Bounce Rates – Understanding what it is telling you

Numberminds Bounce RateBounce rate is a key metric for anyone studying web analytics.  It allows you to understand what pages are causing people to leave your site.  Or is it?  If people come to your site to read a blog post and then leave does that necessarily mean that the website failed?  If the customer lands on a marketing landing page and then doesn’t engage further with the campaign (at that time) does that mean the website failed?

There is so much supposition in any metric that its key for any web analytics (or management information professional) to be able to explain and add context.

So if you write a blog post you write something that is of interest to the masses.  Lets take this blog as an example.  I want to tell people:

  • What a bounce rate is
  • How to use it in your reports
  • What are the pitfalls

So you should have a reasonable idea already what the bounce rate is.  So lets see how you can use it more effectively.

Use of Bounce Rate in Reports

I love creating reports which show insight.  Segmenting report data – look at the areas of your site.  Your blog.  Your whitepapers, Your secure sales site, Brochure Ware Site, Landing Pages etc.

This will allow you to segment your website into areas where KPIs can be relevant.  So KPIs for the blog would be seperate to those from the secure site – the KPIs themselves are the same but the numbers are only relevant to the specified areas of interest.

If you wanted to add context then you could do a whole website KPI view and benchmark individual areas against it.

Bounce Rate Graph Example

So you’re seeing the same information but segmenting it by similar areas.  You can also see a comparison against the whole site.  I’ve also shown that having a 7 day average and a last month daily average gives you the insight to see (on a rolling basis) whether you’re yesterday’s stats are under/over performing.

Please see an upcoming blog-post on content grouping.  This will help you understand how it is possible to group together pages and analyse these.

Pit Falls

Having a high bounce rate isn’t necessarily a bad thing.  Bounce rate calculations in Google Analytics only calculate if a page had a single pageview in the visit/session.  It doesn’t add in the ‘time on site’.  This is an essential element when adding context to report.

You need to educate your report audience.  If the audience doesn’t understand the behaviour of people on particular pages then they can try and ‘blame’ or ‘berate’ your results.  I would recommend that you explain that people reading blog posts may only be interested in the article.  They may not have an interest in your product – at that time.  The important thing is that the branded blog pages are keeping the brand infront of your potential customers.

If on the other hand they were to be interested in your products and landed on a product page directly but, the page itself didn’t lend itself easily to converting customers (or acting as an influencer to purchase decisions) then this is an action you can take away.  Perhaps the products terms and conditions are too long and would be better served in an easily downloadable PDF.  Perhaps the benefits of the product aren’t being promoted effectively or perhaps the client was looking for third party reviews (a great influencer) and left to go to a competitor site.

This is delivering insight which can be actioned on.  This is delivering something which will easily have a bottom line impact (with and without improvements).

When you make changes to web pages (in content, design or structure) it is important to known when these changes were m

Conclusion

If you’re not working on some form of scorecard then it will be impossible to answer questions of stakeholders and senior managers.  You need to have metrics like bounce rate but, with context.  Segment the pages according to the behaviour.  If you have a product which has been re-called for safety reasons then this may have served its purpose and the visitor may not want to see anything.  Was this a successful or unsuccessful visit?  I would say the former.  However, if you didn’t include this behaviour into your metrics then your bounce rate would crash through the floor.  In short add context to your bounce rate metrics!

 

 

Leave a Reply

Your email address will not be published. Required fields are marked *