The Bounce Rate Myth

Pogo stick

Don't judge content by bounce rate alone.

For advanced web analytics users, bounce rate is quite possibly the most popular stand-alone metric — analytics evangelist Avinash Kaushik suggests it’s the sexiest web metric ever. It’s certainly at the top of my list. Google defines bounce rate as "the percentage of single-page visits or visits in which the person left your site from the entrance (landing) page." It’s great because it tells you right away that something is wrong with your website. Or does it?

Analysts typically use bounce rate as a measure of poor quality content — or as an expression of dissatisfaction with your site. But bounce rate has a lot more to say than simply "your website stinks." In fact, it might even say something good!

Every web metric has more than one angle.

As with all web metrics, we need context to provide meaningful insights. Maybe one bounce means a visitor left because she immediately found what she was looking for or bookmarked the page to view it later. Every web metric has more than one angle.

Look at "time on page," often regarded as a positive stand-alone metric. Surely, the longer someone stays on your page the more engaged they are with the content, right? But maybe a long time on page means the visitor had a difficult time using or understanding the content. One metric never says it all.

What’s the Value of a Visit Without Action?

Contrary to popular belief, a bounce does not mean a visitor didn’t take action.

Answer: likely, none. Indeed, as Tim Nekritz says in our Meet Content introductory video, "Great content moves people to action." But contrary to popular belief, a bounce does not mean a visitor didn’t take action.

Here are common visitor actions that people don’t often account for with bounce rate:

  • Followed you on Twitter
  • Liked you on Facebook
  • Mentioned you online
  • Left a comment
  • Bookmarked your site
  • Emailed the web link to a friend
  • Printed the page
  • Watched embedded video or listened to embedded audio
  • Recommended you to friends or peers offline
  • Returned to your site later to take action or took action offline

Some of these actions can be properly tracked with web analytics tools to reduce your bounce rate, including establishing events and virtual pages for external links, downloadable files and user comments. Others, such as social media mentions or offline actions, require additional measurement solutions.

Make Your Bounce Rate Meaningful

Segment new visitors

Loyal, repeat visitors will not take an action every time — they already know where to look for news and information — but the fact that they are returning and reading your content regularly is positive.

Narrow your definition of bounce rate

Kaushik uses time on page to more narrowly define bounce rate as "the percentage of website visitors who stay on the site for a small amount of time (usually five seconds or less)." To understand the complexity of measuring bounces by time on site, check out What’s wrong with bounce rate? by higher ed analytics expert Shelby Thayer.

Track all online actions

Not all analytics tools track all online actions by default. Make sure that your bounce rate accounts for links to Twitter, Likes on Facebook, PDF downloads, video views, user comments and print-button clicks.

Track offline actions

Understand web content that encourages offline action so you can measure it.

One of the things analytics can’t tell you is what actions people take offline. But don’t let that deter you from learning. Conduct user research. If prospective students inquire by phone or email, ask them how they learned about your institution. Understand web content that encourages offline action so you can measure it.

You can’t automatically tie offline action feedback to web analytics, but you can identify it as a key performance indicator alongside your analytics data to identify trends, including successes and unmet goals.

Prepare Your Content for Action

Our content plan needs to recognize and account for the many ways people take action. Content can’t be measured by clicks alone. We need to define the right combination of web metrics to support our measurement goals.

However, while no stand-alone web metric is enough to make informed decisions, bounce rate is often a great place to start. It may not account for every action a visitor takes, but it accounts for a large chunk of them and is a great lead indicator for content problems.

For learning more about bounce rate and ways to effectively use the metric, check out Kaushik’s post Standard Metrics Revisited: #3: Bounce Rate.

How do you evaluate bounce rate at your institution?

Photo by woodleywonderworks / Flickr Creative Commons

  • RSS
  • LinkedIn
  • Twitter
  • Facebook
About Rick Allen

Rick Allen has worked in higher education for over twelve years, helping to shape web communications and content strategy. As principal of ePublish Media, Inc, a web publishing and content strategy consultancy in Boston, Mass., Rick works with knowledge-centric organizations to create and sustain effective web content. Keep going »

Comments

  1. Your post is a perfect example of why content analytics require time, effort, and THINKING to get right. Your point about bounce rates is exactly why I advocate a data-informed approach to content, not a data-driven approach. You can’t simply take a bounce rate and make decisions. You have to consider the whole situation. Very nice post.

    • Thanks for the feedback, Colleen! Indeed. You can’t make good decisions with dashboard metrics. As you say, it takes *thinking*. I wholeheartedly agree with the distinction between data-informed and data-driven content. Using analytics effectively means understanding what the numbers *don’t* tell you just as much as what they do tell you.

  2. Mikey Cooper says:

    I just assumed users were finding my page from their google search, getting the solution to their problem, and then leaving. Bounce Rate seemed kind of meaningless as a negative metric to me. For something where you want to funnel them through a conversion pipeline, I can see the use. For an informational blog, not so much.

    • Hi, Mikey. Thanks for chiming in. I hear ya. Although, I consider bounce rate a super valuable metric if used appropriately. Avinash Kaushik cites some great examples in the post I mention above, including evaluating bounce rate for traffic sources and keywords. However, even these examples require you to have clear measurement goals in order to make good use of the numbers. For blogs, as you mention, bounce rate is much less useful.

  3. Good stuff, Rick. I’d also add that some of the most important visitor actions on higher ed sites often result in “bounces”. At many institutions (or most?) forms such as visit registrations, inquiries and applications are hosted outside of the website and aren’t tracked, resulting in the appearance of bounces. The same is true of student systems (portals, self-service, course management) and the dozens of other separate websites that make up a higher ed website (schools, colleges, library, etc.). Until you’re tracking ALL of these, your bounce rates aren’t really telling you much of anything.

    • Thanks, Todd. So true. Easily half of the online actions on most higher ed websites are not accounted for in bounce rate, including PDF views, video views, off-site application and registration forms, library databases, and email links. As you mention, portals and internal systems are often overlooked too. In lieu of technology that auto-tags these links with tracking code, it requires lots of manual work. But, the effort is needed to make use of the data.

  4. Hi Rick,

    Thought I would ask you about a dilemma I am facing:

    There’s a video section on my client’s blog that posts relevant content related to his industry. In google analytics, I can see that this page is getting a 100% bounce rate.

    1100 %. WTF? The page is decent and professional looking. The blog is clean and has no ads in the sidebar – no banners – no pop ups or pop under – How do I further analyse this phenomenon in analytics?

    What questions should I be asking and how do I go about reducing that rate?

  5. Brian Holda says:

    Good stuff! Thanks for sharing. Another thing I’d add, which seems like a huge factor, is if someone makes your website their home page (at least for a season). If they do this, then there will be a high bounce rate, but it really shows that they value your content so much they want it to greet them every time they start their web browser.

    • Brian Holda says:

      I should also add that if your site serves as a link to other sites, then if they are doing their job and following the links, it will be a high bounce rate.

What do you think?

*