Bounce Rate Benchmarks 2026: GA4 Industry and Channel Data

What Is a Good Bounce Rate in 2026?

There is no single cross-industry bounce rate, because publishers measure different things. The most recent figure here is Littledata’s: a median GA4 bounce rate of 41% across 438 Shopify stores over the 90 days to October 31, 2026, with the best 20% of stores below 21% and the worst 20% above 56%. Databox’s August 2023 median across 15 industries was 44.82%, and Similarweb’s October 2021 panel average was 30.5% on a different definition.

Treat those as three separate yardsticks rather than one number with a margin of error. A store that sits near 41% is typical for Shopify; a bounce rate far from that is a prompt to look at traffic mix and tracking before changing the page. Littledata’s own FAQ gives “top 20%” values of 53% to 59% by industry that match its worst-20% threshold rather than its table’s best-20% column, so use the table.

How Each Source Defines Bounce Rate

The same label covers two different measurements. In GA4, bounce rate is the share of sessions that were not engaged: sessions that ended in under about 10 seconds, with no key event and fewer than two page or screen views. Universal Analytics, retired in 2023 and 2024, counted a session as a bounce when it viewed one page and sent one request, and gave it a duration of 0 seconds.

Similarweb’s help page and Contentsquare’s bounce-rate explainer both use the single-page definition. Because a single-page bounce ignores time on page, a visitor who reads an article for five minutes and leaves counts as a bounce there but as an engaged session in GA4. Google words the time rule two ways: its engagement-rate page says a session lasts longer than 10 seconds, while its comparison page says 10 seconds or longer. The one-second difference rarely matters, but the engaged-session timer can also be changed per data stream.

Bounce Rate by Industry

Databox’s GA4 benchmark remains the broadest industry table with a stated metric and month, though it dates from August 2023 and states no sample size. Its medians run from 27.92% for apparel and footwear to 49.47% for consulting and professional services, with eCommerce and marketplaces at 36.14%, SaaS at 48.24% and technology at 48.02%. Databox’s median bounce rate and median engagement rate do not add up to 100%, because each is the median of a separate distribution.

For ecommerce, Littledata’s current Shopify data is narrower but newer. Beauty and skincare stores had the lowest median (38%), while health and supplements and food and beverage were joint highest (46%). Contentsquare’s retail and financial services reports add context on engagement rather than bounce: for example, 77% of retail traffic was mobile in 2025, and financial services sites averaged 3.6 page views per session.

Bounce Rate by Traffic Channel

No publisher checked here gives a full current table of bounce rate by channel. What exists is narrower. Contentsquare’s 2026 benchmark reports a 53.6% bounce rate for AI-referred traffic in Q4 2025 (the report does not define bounce), just behind social and ads, and notes that AI platforms sent only 0.2% of traffic, up 632% year on year. It also reports falling bounce rates in AI-referred, paid search and organic search traffic against Q4 2024.

Adobe Digital Insights reports the opposite pattern for retail and travel: AI-referred visitors were 34% less likely to bounce than other visitors to US retail sites in July 2026, and 42% less likely on travel sites. Adobe does not define bounce, and Contentsquare measures all sites, not retail alone, so the two findings are about different populations. Both show AI referrals growing. For the wider picture, see our conversion rate benchmarks.

How Engagement Changed Year on Year

Contentsquare’s 2026 benchmark covers 99 billion sessions across more than 6,500 websites, comparing Q4 2025 with Q4 2024. Over that year overall engagement fell 10%, with time on site down 7%, page views per session down 1% and scroll rate down 2%. That is the context for any bounce-rate comparison: the same site can show a worse engagement profile than a year earlier without any change to its pages.

Its 2025 edition reported something related: businesses that raised session depth by 10% or more saw an average 5.4% gain in conversions, and a rise in reliance on paid social was associated with a 9.2% increase in bounce rate. Those are associations in vendor data, not experiments.

Bounce Rate and Engagement by Device

No current device bounce rates were found at the publishers checked, so this section compares behaviour instead. StatCounter measured mobile at 59.9% of worldwide page views in October 2026, and Contentsquare found mobile at 69.9% of its sample in Q4 2025. Across Littledata’s Shopify stores, the median store received 66% of sessions from mobile phones.

