An exit is the metric referring to the number of times visitors have left a site from a specified page.
The page’s exit rate indicates how often visitors exit from it after visiting any number of pages on the site; as a percentage, exit rate is calculated as the number of exits / number of pageviews for the specified page.
For example: a visitor lands on the homepage of an ecommerce site, navigates to a category page, then to a product page, and leaves. That’s an exit on the product page. The overall exit rate for the product page is then calculated as number of exits / number of pageviews the product page received.
To see exit rate in GA, visit Behavior > Site Content > All Pages and look at each page’s % Exit number.
Monitoring exits helps you understand the performance of specific pages and/or page groups on your site: when you understand which page(s) people are exiting the most on and at the highest percentage (i.e. the exit rate), then you know which pages might need improving the most.
In the sentence above, we wrote ‘might need improving’ because, quite obviously, people eventually have to leave a website—so the context of where this happens is important. For an ecommerce website, a high exit rate on a thank-you page isn’t something to worry about, because people have completed their action successfully and there’s little left for them to do but leave. However, a high exit rate on the last step of a checkout process (for example: a payment page) could be a sign that something is wrong and needs investigating.
Before talking about how to do that, we’re taking a quick detour to address a frequently asked question about exits:
As metrics, exits and exit rate report on the action of a visitor leaving a page. Two other GA metrics, bounces and bounce rate, also report on people leaving—but only do so in a specific scenario, which we show you below.
First, a reminder: exits are recorded when people leave a page; the page’s exit rate indicates how often visitors exited from it after visiting any number of other pages on the site in the same session.
For example, if people visit the Hotjar homepage → the blog → a specific blog post and leave from it, the exit is counted on the blog post. The blog post’s exit rate is then calculated as number of exits / number of pageviews to the page.
Unlike exits, bounces measure the percentage of visitors who, after landing on a page, leave it without taking any further action (such as clicking a link or navigating to a second page).
For example, a visitor who lands directly on the blog post, reads it, and leaves counts as a bounce. Bounce rate for the page is then calculated as the percentage of all sessions that started and ended on the page / all sessions that started on the page and continued towards more pages.
Technically speaking, the bounce from the blog post is also an exit, in that the visitor did leave the website; but because the visitor exited on the same page they entered from without going anywhere else, it gets recorded as a bounce.
Another way of saying it: all bounces are exits, not all exits are bounces.
As a traditional web analytics tool, GA will show you that something is happening on your website exit rate-wise (remember: Behavior > Site Content > All Pages), but it will not be able to address your next two questions:
To answer both, you need to collect additional context and information. Here, we cover three behavior analytics and feedback tools that would be immediately useful when investigating why exits are happening and how to fix them: heatmaps, session recordings, and on-site surveys.
What: once you know that your visitors are leaving from a specific page, you’ll want to investigate what may be working, or not, on the page itself.
How: as people visit the page and data gets collected, your Heatmap tool will start rendering the behavioral data points visually via scroll maps, which you can review to see how far down the page people scroll (or don’t), and click maps, which show you what element(s) on the page people are interacting with:
Based on the evidence you accumulate, you might already be able to spot behavior patterns that explain the high exit rates: for example, people missing information that sits below-the-fold on the page or clicking on non-clickable elements and becoming frustrated.
If you are using Hotjar: from the Heatmaps dashboard, create a new heatmap and enter the specific page URL during the ‘Targeting’ step, then wait for the tool to collect behavioral data such as clicks, taps, scrolls.
What: to get more context, you’ll also want to review the wider journey(s) which terminate on the page.
How: using session recordings, you’ll be able to rewatch user sessions that end on the drop-off page and get a better sense of whether visitors were:
If you are using Hotjar: after setting up the Session Recordings tool from the dashboard, which allows it to start recording sessions from the users who come to your website, filter your existing recordings by adding the URL of the page into the Exit URL field.
What: using heatmaps and recordings may have given you a few clear ideas of what’s happening, but instead of making assumptions you can ask your users directly.
How: place an on-site survey on the page and set it to ask a specific question, such as:
If you are using Hotjar: set up a Feedback Poll and choose to show it on the specific page URL:
After you’ve collected some answers, analyze them and start looking for patterns or insight that might point you in the direction of a solution.
Content credits: we created this piece with the help of web analytics and CRO expert Rich Page, who was interviewed by our team and talked to us about all things GA in August 2019.