Customize report > Metrics > Add metric. Add Engagement rate, Bounce rate, and Add-to-carts; apply and save. Editor or Administrator access is required.
- Sort by sessions. A page with 90% bounce and forty monthly sessions should not automatically consume a week's work. Consider impact as well as severity.
- Add the source. A source/medium secondary dimension distinguishes a page that struggles across sources from one poor campaign. For the latter, review ad-to-landing-page alignment.
- Use the grid. Compare add-to-cart performance with your store's baseline. The add-to-cart guide supplies context, and revenue per visit adds commercial meaning.
- Check configuration. In Admin > Data display > Events, ensure ordinary events have not inadvertently become key events. Keep the ten-second setting or date any change in your operating log.
The pitfall
Do not chart GA4 bounce as a continuation of Universal Analytics bounce. A graph spanning 2022 and 2026 can show an apparent improvement caused entirely by the definition change. The team may take credit for behavior that never changed.
The same problem exists within GA4. Raising the engagement timer to thirty seconds or adding a key event moves the curve immediately without any change in shoppers. Settings belong beside the chart's interpretation.
What to tell your team
Replace these fictional figures with your own:
“The overall bounce rate does not identify an action. Our three spring-campaign landing pages have 62% non-engaged sessions, compared with 31% on product pages. Collections hold attention but send only one session in six to a product. Those are this month's priorities.”
