Working as a conversion optimization specialist there are some things that really stand out. One of them is the horror my clients all have towards their bounce rates and the burning desire to lower it. It seemed to weigh more than many other metrics, the bounce rate simply has to go down.
Of course it’s absolutely understandable that you would want as many users as possible to stick around on your site, especially if you’re paying for traffic.
What many of them failed to understand before that was what that bounce rate actually meant and how it was calculated — what are the elements and interactions that affect it.
A bounce, as defined by Google, is a single-page session that triggers a single request to the Google server. The bounce rate of any page is calculated as
Bounce Rate = Sessions with single request (bounced sessions) / Overall Sessions X 100
See how the homepage in our example below has a 47.34% bounce rate. This means that out of the 101,629 sessions that landed on the homepage, 48,111 have bounced or had a single-page session with zero interactions with the page.
Now, a high bounce rate can many times indicate that your landing page is not relevant to your visitors. However, a high bounce rate is not always such a bad thing. This depends on the goal of your site — if your goal is to get users through a funnel where they need to see multiple pages and so on, then it is a bad thing! If your goal is for users to read your content, blog article, press release, etc. — put short, when the goal is a single-page session than a high bounce rate is more than expected.
Let’s have a quick walk through the ways you can analyze your bounce rate and see how to identify when you should be alarmed.
Your website’s average bounce rate
First, you will see your website’s bounce rate in the Audience → Overview report in your Analytics. It looks something like this:
Like most averages, this can be slightly deceptive. What you want to do is to have a look at bounce rates for your major traffic sources, landing pages, keywords, devices, networks even.
Your bounce rate by your source/medium.
This report can be easily found in Acquisition → All traffic → Source/Medium. What this report will show you, first of all, is the quality of the traffic that’s coming from your major traffic sources.
For this website you can clearly see how the paid traffic has very high bounce rate and the organic traffic has a low bounce rate. This can mean that the paid traffic campaigns are not very well targeted or that the landing pages do not match the expectations created by the ads.
Another very interesting aspect that we can see in this report is that the organic users are spending an average of 3:40 minutes per session and have an add to cart rate of 3.57%, while the users coming from PPC campaigns are spending less than half a minute per session and they’re converting at under 1%.
One crucial thing when looking into your analytics reports is to drill down into your traffic and content, avoid looking only at website averages.
Your bounce rate by each landing page
Find this report in Behavior → Site Content → Landing Pages.
Pro tip: Sort your landing pages by your bounce rate and then select sort type: weighted. What this does is that it shows you the pages with the highest bounce rates but with high volumes of traffic as well.
Your bounce rate by devices.
You can find this report under Audience → Mobile → Overview. Have a quick look through it and make sure nothing stands out that much. If you see a much higher bounce rate on one device, look into the experience on that device and loading times. Many times you can have a high bounce on mobile devices if the loading time is too high — users are not really that patient when it comes to loading websites.
In this example, we can see that the bounce rate on mobile is almost 25% higher than the bounce rate on desktop so that can be a first sign that something is wrong with the mobile experience. Of course, then you can see that 30% of the traffic that’s on mobile is bringing in only 0.66% of the total revenue so there are several signs there.
Now that you know how to look at your bounce rate and analyze it correctly, I should also tell you how you can adjust your bounce rate, especially if you have a blog or content website.
Adjusting your bounce rate
What is the purpose of adjusting your bounce rate? Simply put, it’s to understand if your users are actually engaging with your content. It might be that your users land on your page, read the information there, write down a phone number or address and then close the page. You would normally consider that as a bounce when you actually had an engaged user.
The simple way to do this is to send events into your Google Analytics that tell you when a users spent a certain amount of time on page, scrolled a certain % on the page or saw a specific element on the page.
You can do this in two ways:
- by adjusting your GA code (learn how to do this here: https://analytics.googleblog.com/2012/07/tracking-adjusted-bounce-rate-in-google.html)
- by sending an event from Google Tag Manager (see my step-by-step instructions below)
Adjusting your bounce rate through scroll % events
Recently, Google Tag Manager have implemented a “Scroll depth” trigger so you can easily create custom events based on the scroll depth for each user.
This is how you set up the Scroll Depth event in GTM. First, create a new tag.
In the tag setup screen, name your tag (GA — Scroll Tracking 75) and then select “Universal Analytics” for tag type and “Event” in the track type dropdown.
You will then get the three inputs for your custom event. Type in the Event Category — I kept it simple and named it Adjusted Bounce Rate.
Then, for event action, click the little plus sign for the variable and select Page Path. For the event label, look for the data layer variable named “Scroll Depth Threshold”.
If you don’t see this data layer variable, then you need to go to your “variables” screen and enable the scrolling built-in variables:
Select “Non-interaction Event” as “False” and add in your UA tracking ID — we are doing this as we want the scroll to 75% of the page to be considered an interaction.
Now, you need to create a trigger for the 75% scroll (you can create events for each of the thresholds you are interested in but I recommend setting anything under 50% scroll as non-interactive as to not skew your data, especially for shorter pages).
Name your event accordingly, select “Scroll Depth” as the trigger type and then in percentages, put down 75. Select if you want to trigger it on all pages or only on certain pages. Save, preview and debug and publish.
For a lot more info on Scroll Depth Tracking, I definitely recommend reading Simo Ahava’s blog article here.
Adjusting your bounce rate through the timer function
Another way to adjust your bounce rate is through counting the time spent on your page. You might think you are already doing that as you can see the average time spent on page but that’s actually counting between interactions so bounces are not in that category.
Here is how you create a custom event triggered by a timer:
Create a new tag, name it “UA — Adjusted Bounce Rate — Timer” for example. Just keep to your naming conventions and make sure you are consistent in the way you name things in GTM as it will get very complicated to find something once you start adding dozens of tags.
Now, create your 30 seconds trigger. Add a new trigger, name it “Timer — 30 seconds”.
Fill in the interval with 30000 (these are milliseconds, adjust the time as you see fit for your site here), then put a limit of 1 (you don’t want to send more than one of these events). Then, in the conditions, select “Page URL” matches RegEx “.*” as this would include all pages.
It is absolutely critical that you make sure you debug all of these tags and triggers in the GTM debug pane. Once they are live, go ahead and verify the events are being sent correctly to your Google Analytics account, through the real-time events report you can find at the top of your reports list.
I hope this brings some more clarity into the way you analyze your site performance and maybe you got one or two ideas extra on how to go further into your insight generation.
You can also contact us for a free no obligation page evaluation to find out how we can help you reach your business goals. Just drop an email at email@example.com or leave a comment in here and we will make sure to get back to you in less than a day.
Now it’s your turn: what other ways have you used to fine-tune your analytics implementation and analysis?