What’s the deal with bounce rate?
In my years working in website analytics, Bounce Rate has consistently been the most difficult website behaviour metric theme to broach with non-data savvy colleagues.
Many conversations start with “what is a Bounce Rate?”, “Why is the Bounce rate so high!?” or “the Bounce Rate is zero, should we pop a bottle of sparkling”.
So what IS the deal with Bounce Rate!?
If you are using Google Analytics or Adobe Analytics you will recognise the Bounce Rate metric.
Both Adobe and Google calculate a bounce as a website visit that triggers only a single request to the Analytics server, such as when a user opens a single page on your site and then exits without triggering any other requests to the analytics server during that session.
Bounce rate is then calculated as the number of bounces divided by the number of visits or sessions to your website.
Knowing this, what is the bounce rate of a great website?
This is actually a very tricky question, as an ideal bounce rate is dependent on the function of your website and the pages users are landing on.
Below is an image from Similarweb.com showing the bounce rate of some category leaders for various industries.
As you can see the Tesco.com bounce rate is over 20% lower than that of Wikipedia.org, however, this doesn’t mean Tesco has a better website.
If you think about the purpose of a user visiting Wikipedia, it may just be to discover how old a celebrity is or where a foreign city is located. The user would land on the page and scroll to the desired information, read it, and then leave the site; registering a bounce. In doing so they have not interacted with the page, however, this would still be a quality session in the eyes of the Wikipedia web analytics team.
A user on the Tesco site is more likely to interact with the page, click through to a product, locate a store, or make a purchase. Therefore the Tesco web analytics team would view a bounce as a low-quality session.
With this in mind, when you look at your own bounce rate you need to consider the function of your website and the purpose of the page the user has landed on. If you want the user to interact with the page then a low bounce rate is important, but if you just want to expose them to information on the landing page then a higher bounce rate is not so consequential.
- An example of when a high bounce rate is worrying would be on a home loan application page. If users are leaving before interacting with the page and entering any of their information, it could be a sign they are not finding what they are looking for, the user experience is bad or the interest rate is too high!
- An example of when a high bounce rate is not very important is on a news article page. Users can read an entire article and be exposed to any advertising housed beside the content without interacting with the page. In this case, a bounce would still be fulfilling the desires of the reader, writer, and advertiser.
Wait there is one more spanner in the works, a very low bounce rate is also potentially a bad thing!
If your bounce rate is below 25%, this is a sign that you may have additional events automatically firing in the data layer after the page has loaded. Depending on the event that is automatically triggering a request to your analytics platform, the bounce rate data being reported could be inaccurate.
- An example of a good auto-firing event is a 2-minute session timer, which would indicate the user has been reading your page for 2 minutes.
- An example of a bad auto-firing event is an autoplay video starting, as the user might not even see it.
If you are still can’t quite narrow down why your bounce rate is the way it is, you might be looking like Nick Young below.
If so, you should set up a visual behavioural analytics tool on your website to garner additional insights.
Tools such as Microsoft Clarity and HotJar have the capability to record videos of live user sessions, documenting the user web portal view and cursor movements of visit.
With live user session recordings or screen recordings, you can view exactly what users are doing when they land on your website and watch their on-screen behaviour to identify trends that could help explain your bounce rate.
More information on these tools can be found here:
Hopefully, now you understand exactly what the deal is with bounce rate; the true meaning of bounce rate is dependent on the purpose of the landing pages or website.
So next time anyone asks you why they are seeing a bounce rate of XX% you can give them a consolidated slam dunk of an answer.
If you have any questions, please reach out to Orchard and their data team for help.