As marketers, we like to think people want to engage with us. We want them to spend all day on our site, publicly express their love for our products, voice their affinity for us on social media, and generally become evangelists for our business.
However, it should come as no surprise that people are busy, and they may not actually want to engage with you and love your brand all day long. (If you’re a marketer and this hurts, I’m sorry. But it’s true.)
I’m not saying engagement and brand love aren’t important. They are. But sometimes people just need answers, and most people typically find answers to questions by:
1. Searching Google
2. Finding (and clicking) a link to a page on your website
3. Reading what they need to read
4. Closing their browser (without viewing any subsequent pages or clicking any more links)
And going on with their day.
The aforementioned scenario is what Google Analytics defines as a bounce. A bounce, by its official definition
, is a single-page visit. That is, a user left the site from the same page they landed on, without clicking on anything.
Consider a busy professional. Let’s call him Joe. Joe wants to know what kind of furnace filter is best for his kids, who have allergies. He also has two dogs and a cat, and doesn’t vacuum very much.
With a not-so-simple Google query — “What kind of furnace filter is best for kids with allergies or if I have pets and/or I hate vacuuming?” — Joe has discovered a link to your site that contains the information he needs. Joe clicks the link from the Google search results page, reads the content on your page, gets his question answered, and leaves the site. He’s most likely going to buy your furnace filter because your content was so helpful (nice job on your SEO efforts, by the way).
By default, Joe’s interaction would increase
the bounce rate in your Google Analytics account. A high bounce rate is considered bad – you don’t want it to increase. And this is why bounce rate can be a misleading metric.
After all, Joe may not have clicked anything, but he read the entire page. And he’s probably going to buy your product. To me, that’s an interaction – nay, a possible conversion! You shouldn’t be penalized by having the bounce rate increase.
One solution: Tweak Google Analytics to measure attention
Before I suggest a solution to the bounce rate problem, I have to make a disclaimer: There are lots of ways to make bounce rate work better, and there isn’t a one-size-fits-all solution. Also, a high bounce rate can be indicative of a usability problem on your website, so be sure to address UX and content issues before implementing a technical solution. (Avinash Kaushik, my analytics hero, has lots of ideas on how to interpret the bounce rate metric
Justin Cutroni, who works at Google, has created an extremely helpful script
that allows you to measure attention using your Google Analytics account. All it takes is a simple script that you place on your website. In a nutshell, you can measure attention by tracking the following activities:
1. When a page loads (already tracked by Google)
2. If / when a user starts scrolling
3. If / when a user gets to the bottom of the content
Better still, the script tracks how much time elapses between steps. So not only can we see if our content is being read, we can also see how much attention users are dedicating to each piece of content.
But what does all this have to do with bounce rate, you ask? Good question. Using Mr. Cutroni’s script, when a user scrolls on a page, this is counted as an interaction
. If an interaction occurs, then that user’s session will not be counted as a bounce – even if they only visit one page and don’t click any links.
Consider the following three scenarios:
1. User A lands on a website, does nothing, and closes their browser without clicking any links or visiting any subsequent pages
2. User B lands on a website, reads the entire page, and closes their browser without clicking any links or visiting any subsequent pages
3. User C lands on a website, visits three more pages, then closes their browser
In the default Google Analytics tracking, Users A and B would be counted as bounces, and therefore the bounce rate of the above three sessions would be 67%.
With the attention-measuring script, only scenario A would be counted as a bounce. User B’s scrolling would be counted as an interaction, which would mean their session would not count toward the bounce rate. Therefore, with the script in place, the bounce rate of the above three sessions would only be 33%.
The interaction of User B means they don’t get counted as a bounce – even if they don’t visit any more pages or click on anything else. And that’s the way it should be: A user who takes the time to read your content should not count toward the bounce rate.
If you’re using Google Analytics, I strongly recommend implementing the aforementioned script, or something like it. As attention becomes an increasingly prevalent way to measure how users engage with Web pages, the bounce rate metric should take interaction into account and adjust accordingly.