bounce rate problemsMany people point to the bounce rate in Google Analytics and assume that those who bounced are worthless visitors. This is not the case.

Before I explain why, I want to say that I have changed the GA code so that Bounce Rate is calculated from those who spent less than 10 seconds on the site, not those who viewed one page, which is the default. The reason I did this is that many of my sites are blogs or are using blogging extensively. If a visitor enters on a category page, they can spend up to 10 minutes reading posts, yet the default GA code only sees one page viewed, and says that they bounced.

If you would like to use this bounce rate, you need setTimeout, and here the code and where to place it:

var pageTracker = _gat._getTracker(“UA-xxxxx-x”);
setTimeout(‘pageTracker._trackEvent(‘NoBounce’, ‘NoBounce’, ‘Over 10 seconds’)’,10000);

Now, why are visitors who bounce not worthless? Simple, GA seems to be calculating bounces off of the most recent visit. This does not mean that the visitors have not spent time on your site before. Here is a screen shot of the Visitor Loyalty report, segmented by those who bounced.


Bounce Rate and Visitor Loyalty

You can see that although the majority (93.5%) of visits were the 1st time to the site, around 6.5% of visitors were returning, some of them up to 9 – 14 times.

This means that these returning visitors liked what was offered on the site enough to come back, but then why did they bounce?

I believe that many of the Loyalty Bounces are looking to copy URL’s to send to other people. Yes, that may sound like wishful thinking, but when I go back to a site and leave immediately, it is usually for that very reason. Another reason may be to verify some piece of information. Every case is different, which is why Custom Segments are so important for your analytics.

To see this in action, you will need to create a custom segment by clicking on Advanced Segments in the left hand navigation of GA. Originally I used Metrics-Site Usage-Bounces with the Condition Greater than or equal to, and the Value as 1. This gave me the bounce rate of the site. Then I created another segment to combine bounce rate and returning visitors;

Dimensions – Visitors – Visitor Type – Matches Exactly – Returning Visitors

Metrics – Site Usage – Bounces – Greater than or equal to – 1

There you have it. You created a custom segment, and you can see what I call a Loyalty Bounce.

Update: When I applied this to another of my sites, I believed it confirmed my suspicions about why people are returning to the site, then bouncing. The Top Landing Pages report segmented by Loyalty Bounces was very similar to the Top Content report for all visitors. The pages that are most popular on my site are the ones giving me the Loyalty Bounce.

So a visitor that bounces is not always a bad thing. They may be spreading links.