This is an intermediate level post for people who want to know more about Bounce rate within Google Analytics and how to use it in analyzing websites. I have a site with a bounce rate of 72%. This is a a statement that I make with no shame at all. If you look at Google’s interpretation of the metric, it suggests that a high bounce rate is likely caused by something wrong you’ve done. Maybe you’ve incorrectly setup your Google Analytics, or have a bad design, or maybe your site isn’t a good user experience. I’m here to tell all of you “high bounce rate” site owners: the sky is not falling, and there may not be anything “wrong” with your site.
High bounce rate pages that are doing their jobs
Unsubscribe pages: A user’s intent on a page like this is to stop your non-awesome emails from bothering them ever again. Great unsubscribe pages defuse user frustration and maybe recover them into a newsletter or email subscription that makes more sense for them. Groupon has a hilarious unsubscribe page that probably has a huge bounce rate, but that doesn’t make it a bad page. By the way, make your emails awesome so that people don’t unsubscribe so often.
[Some] Product Pages: Do your pages sell a commodity or product that has a quick and simple pitch? Have a short sales cycle of 1-5 days, or even same day? A high bounce rate page might be fan-diddly-tastic! Consider a user searching for “volvo hubcaps” and they land directly onto your product page for factory volvo hubcaps. There are images, a size selector, and a buy now button. Let’s propose that only 15% of users click the buy button, and of that 15%, only 10% total actually complete the purchase. To put this in perspective your “crappy” 85% bounce rate hubcap page, would be a Christmas miracle for most eCommerce sites because a 10% convert to purchase rate, is pretty damn good!
Location Pages: So. . .the user gets directly to the store location page, because the user was looking for the store location page. . .and then what? Well, their journey with the website is over. The whole point of searching for and going to a store location page for most users is to get the phone number, store hours, and address. Few of those fulfillments are well tracked on a webpage, so we’re stuck with a “sucky” high bounce rate for our page, that in fact, just fulfilled our customer’s every need. That doesn’t sound fair, does it?
What are you trying to measure?
The thing that most people equate bounce rate to is as a “crappiness meter” for individual pages on a site. One page has an 80% bounce rate while another is 40%? The 80% one must be crappier, right? In reality, any accurate measurements needs to be more complex and take more into context. Pages that fulfill everything a user came for in 1 click are doing their job, yet are plagued with high bounce rates. If your site is well optimized, you’ll have visitors searching in Google, and landing on the most relevant, informative page you have on the topic.
Where improving bounce rates means something
Blog posts: Most company strategies around blogs involve using them as thought leadership, education, awareness building, and ultimately increasing sales. The thing with blogs is that they usually come with high bounce rates due to users’ single serving needs when they arrive on a post. Within this strategy framework, great blog posts inform and drive some traffic over to the “rest” of the site where products and services are offered, and hopefully generate a lead or sale. An alternative measurement to see if any of your thought leadership is sinking in, is to use Google Tag Manager and a timer event (future article coming). If a user spent 5 minutes on a meaty thought leadership article, maybe that’s success and needs to be measured.
Complex Product/Service: With more complex product or services, clicks to “more information”, videos, datasheets, and other supporting material is important, so bounce rate will mean more to those websites. There is also the issue of lack of tracking on “long form” pages that are really long and full of information. Sometimes those pages make sense, but much of the time you get a high bounce rate because you offer lots of information on a single page that they don’t need to leave. Running another service like Click Tale will tell you more about user scroll rates and how engaging your long copy page is.
Lack of CTA (Call To Action): Sometimes a page has lots of great information, but no logical next step. Urging the user to download a datasheet, attend a webinar, or request a quote is helpful because it doesn’t force the users to back up and figure out what to do next on their own.
Identifying “dead end” pages: Sometimes, especially with larger sites, or sites that have gone through multiple revisions, you carry over legacy content that ends up being a website dead end. Sorting pages by bounce rate can sometimes identify these pages.
Making Bounce Rate suck less
Bounce rate can be made more meaningful when:
- Clicks to external websites are tracked as events. This prevents the visit from being tracked as a bounce and that’s correct, because if you sent them to another site, they’re still on their journey as guided by you. When this occurs it is not a failure. You suggested another site, they went to it, it’s all good.
- Clicks on interactive page elements tracked as events. So, the user went to your super duper home builder page, custom built 4 different house packages, watched 3 videos, viewed content with a phone number on it and left. . .all without leaving that 1 interactive page. Is that bounce a failure? Of course not, that’s why custom event tracking of on-page elements is necessary to get meaning out of usage, on highly interactive pages.
- Time on page is tracked with a timer. A feature in Google Tag Manager but also present in other tracking packages, is the ability to accurately record time spent on page beyond what’s built into Google Analytics. If 5 minutes of client research on-page prevents 10 minutes of explanation by a customer service rep on the phone, 5 minutes of on-page use is a success!
A story about Bounce Rates
There once was a website with hundreds of pages of great content, but no real next step goal other than having the users read the content.
- Average bounce rate before: 70.2%.
- A 2 minute on-page timer was added which would throw a Google Analytics event once a user spent 2 minutes on site.
- An event listener for external link clicks was added.
- Average bounce rate after: 35.5%
Moral of the story: although the 70% bounce rate suggested a high rate of failure, it was found afterwards that 60.1% of users were in fact spending more than 2 minutes on site or clicking links to external websites. More robust tracking unlocked better use of the bounce rate metric and told a more complete story of how people used the website.