Bounce rate is beautiful

By Laurence Veale | 8 Comments

If you don’t know Jared Spool, he’s the highly respected and influential principal at uie.com. A good proportion of our library is represented by publications from his various UI conferences so we hold him in pretty high regard here at iQ Towers.

However, I spotted something he tweeted last week.

"Analytics like bounce rate or time on page don't actually measure anything useful. They are amplifiers to whatever kind of spin you're looking for."

Yes, some metrics are truly awful, even dangerous. Time on page could be one good example. But other metrics, like bounce rate are awesome. So, in my book, he’s only half right.

In defence of the humble bounce rate

First, what is bounce rate?

Bounce rate measures the percentage of people who come to your website and then leave without having visited a second page (here’s a better definition – watch the video that follows it too).

So, with that in mind, here’s why I think bounce rate is one of the most useful metrics in analytics.

1. It’s simple to understand.
2. It can be applied to many of your efforts.
3. It’s actionable.

Most of the time, a high bounce rate is a good indicator that you really suck at something (though it’s not always the case).

So, what can bounce rate do for us?

1. Give you insight to the quality of your marketing efforts

  • When you segment your traffic sources, bounce rate can tell you which sources of traffic are sending you the right traffic and which ones are sending you accidental tourists.
  • It helps us answer the question “Should we spend less here or focus more of our attention there”?

2. Measure the warmness of your welcome

Assuming that the offsite traffic acquisition piece is working, perhaps your landing page sucks at welcoming all that great traffic. Why?

Well, maybe the page purpose could be so unclear so that what drove customers to the page and what they expect to be able to do on it are not aligned closely enough.

This can happen for a number of reasons:

  • the content hierarchy may be poor
  • your content may not be concise or compelling enough or just badly written
  • the call to action, if there is one, may not obvious enough.

3. Helps identify the “fix me now” issues

Ultimately, bounce rate is a metric that can help quickly identify “fix me quick” issues – those things you need to work on right now whether it’s those landing pages which need some love or poorly performing traffic acquisition sources.

When bounce rate is less useful

That said, bounce rate is not without limitations – it’s a quality assurance metric, not a conversion metric – you still need to set up conversion metrics specific to your site to measure “success”.

And sometimes, a high bounce rate is perfectly fine. A page with your contact details or a site like jamieoliver.com, where visitors got exactly what they were looking for in a one page visit, then left. Context is what matters here.

UX & marketing need to get along

When well known UX professionals tweet absolutisms, it serves to shut down conversation and pitches UX against analytics/marketing when the different disciplines should be looking for common ground and a shared understanding and appreciation for what each can bring to the table. The user experience begins well before landing on your website, and as UX peeps we need to acknowledge that.

Your turn

What I’d love to see here or over on uie.com, is some deeper discourse as to exactly why bounce rate is useful or useless and for that you’ll probably need more than 140 characters.

So, let’s get the ball rolling. What do you think? Bounce rate – useful or useless?

 

 

8 responses to “Bounce rate is beautiful

  1. I think, as I’ve said elsewhere, that all of these data points are useless in aggregate and useful in context.

    So for example, I just read this post and thought, “Hmm, I wonder who sends us our bounciest traffic”, and the answer was one of our paid adverts.

    However, telling me I have 11% bounce means jack shit. There is nothing meaningful to infer from it.
    The same arguments from my post here: http://blog.intercom.io/understanding-distribution-conversion/ apply to bounce rate.

  2. Thanks Des. Agreed – but bounce is one of those things that almost begs for deeper analysis. 11%, that’s great! or is it, let me dig into what that might mean for these users or these traffic sources etc.

    One thing I left out is that it’s the analyst, not the metric which is most important.

  3. Good to see you guys blogging more :)

    I think Bounce Rate is a great metric if in context and you interpret it right. It’s a lot more straight forward that wishy washy engagement metrics. I’ve sat with people who look at things like Time on site etc when making decisions on how their site is performing. Metrics like that cause a lot of people to jump to the wrong conclusions. At least with bounce rate it means this page sucks (in general) or people are getting exactly what they need from this page and don’t need to view anything else.

    I always harp onto site owners about segmentation though. Looking at bounce rate for all traffic types can make a page look bad, when in actual fact it suddenly ranks for an image or some other form of traffic is driving in crappy traffic. As you say, it’s an indicator of failure, you need to have your own indicators of success.

    I think bounce rate for home page is the most difficult. If you are a big brand, everyone wants to get a slice of that real estate, so you can often end up with over complicating your home page. Segmenting by traffic type and tracking internal clicks of the home page gives you a lot more data points to go on.

    Anyhow, waffling aside. I think bounce rate + segmentation and a little dash of content is still an awesome metric. Agree on things like Time in Site where you are basically putting your own interpretation on what that means.

  4. Wow, so many typos in the above. Should really read over things. A couple of these:

    “when in actual fact it suddenly ranks for an image or some other form of medium is driving in crappy traffic.”

    “Anyhow, waffling aside. I think bounce rate + segmentation and a little dash of context”

  5. Hi Lar

    As Des said, segmented bounce rate is the way to go, particularly by traffic source (Email, Paid Ads, Organic Search, Affiliates etc.). If you are any way advanced in UX on your site, high bounce (particularly if it is only evident in some traffic sources) is more likely related to traffic quality that landing page experience.

    I recommend analysing Visits Net of Bounce (i.e. subtract the number of bounces from your total visits from each traffic source). A traffic source you thought was growing year on year may actually have declined in Visits Net of Bounce.

    Then I re-run all my Conversion Rate numbers etc. based on the Visits Net of Bounce.

    It can be a really interesting way of looking at the data. It highlights scenarios where “junk” traffic is growing but the conversion rate of engaged traffic is actually improving. Your overall conversion rate would be down (because of the bounces) but you might just find that your Conversion Rate Net of Bounce has actually improved (justifying your UX spend!)

    Best wishes
    Niamh

Comments are closed.