Analytics Focus: “The Sexiest Metric Ever”

High bounce rates indicate issues on your siteThis post marks the first in a series of posts I’ll be doing in the coming months, focusing on one specific web metric and how you might want to use it.  I think this series will be of a lot of use to the Web Managers out there who are responsible for the bottom line generated by their web site.

Today, I’m going to talk about what Google’s Avinash Krausik calls the “sexiest metric ever” and one that you might just be ignoring completely.  I’m going to be talking about Bounce Rate.

Exactly What Is Bounce Rate?
Typically, a “Bounce” is defined either as a visit that consists of only one page viewed.  Most analytics packages, including Google Analytics, define a bounce this way.  Some more aggressive definitions include any visit that is five seconds or less, regardless of whether there was a second page clicked.  In any case, a bounce is an instance of someone coming to your site and realizing leaving without exploring the site to any meaningful extent.

Bounce Rate is therefore the number of bounces divided by the total number of visits, expressed as a percentage.  Like all metrics, it’s tough to give a single benchmark, but I like to feel that for a site’s overall bounce rate, somewhere in the neighborhood of 50% might be considered an overall web average.

Of course, bounces are generally not good.  If you’re trying to make a conversion, such as an online order or a newsletter signup, or almost anything, your visitor will have to spend more than five seconds and visit more than one page to do that.

It should be noted that not all bounces are bad things.  Reference sites will often have a higher than average bounce rate.  I would bet that Wikipedia’s bounce rate is pretty high.  If you search on “Peyton Manning”, Wikipedia will probably be one of the top results, and most of the people who click on it will either find or not find the information they need, but in either case will be unlikely to click around to other Wikipedia links.  Typically, you enter Wikipedia and other reference sites on the page you needed, and leave thereafter.  But for most, a high bounce rate is a cause for concern.

Why Is It So Important?
Why does Avinash call it sexy?  Well that’s easy.  Your bounce rate can really help you hone in on what is holding your conversions back.  If your site’s conversion rate is 2% and your bounce rate is 80%, then really only 20% of your visitors even have a chance at converting.  Actually, with an 80% bounce rate, I’d be reasonably impressed with your 2% conversion.

So okay, a high bounce rate is a problem.  Is that any “sexier” than Cart Abandonment Rate or any other negative metric?  Absolutely!  And the reason is because with a little bit of refinement, it can point out so many things you can do to improve your site (and therefore improve your conversion rate).

So What Can It Tell Me?
While looking at your site’s overall bounce rate can quickly indicate that you might have a problem, the real magic comes in segmenting out your bounce rate by slicing it in different ways to see what turns up.  Consider:

  • Bounce Rate by Landing Page:  Focus on your top landing pages by Entry Views.  Do some pages outshine others?  What is different between the high bounce rate landing pages and those with lower bounce rates?  Are the Calls To Action noticeably different?  Are there differences in page layout, organization of copy, or graphics?
  • Bounce Rate by Keyword:  This is especially important if you have large PPC spends.  But whether paid or organic, are poor performing keywords sending traffic to the most appropriate pages?  For keywords with higher bounce rates, try to redirect them to different landing pages.  This segmentation can also help identify some strong keywords that are not being emphasized enough.
  • Bounce Rate by Referring Source:  Do you have sources with high volume and high bounce rates?  If so, perhaps you should spend less on those sources or reevaluate the relationship with them.  Cross-reference this with landing page bounce rates as well.

How Can I Fix It?
A high bounce rate seems like a problem, but to me, it’s really an indicator of a different problem.  It’s a symptom.  So use it to diagnose the real problems of your web site, such as substandard landing pages that should be redesigned, poor Calls To Action, PPC campaigns that are wasting money, etc.)  Addressing problems like these will have the double-effect of improving bounce rate and improving conversion rate, which is the real goal.

Still, there are some quick things you can look at even before you start getting into the nectar of the bounce rate segmentation.  The most important of these is to check your site in all major browsers and platforms.  How well does your site work on tablets and smartphones.  Do you rely on Flash?  If so, you’re probably bouncing 100% of your iPhone, iPad, and Android visitors.

Also, check your page load speed.  If your site takes a long time to load, many people might just not want to wait around for it and bounce before it does.

A Word About Blogs
As I said above, for some sites, a high bounce rate is to be expected and is not necessarily an indicator of a problem on its own.  This can be the case for many blogs.  Many outstanding blogs will appear to have a high bounce rate because they have a great following of people who come each day to look for and read the newest post and then leave without ever clicking on a second page.  That will produce a high bounce rate, but it is an enviable situation.  Such a dedicated following is every blogger’s goal.  In cases like this, you would want to segment out your bounce rate by new versus returning visitor.  The scenario described above is for returning visitors, and a high bounce rate is perfectly okay for them.  But for first time visitors, even if they arrived at just the article they were looking for, you want them to at least explore the site a little bit to understand what other kinds of articles you write, etc.  Doing this would increase the changes that they become one of those loyal followers.

This entry was posted in Analytics. Bookmark the permalink.