Analytics and Search Marketing Tips
Bounce Rate doesn’t lie, but it can definitely DECEIVE you!
Did you know that a high bounce rate isn’t always bad? This post will help you better understand bounce rate and avoid common analysis mistakes.
Read on to understand;
The definition of bounce rate is the percentage of people who arrive on your site and leave without visiting a second page. More importantly, Bounce Rate was designed to tell you if you have the right audience coming to your pages and if you are meeting their expectations.
Or as Google analytics evangelist Avinash Kaushik frankly sums it up from the user’s perspective, “I Came, I Puked, I Left.”
To fully understand bounce rate, it is important to understand the difference between bounce rate and exit rate.
Bounce rate applies to a visit entry/landing page (i.e. the first page a person visits) and exit rate applies to the page a visit exits/leaves on. More specifically, think of exit rate as a way to identify where people are exiting mid-stream from your conversion funnel.
If most visits exit/leave at step 3 of 5, (as shown in the checkout goal funnel visualization (on the right) where the weak link is the billing information page), then you can identify that something is likely wrong with this specific step. Now you know where to start your optimization efforts.
Keep in mind that we do expect high exit rates on the final step pages such as step 5 of 5 in a checkout funnel. This doesn’t mean it can’t be and shouldn’t be reduced, but it is not a critical concern.
Now let’s get back to bounce rate…
These four visitor actions will be identified as a bounce from your site and typically signal that the visitor’s expectations were unmet.
Be aware that there are two common tracking issues that will improperly inflate your bounce rate that you need to be aware of to ensure your data is more reliable.
Clicked an external link
There are times when a visit is identified as a bounce but it isn’t necessarily negative. For example, during a visit a user clicks on a link that brings the user to an external site (this could be a partner link that you actually want people to go to or it could be another one of your own domains). One of the most common scenarios we see, is an account login that requires secure authentication and is on a separate domain. This definitely shouldn’t be counted as a bounce.
To better identify this, we recommend tracking outbound links and ensuring that you are properly tracking traffic across your various domains. Note that in Google Analytics, you can choose to have an outbound link that is tracked as event be an interaction or non-interaction event. This choice will determine whether or not a link click affects bounce rate.
Improper Analytics Tracking Configuration
If your site is improperly tagged with tracking code, you may be looking at inaccurate bounce rates. For example, if landing page X has different tracking code settings than page Y and a visitor moves from X to Y, the result will be that page X (and possibly even Y) will have an inaccurate bounce rate. The session is killed between these pages and will be incorrectly reported as a bounce. If you aren’t sure that your tracking is correct, get it audited by an expert analytics consultant.
It depends. The average website bounce rate is 40% (source: Google). But this is completely meaningless, because what constitutes a good bounce rate varies by:
But I digress and would like to give you some high level bounce rate examples for your reference.
When the visitor has a POSITIVE EXPERIENCE! There are many visitor interactions that are often undervalued and ultimately lead to desired conversions in either the short or long-term. This is hard to gauge when you are new to web analytics. I have shared some examples of acceptable high bounce rates below that should accelerate your ability to identify other examples.
Don’t think we don’t believe in reducing bounce rates everywhere. Just know that you must prioritize your efforts. You shouldn’t prioritize one of these example pages over top landing pages in your conversion funnel that have high bounce rates.
Remember that your purpose is to serve the customer and every visit doesn’t need to generate a lead or transaction.
Marketers should embrace this and understand that customers have many interactions with a company which should be seen in a positive light, even when the business doesn’t see immediate direct value. These touches support the visitor’s needs throughout the complete customer lifecycle and lead to desired goals and most importantly, acquisition, retention and brand affinity.
Further, these visits can be tracked and measured with engagement goals that take into account and demonstrate the value of frequency, recency, loyalty and etc. Better yet, attribution modeling can demonstrate how these early stage visits lead to transactional conversions.
Do you know how to determine a good or bad bounce rate? Are you sure?
After being an analyst for many years, I can tell you that I misinterpreted bounce in my early days and I run across many people who don’t spend enough time in analytics tools to know how to accurately interpret this metric. The primary reason is context.
For example, you might see a high bounce rate on a high traffic page and rush to the judgment that it is performing poorly and needs to be improved. When in fact everything is fine.
How can this be, you ask? I have provided a case study below of a bounce rate that would be easy to misinterpret on the content/pages report.
Did you know that a 100% bounce rate isn’t always bad? This could mean that only 1 person bounced from your page while thousands funneled through the page and had a great experience.
When bounce rate is being analyzed in the ‘Pages’ report, you need to be aware that bounce rate loses important context because it lacks the entrances metric. Instead, leverage the ‘Landing Pages’ report to view the bounce rate relative to the number of entrances.
Better yet, build yourself a custom report that shows the number of bounces in addition to the bounce rate as well as other metrics that are important to you.
Consider this case study on bounce rate deception. You see a high bounce rate on the content ‘Pages’ report.
The bounce rate is 63% and only 13% of visits exit on this page. Considering that this page is not a primary entry point and is a supporting tertiary shopping funnel page, we are more concerned with people exiting at this point versus bouncing. And approximately 1 out of 10 people exiting, is actually great on this high commitment page.
Let’s cross-reference the bounce rate with the number of visits who entered on this example page. As described below, we calculate that only ~2% of total pageviews for this page bounced! That doesn’t sound near as bad as 63% does it?!
The bounce rate only applies to visit entrances which was a total of 229. 63% of 229 is only 144 visits out of the total 8,634 pageviews.
Again, this is the equivalent of only 1.7% of total pageviews that actually bounced.
Most people incorrectly assume that it is 63% of 8,634 (5439) pageviews, which would be a major concern. This is roughly a 37x difference in the assumed severity of the issue.
(Note: entrances are similar to pageviews but be aware that this technically isn’t an apples to apples metric analysis.)
Keep in mind that user expectations and intent varies based on many factors. To avoid bounce rate from deceiving you, we highly recommend segmenting your data. In aggregate, the bounce rate can look good or bad, but be just the opposite or be hiding substantial problems.
Here are a few segmentation examples to consider that can help you better evaluate your site and marketing performance.
Want more tips on segmentation and how to setup segments? Check out this post: Don’t be misled by averages – Use advanced segments
If so, leave a comment. If nothing else, we hope this post will prevent you from being fooled by the bounce rate on the ‘Pages’ report.
In case you didn’t get enough on bounce rate and exit rate check out these related posts:
Make it a great day!
Share this Post
Kayden is the founder and Managing Director of Blast Analytics & Marketing. He lives to do better and loves sharing his ideas and experiences. Kayden uses his seemingly infinite knowledge of web-based technologies to lead from the front and constantly push the edge of what is possible.51 posts on the Web Analytics Blog.
Web Analytics Blog
We're here to help with tips and insight on the following topics:
Optimize your website and marketing campaigns