This week I was pleasantly surprised by an old blog post on charity bounce rates getting picked up. It sparked a conversation with @charitychap and @LondonKirsty about reducing your website bounce rate that I thought worth elaborating on.

Having assumed a view on average bounce rate for charity websites from my old post. We all agreed the key challenge is that measuring your bounce rate might be easy (and is critical) but it’s harder to know how to reduce it apart from following the generic tips out there. The conversation got me thinking about the analysis we started to consider in my last job.

We got to the stage of considering how best to segment our view of bounce rate. This came from the recognition that taking a site wide view of bounce rate might work for some websites which have a single purpose, but many charity websites have multiple purposes and so a single view isn’t good enough.

This means a good definition of your key audience groups and their goals is very important. Along with recognising that goals may vary by personal circumstance and time of year. For example, for health charities there is a clear ‘patient pathway’ view that can be taken – from diagnosis, through treatment, to recovery, and often to maintenance.

But how exactly can you take a segmented audience view of bounce rate without requiring all of your users to be logged in or personally identifiable? Well – we didn’t find a concrete answer to this. But here are some ideas:

• Use your content as an approximation of the audience; look at bounce rate by section rather than site wide and make assumptions about the audience consuming that particular section and how you might cross-sell or up-sell to that group.
• Use the traffic referral source as an approximation of the audience; look at bounce rate segmented by traffic source and see where you can make assumptions about the audience based on this eg those from BBC Vs The Sun, Google Vs Bing are different demographics.
• Segment those that don’t bounce from those that do; these are two high level segments that could shed some light on things when looked at within a content section.

Finally, of course, the best way to reduce bounce rate is to test, test and test again. I’m not sure enough testing of the ‘bread-and-butter’ online activity happens in charities. But the surest way to find out what improves your bounce rate is to test variations and find the winning combination until you start to spot potential to improve further.