Bounce rate is one of the most quoted and most misunderstood numbers in website analytics. A high figure is often treated as a warning sign, yet for many pages it is perfectly normal. This guide explains what bounce rate measures, how different tools calculate it, what ranges you can reasonably expect, and how to use related figures such as pages per visitor to judge whether your content is doing its job.
What bounce rate measures
In its classic form, a bounce is a visit in which someone views a single page and then leaves without viewing another page on the same site. Bounce rate is the share of all visits that were bounces:
bounce rate = single-page visits ÷ all visits × 100If 400 of 1,000 visits ended after one page, the bounce rate is 40%. The definition sounds simple, but each part of it depends on decisions made by the analytics tool: what counts as a visit, when a visit ends and whether some activity on the page, such as watching a video, should count as engagement even without a second page view.
Why different tools show different bounce rates
Not every tool uses the classic definition. Google Analytics 4, for example, defines bounce rate as the opposite of its engagement rate: a session counts as engaged if it lasts longer than a set time, includes a conversion event or has at least two page or screen views. A visitor who reads one long article for three minutes and leaves is a bounce in the classic sense, but not a bounce in GA4.
Because of this, comparing bounce rates between tools, or between your site and a figure from an article, rarely makes sense. Compare a page with itself over time, or compare similar pages measured by the same tool.
Is a high bounce rate bad?
Not necessarily. The right question is whether the visitor got what they came for. Consider these examples:
- A recipe, definition or opening-hours page: visitors find the answer and leave satisfied. A high bounce rate is expected.
- A single long article from a search result: readers may read everything and close the tab. That is a successful visit even if it is recorded as a bounce.
- A home page or category page: its job is to send people further into the site. A high bounce rate here deserves attention.
- A product or sign-up page: if visitors leave without the next step, you may have a problem with the offer, the price, the form or the page speed.
A useful rule: judge bounce rate against the purpose of the page, not against a universal target.
Typical ranges and why to treat them with care
You will often see broad ranges quoted: roughly 25–55% for many business and e-commerce pages, and 60–90% for blogs, news articles and reference pages. These ranges can help you notice extremes, but they are not standards. Traffic mix has a large effect: visitors from search engines who land on a specific answer bounce more often than visitors who arrive from your newsletter and already know your site. A sudden move, up or down, is usually more informative than the absolute value.
Unusual values that point to a measurement issue
- A bounce rate near 0%: often means the tracking code runs twice on each page, so every visit looks like two page views. Check that the code is not installed in two places.
- A bounce rate near 100% on every page: may mean the code is only installed on landing pages, so the second page view is never recorded.
- A sudden jump after a redesign: the new template may be missing the tracking code on some page types.
- Many very short single-page visits: may come from automated traffic rather than people. Our guide to traffic sources explains how to spot suspicious referrers.
Pages per visitor: a clearer companion figure
GNRCounter reports visitors and page views, and the dashboard shows pages per visitor: page views divided by visitors for the selected period. It answers a related question in a simple way: on average, how many pages does a visitor open?
- A value close to 1.0 means most visitors look at one page. That is normal for a site that mainly receives search visitors to individual articles.
- A value of 2 to 3 or more shows that people often continue to other pages, which is what you want from home pages, catalogues and documentation.
- A falling value after a change to your navigation or layout is a signal to review that change.
Because pages per visitor uses only two counts, it is easy to understand and hard to misread. Look at it together with your visitors and page views for the same period, and with the list of your most viewed pages.
How to improve engagement on the pages that matter
If a page should send visitors further and it does not, these changes usually help more than any general trick:
- Match the page to the search or link that brings people in. If the title promises a tutorial, the tutorial should start near the top, not after a long introduction.
- Make the next step obvious. Add a clear link to the most relevant next page: a related guide, a product category or a contact form.
- Link inside the text. Readers follow links that appear where a question naturally arises far more often than links in a sidebar.
- Improve loading speed. Visitors on mobile networks leave slow pages before they finish loading. Compress large images and remove scripts you no longer use.
- Check mobile layout. Overlapping elements, tiny buttons and intrusive pop-ups push people away quickly.
- Remove distractions near the main content. Too many competing boxes make it hard to see what to do next.
Change one thing at a time and compare the same period before and after, for example two full weeks, so that weekday patterns do not mislead you.
Avoid inflating your own numbers
Your own visits can distort small sites. If you browse your pages while editing them, you add page views and reduce the share of single-page visits. GNRCounter lets you exclude your own browser: open any tracked page once with ?gnrc-hideme=1 added to the address, for example https://example.com/?gnrc-hideme=1. The browser is then excluded from counting on that domain. To undo it, open a page with ?gnrc-hideme=0. The setting is stored in a cookie, so repeat it in each browser or device you use and after clearing cookies.
Frequently asked questions
What is a good bounce rate?
There is no single good value. A reference page with a 75% bounce rate can be doing its job perfectly, while a home page with 60% may need work. Compare each page with its own history and purpose.
Does bounce rate affect search rankings?
Search engines do not read the bounce rate from your analytics tool. What matters for search is whether your page satisfies the people who arrive. Improving that usually improves engagement figures as a side effect.
Why is my bounce rate different in two tools?
The tools define visits and engagement differently, may count bots differently and may be blocked by different browser extensions. Use each tool consistently instead of trying to make their numbers match.
Can a visible counter change bounce rate?
No. Whether the counter is visible or hidden, the same page views are recorded. Display choice changes the layout, not the measurement.
Next steps
Start with your most viewed pages and ask, for each one, what the visitor should do next. Then watch pages per visitor for those pages over a few weeks. To understand the terms behind these numbers, read sessions, visits and users explained, and to see where your single-page visits come from, use the traffic sources guide. If you do not have a counter yet, create one for free.