When A High Bounce Rate is a Good Metric
Most people examine bounce rates and immediately assume that if the bounce rate is high, the page or traffic source is bad.
That’s not always true.
First, bounce rates do change by analytics systems. Some use true time on site based upon server and browser behavior. Others, such as Google Analytics only look at when the script executes.
So, let’s say you are using Google Analytics and someone comes to your site. When the first page loads, Google knows you have a visitor and how the visitor got there, along with a lot of metrics about browsers, operating systems, etc.
They have no idea how long someone is on that page until that user goes to another page and the script executes again.
Phone Calls Lead to Bounces
Therefore, if someone goes through these actions:
- Goes to your site
- Examines your offer
- Gives you a call
- Schedules a visit (some phone conversion activity)
- Hangs up
- Leaves your website a happy customer
This person is considers a bounce. That’s a very nice bounce, but its still a bounce.
A bonce in Google is a one page visit. That’s all it means.
To even further clarify Google Analytics metrics, if someone spends an hour on a single page and leaves, they are not only a bounce, but time on page isn’t counted. Google has no idea how long someone is on your page until they cause the script to execute again, usually by going to another page. If someone goes to another page, Google Analytics now knows time on the first page, but not time on the 2nd page. So, time on site is time on site minus last page visited.
3rd Party Integrations Lead to Bounces
In the past, we used Eventbrite for our AdWords Training signup process. Eventbrite hosts the pages themselves, so this is what our marketing funnel looked like:
- We sent out email / advertised / etc
- Users came to our landing page
- Users choose tickets on the landing page
- Users then went to eventbrite (off our site) to convert
If a user converted, they were usually a bounce. In fact, if a user spent a lot of time on our site, they were less likely to convert. Therefore, a high bounce rate usually meant more sales.
Some 3rd parties support cross domain tracking so you don’t see these high bounce rates and you can see more accurate metrics – but not all of them support analytics tracking.
Many driving direction pages go to Google maps. Many Paypal integrations go to Paypal.com. In these cases, you might have very successful bounces.
Adjusted Bounce Rates
Google Analytics (GA) recently launched a feature to track adjusted bounce rates. With adjusted bounce rates, you can set an event which will cause Google Analytics to track bounce rates based upon a time interval. For instance, if you wanted a bounce to be anyone who spend less than 30 seconds on the site, you can set a 30 second event to occur. At the 30 second mark, GA records an event and since GA did something, that user is no longer considered a bounce.
If you are using a plug-in for your CMS system, such as WordPress or Drupal, the plugins do not yet support this event. I tried to use this with Yoast’s analytics plugin and it not only didn’t work, it stopped the script from working properly so Google didn’t track any data. Hopefully, the plug-in authors will update their plug-ins soon with better control over additional information you can add to the Google Analtyics scripts.
Event Tracking & Virtual Page Views
Another way of handling users to leave your site (for a good reason, such as driving directions, affiliate link click-outs, etc) is to use events or virtual page views to track those clicks. With either of those tracking methods, Google Analytics will track that something happened on the site (and if you use events, you can even make them goals) and will adjust the bounce rate accordingly.
If you are using the default version of Google Analytics, then a bounce is just a one page visitor regardless of how much time was spent on the site. Bounce rate is just a metric. It’s neither a good nor bad metric – it’s just a metric. What’s important is the interpretation of that metric. So, next time you look at bounce rates, think about how your site functions and what you want users to do. If your goal is a call or sending a user off your site, a 100% bounce rate might be the best success metric you can reach.