Mobile visits are shorter and shallower: Contentsquare recorded 4:46 per desktop session against 2:20 on mobile in Q4 2025, a scroll rate of 50.5% against 45.2%, and a conversion rate of 3.4% against 2%. A higher bounce rate on mobile is therefore plausible, but the numbers to compare are your own device segments in GA4, not a published gap.

Bounce Rate and Page Speed

Google recommends a Largest Contentful Paint within 2.5 seconds, an Interaction to Next Paint of 200 milliseconds or less and a Cumulative Layout Shift of 0.1 or less, each measured at the 75th percentile. INP replaced First Input Delay as a Core Web Vital in March 2024. The published bounce-rate evidence for speed comes from company case studies, which show direction but not a universal rate.

At Renault, Google reports that a one-second LCP improvement was associated with a 14-point lower bounce rate and 13% higher conversions, and with a 5-point fall on pages whose LCP was above 1.6 seconds. NDTV and The Economic Times reported bounce-rate falls of 50% and 43% after speed work, and Contentsquare’s 2025 release says slow-loading content was a key contributor to bounces, causing 53% of users to exit after a single page. None of these supports a fixed “bounce rises by X% after 3 seconds” rule. For landing-page effects, see our landing page statistics.

How Long Visitors Stay Before Leaving

Nielsen Norman Group’s reading of Microsoft Research dwell-time data, covering 205,873 pages, found that users often leave within 10 to 20 seconds, and that the first 10 seconds are critical. For 99% of pages, the longer a visitor stayed, the less likely they were to leave, so the chance of leaving flattens after about 30 seconds.

This is why a 10-second engaged-session threshold is a reasonable default but not a law. NN/g’s guidance is to judge bounce rate per page against pages of the same type, since a site-wide figure supports no specific action. A newsletter landing page, a blog post and a checkout page should not share a target.

Reading GA4 Engagement Rate Against Old Bounce Rates

Google publishes no conversion factor between a Universal Analytics bounce rate and a GA4 one, and says the two properties define sessions differently. GA4 does not restart sessions at midnight or on new campaign parameters, and Google says discrepancies of up to 20% in user and session metrics, and 10% in pageviews, can be expected even with comparable setups.

That is the reason any “UA 60% equals GA4 48%” table should be discarded. Compare GA4 with GA4, keep annotated notes where the property or timer changed, and read GA4’s engagement rate and bounce rate as complements: bounce rate is one minus engagement rate. For reporting setups, see our GA4 dashboards guide and our analytics services.

Older Universal Analytics-Era Figures

These figures are kept because readers still meet them, not because they describe 2026. They use the single-page definition and come from 2017, 2020 or 2021 data. CXL’s 2017 table puts email lowest at 35.20% and display highest at 56.50%; Contentsquare’s 2020 data gave 51% on mobile and 43% on desktop; and Similarweb’s 2021 analysis put quick-hit reference sites at 46.1% against 20.5% for streaming.

Do not use them as targets for a GA4 property. Their value is in the ordering, such as email and referral low and social and display high, which later sources broadly echo, and in showing how much a definition and an era change the answer.

How to Use These Benchmarks

Pick the benchmark whose definition and population match yours: Littledata for a Shopify store reporting in GA4, Databox for a broader GA4 sanity check, Contentsquare for engagement and traffic-mix context, and Similarweb for single-page comparisons. Then segment your own data by channel, device and page type before comparing, since a blended site-wide figure hides most of what matters.

Bounce rate is a diagnostic, not a goal. A high rate on a contact page or a recipe is normal; a high rate on a paid landing page is a prompt to check message match and speed. Pair it with conversion and revenue per session, and see our content engine and agentic SEO work for how we connect engagement data to content decisions.

Widely Quoted Figures, Checked

Several bounce rate statistics circulate without a source, or with a source that says something different. Each is checked against the publisher’s own page.

Sources

Every figure on this page comes from one of the sources below.