Explaining the Bounce Rate
February 22, 2019
The bounce rate of a given website is a key metric that can help us understand how people interact and engage with that website. While important, the bounce rate is often used incorrectly and frequently misunderstood. So, what exactly is the bounce rate, why should you use it, and what is a good bounce rate you should aim for?
Defining the Bounce Rate
Let’s begin by defining what exactly the bounce rate is. At its simplest, to say a visitor bounced away from the website means that person looked at one page of a website, did nothing else and then left the website. If 100 people visit your website and 60 of them leave after viewing just one page and do nothing else (that is, 60 of the 100 visitors bounce away), your website would have a 60% bounce rate.
It’s about interactions, not pages…
Of course, that definition is a bit simplistic. When measuring the bounce rate in Google Analytics (and other web analytics tools), what you are actually measuring isn’t people who visit only one page, but instead measuring people who only have one type of interaction. Visiting a page is only type of interaction, but there are other types of interactions that can be tracked in Google Analytics—the most common of those interactions is events.
You can use event tracking to measure different types of interactions (or different ways people engage with a website). This allows you to understand more about how people use your website beyond simply looking at pages. When you use event tracking in Google Analytics, you get to decide if the event you are tracking should affect the bounce rate or not. For example, you may want to know how many people scroll down the page and you can collect this scrolling interaction with event tracking. The question is, if people interact in this way should it affect your bounce rate?
Google Analytics lets you decide. Events can be configured to affect the bounce rate (by setting non-interaction to false) or may not affect the bounce rate (by setting non-interaction to true). If you are using event tracking, it is extremely important you intentionally decide which setting to use so that you are in control of how your bounce rate is affected.
So, in that scrolling example, if people scroll down a page, should that reduce your bounce rate or not? Let’s say we decide scrolling should affect the bounce rate. In that case, if 100 people came to your website, 60 left after viewing only one page, your bounce rate would no longerbe 60% because your bounce rate would now account for the people who scrolled on the one page they viewed. Maybe of the 60 people who left after viewing just one page, 30 scrolled. In that case, we have 40 people who viewed multiple pages and 30 people who scrolled while only viewing one page—or 70 people who didn’t bounce in our modified definition. With that modification, we’d have reduced the website’s bounce rate from 60% to 30%.
Bounce Rate TL;DR
Even once you factor in the ability to redefine the bounce rate with event tracking, the idea remains the same: people who come to your website, interact in just one way, and then leave are considered to have bounced away. The bounce rate is a ratio of how many visitors bounced.
Locating the Bounce Rate in Google Analytics
Now that we know what the bounce rate is, where do we find it in Google Analytics? The bounce rate can be seen in different contexts. You can locate the website’s overall bounce rate on the Audience Overview report, you can view the bounce rate by traffic source under the Channels or Source/Medium reports, or you can view the bounce rate for a given page on the All Pages report.
This level of granularity is helpful. Your overall website’s bounce rate may be 75%, but you might find that traffic arriving on your website from a Google organic search only bounces 50% of the time. This would indicate traffic from Google organic searches might be more interested in what your website has to offer and might be more willing to engage. Equally, you may find that some pages on your website have higher bounce rates than others—indicating that people who visit certain pages may be more or less likely to engage.
Bounce Rate vs. Exit Rate
As you review Google Analytics, you’ll start to notice there is another metric that seems suspiciously similar to the bounce rate—the exit rate. Although there are similarities in what these numbers tell us, and although both metrics can help us understand how people engage with a website, these metrics are two different things.
The exit rate is a ratio telling us how many people who viewed a given page on our website left the website from that page. If 10 people visit our home page and 4 of those people leave our website from the home page, our home page would have a 40% exit rate.
That means, if somebody comes to your website, visits Page A and doesn’t interact in any other way, then leaves, they are considered to have bounced away. As well, that person who left also exited your website from Page A. But, other people enter the website on Page A and continue forward from there, looking at Pages B, C, D, and so on. Those people who entered on Page A will eventually exit the website from some other page—but when they exit from another page, they won’t have bounced away because those visits other pages on the website were interactions with the website.
In short, when it comes to the bounce rate vs. exit rate, the bounce rate tells us something about how sessions begin and exit rates tell us something about how sessions end. By looking at the bounce rate, you can determine how many people came to your website and left before they interacted with your website. By looking at the exit rate, you can see what pages people are on before leaving the website.
As one more point, remember that there is no sitewide exit rate. That’s because the sitewide exit rate is 100%; everybody exits a website eventually but not everybody who visits will bounce away from a website.
What is a Good Bounce Rate?
The short answer is a good bounce rate is one that decreases over time. A decreasing bounce rate means that the changes you are making to your website are helping more people find ways to interact with your website. This is especially true when considering the sitewide bounce rate.
A good bounce rate at the page level will vary depending on the type of page in question. A blog post designed to answer questions quickly may have a very high bounce rate since people not need to interact beyond the single pageview to answer their question. But you’d want a considerably lower bounce rate for a product detail page located on an ecommerce website because people can’t buy that product if they are bouncing away.
The same is true of bounce rate by traffic source. Some sources of traffic will probably lead to more people bouncing away. Test different traffic sources and use the bounce rate as one metric to understand the results you got from that traffic source.
There simply aren’t any set numbers for what a good bounce rate should be and anybody who says otherwise is trying to oversimplify reality. That doesn’t mean a bounce rate is useless—far from it. Rather, this means that the bounce is relative to your website.
Given that, you need to clearly understand what engagements look like on a website and what your expectations are for how people engage. With that in mind, you can look at the bounce rate alongside other metrics to get a realistic portrayal of how people are engaging with your website and if the way people engage meets your expectations. For some websites, that means a good bounce rate is over 70% and for other websites that means a good bounce rate is well below 50%.
If you have questions about your website’s bounce rate or need help improving how people use your website, let’s talk. Please contact me today.