Happy Easter! Or, should we say Hoppy Easter! With the chocolate-filled, bunny-hopping occasion just around the corner, we thought it would be a perfect time to talk about a digital analytics fundamental related to hopping: bounce rate.
By definition, bounce rate is quite simple – it is the percentage of single-page sessions. More specifically, it is the percentage of visits in which someone lands on your site and leaves without any further interactions on the page. Succinctly put, Avinash Kaushik refers to the customer experience of a bounce as I came, I puked, I left.
But does your bounce rate really measure this?
Ideally, you would like your bounce rate to tell you the percentage of people who aren’t interacting with your site. But this relies on you telling your web analytics tool what you consider to be an interaction.
By default, Google Analytics and other analytics tools will count any single-page session as a bounce. But what if that single page contains everything you want your visitors to accomplish?
- Do you have long pages where your visitors could spend 5, 10, 15 minutes or more reading content?
- Do you have forms that submit dynamically without a subsequent thank you page?
- Do you have videos, expandable accordion sections, or other interactive content?
- Do you have links to partners, subsidiaries, or other external sites?
If you don’t configure custom tracking for these types of interactions, they will be invisible to your analytics tool. Your visitors could be reading, scrolling, watching videos, submitting forms, and interacting with dynamic content on the page, while your analytics tool counts them as bounces.
Your analytics may show that 90% of your visitors are “bouncing,” but this could include a significant proportion of engaged single-page sessions. So how do you distinguish these from the “true” bounces?
This is where an adjusted bounce rate can be useful. You can “adjust” your reported bounce rate by ensuring all meaningful interactions are tracked. Depending on the nature of your website, you may wish to track an interaction when visitors spend at least a certain amount of time on page (e.g. 30 seconds), scroll to a significant depth (e.g. 75%), or complete a desired action (e.g. submit a form).
If you are using Google Analytics, you can track these actions as interactive events or “virtual” pageviews. When one of these interactions occurs, GA will then no longer count that session as a bounce, even if only one “real” page was viewed.
By adjusting your bounce rate, you will have a more representative measure of your visitors’ behaviour, which will help you do better analysis and gain better insights from your analytics!
How do I analyze my bounce rate?
Once you are comfortable with what your bounce rate is actually measuring, you can potentially do some useful analysis. But like most other overall site metrics, it is important to look beyond the site wide average and segment your data. There are likely pages, channels, campaigns, or device categories that are significantly above or below the average.
When looking at bounce rate in GA, have a look at the following reports rather than just the site wide average:
1. Acquisition > All Traffic > Channels
- This report segments traffic into the Default Channel Groupings and allows you to see the bounce rate for each. In the example above, you can see that Paid Search visitors have the highest bounce rate (24%), while Social has the lowest (9%). These facts are obscured if you were to only look at the overall average of 11.24%. Assessing bounce rate per channel and campaign can help you optimize your marketing initiatives.
2. Audience > Mobile > Overview
- This report shows you the device type used to view your site, either mobile, desktop, or tablet. In the example above, you can see that the bounce rate for mobile is higher than desktop. Based on this, you may wish to investigate if your site experience could be improved for mobile users.
3. Behaviour > Site Content > All Pages
- This report shows you the bounce rate for each page on your site. Keep in mind that the bounce rate for a page is only proportional to the number of sessions that started on the page. So, you should also pay attention to the number of entrances to keep the bounce rate in context.
Have you had any challenges with bounce rate? Let us know!