Bounce Rate Calculator

Background.

Bounce rate is the one web metric whose meaning changed underneath everybody. Until Universal Analytics was retired in 2023, Google defined it as "single-page sessions divided by all sessions" — a bounce was "a session that triggers only a single request to the Analytics server". Google Analytics 4 defines it as something else entirely: bounce rate is "the percentage of sessions that were not engaged", where an engaged session is one that "lasts longer than 10 seconds, has a key event, or has at least 2 pageviews or screenviews". This calculator implements both, because a number computed under one definition tells you nothing about a benchmark quoted under the other.

The practical size of the gap is easy to miss. Take 48,000 sessions. Under GA4, 18,240 of them were engaged, so the bounce rate is 62.00%. Under the old rule, 33,600 of the same sessions were single-page, so the bounce rate is 70.00% — eight points worse for identical behaviour. The direction is systematic rather than accidental: GA4's engagement bar is easier to clear, because a visitor who reads one long article for four minutes counts as engaged in GA4 and as a bounce in Universal Analytics. The one way the ordering can narrow is if a Universal Analytics property fired interaction events on scroll or video play, since those hits stopped a session being a bounce.

The GA4 threshold deserves particular attention, and the calculator prints the reason beside the result rather than hiding it. Ten seconds is a default, not a law. Google's own tag documentation states that "the default threshold for engagement time is 10 seconds but can be adjusted here", so two GA4 properties measuring identical visitors can publish different bounce rates purely because someone changed a setting. Before you compare your bounce rate with anyone else's — including your own from a previous agency — check what their threshold was.

The last output turns the rate into a plan. Enter a target and the calculator works out how many more sessions would need to stop bouncing on exactly the same traffic. Going from 62.00% to 55.00% across 48,000 sessions takes 3,360 additional engaged sessions; under the Universal Analytics figure of 70.00% the same target takes 7,200. That framing is more useful than a percentage-point goal, because it is a count of real visits you have to change something for.

One thing no bounce rate can tell you is whether bouncing is bad. Google's own legacy documentation put it plainly: for a site whose success depends on people viewing more than one page, a high bounce rate is a problem, but "if you have a single-page site like a blog, or offer other types of content for which single-page sessions are expected, then a high bounce rate is perfectly normal". A visitor who found the opening hours, read them and left has been served perfectly, and both definitions will record it — GA4 only if they managed it in under ten seconds.

What is bounce rate calculator?

Bounce rate is the share of sessions that ended without whatever the analytics platform counts as further engagement. The definition is platform-specific and has changed. Universal Analytics counted a bounce as a session with a single request to its servers — one pageview and nothing else — so the rate was single-page sessions divided by all sessions. Google Analytics 4 replaced that with the complement of engagement rate: a session is engaged if it lasts longer than 10 seconds, fires a key event, or produces at least two page or screen views, and bounce rate is the percentage of sessions that were not engaged. Both are computed over sessions, and in GA4 a session still ends after 30 minutes of inactivity by default. Because the two rules classify different visits as bounces, a single site migrating from one platform to the other usually sees its reported bounce rate fall without anything about the site changing.

How to use this calculator.

1. 01Pick the definition your data actually came from. If the number came out of GA4 after mid-2023, it is the engagement-based one; anything from Universal Analytics is the single-page one.
2. 02Enter total sessions for the period, then either engaged sessions (GA4) or single-page sessions (Universal Analytics). Both counts must come from the same property and the same date range.
3. 03If you only have the bounce rate and total sessions, multiply them to get bounced sessions and subtract from total sessions to get the other input.
4. 04Set a target bounce rate if you want to know what closing the gap actually costs in sessions. Leave it as is otherwise — it does not affect the rate itself.
5. 05Read the "what this number means" line before quoting the figure anywhere. It names the definition and flags the 10-second setting that GA4 lets each property change.
6. 06Compare like with like over time. A drop in bounce rate that coincides with a platform migration, a tag change or an engagement-timer change is a measurement artefact, not an improvement.

The formula.

Universal Analytics: bounce rate = single-page sessions ÷ all sessions × 100. With 33,600 single-page sessions out of 48,000, that is 33,600 ÷ 48,000 = 0.70, so 70.00%.

GA4: bounce rate = (1 − engaged sessions ÷ all sessions) × 100. With 18,240 engaged sessions out of 48,000, that is 1 − 0.38 = 0.62, so 62.00%. GA4's engagement rate is the direct complement, 38.00%, and the two always sum to exactly 100 because every session is either engaged or not — there is no third category.

The target output solves for the count rather than the rate. Sessions that must not bounce = total sessions × (1 − target ÷ 100); the shortfall is that figure minus the sessions that already did not bounce, floored at zero so an already-met target reports 0 rather than a negative. For the worked example: 48,000 × (1 − 0.55) = 21,600 sessions must be engaged, against 18,240 today, so 3,360 more are needed. Run the same target against the Universal Analytics figure and the requirement is 21,600 − 14,400 = 7,200, which is the eight-point definitional gap expressed in visits.

Rounding stage: nothing is rounded on the way through. Both rates, both counts and the shortfall stay at full decimal precision and are rounded once, at the point results are returned, to ten decimal places. The floor applied to the shortfall is a genuine maximum against zero rather than a rounding step, so a target that is met exactly returns exactly 0 — tested at, immediately above and immediately below that boundary.

A worked example.

Frequently asked questions.

References& sources.

1. [1]Google Analytics Help, "[GA4] Engagement rate and bounce rate" (article 12195621, current edition, retrieved 2026-07-29). Source of the GA4 definitions used on this page: an engaged session is "a session that meets any of the following criteria: Lasts longer than 10 seconds, Has a key event, [or] Has 2 or more screen or page views"; engagement rate is "the percentage of engaged sessions on your website or mobile app"; bounce rate is "the percentage of sessions that were not engaged". Publicly accessible.
2. [2]Google Analytics Help, "[UA] Bounce rate" (article 1009409, retrieved 2026-07-29; the page is explicitly flagged "You are viewing a legacy article about Universal Analytics"). Source of the legacy definition: "A bounce is a single-page session on your site … a session that triggers only a single request to the Analytics server", and "Bounce rate is single-page sessions divided by all sessions". Also the source of the statement that a high bounce rate is "perfectly normal" for single-page sites and blogs. Archived-status legacy documentation, still publicly accessible.
3. [3]Google Tag Manager Help, "Configure your Google tag settings" (article 12131703, current edition, retrieved 2026-07-29). Consulted as an independent second authority on whether GA4's engagement threshold is a fixed constant; it is not — "The default threshold for engagement time is 10 seconds but can be adjusted here." This is why the page treats the 10 seconds as a per-property setting and says so beside the result. Publicly accessible.
4. [4]Google Analytics Help, "[GA4] Session" (article 12798876, current edition, retrieved 2026-07-29). Restates the engaged-session rule — "An engaged session is a session that lasts longer than 10 seconds, has a key event, or has at least 2 pageviews or screenviews" — and documents the separate, configurable session timeout, noting that "the maximum value for 'Adjust session timeout' is 7 hours and 55 minutes". Used to confirm the engagement rule against a second Google page rather than relying on one article. Publicly accessible.
5. [5]Google Analytics Help, "[GA4] About Analytics sessions" (article 9191807, current edition, retrieved 2026-07-29). Defines the denominator both modes divide by: "A session is a period of time during which a user interacts with your website or app", which "ends or times out after 30 minutes of user inactivity". Publicly accessible.