Why Your Google Analytics Engagement Report Could Be Inaccurate And Two Custom Metrics To Improve It
The way you measure engagement in Google Analytics is potentially flawed.
In this video I’m going to explain how two engagement metrics you’re probably using really work and how this affects your insights.
You and anyone reading your reports, should be aware of this if you are using these metrics as KPIs, as it could change how the report is interpreted.
I’ll also show you two ways to configure your analytics tracking to get more insightful engagement data in your reports.
The best part about it, is you can actually combine the data from the two new metrics, with the existing two default metrics to significantly improve the data in your engagement reports.
If you’re watching this video, you probably measure engagement as part of your overall reporting.
If, like most people, you use Google Analytics as your data source, you probably include these ‘out of the box’ metrics in your engagement report. I’m talking about:
- Bounce Rate
- Average Time On Page
Questions you might look to these metrics to answer might be along the lines of:
- Question: “Which pages are people bouncing from?” – Or in other words, which pages do most people leave very shortly after landing on the page
- Possible Outcome: Change the content or remove pages with high bounce rates
Time On Page:
- Question: “How much time do people spend on this webpage?”
- Possible Outcome: Change the content or remove pages with very low time on page
Given the majority of people don’t convert on a blog post, engagement KPIs such as time on page and bounce rate are usually given more importance when trying to measure the success of your content.
The importance of these metrics and the potential to make significant decisions based on this data, is exactly why what I’m about to explain is so important.
Here’s how the default bounce rate and average time on page in Google Analytics really work
In order to record the time a user spends on a page, Google Analytics needs to measure the time between two ‘hits’.
When a person lands on the first page, a hit is recorded.
When the second hit is recorded (which can be an event or page view), Google Analytics measures the difference between the two timestamps to calculate time on page.
- Lands on page 1 at 10am
- Clicks to page 2 at 10:05am
- Clicks to page 3 at 10:08am
- = time on page 1 recorded as 5 minutes
- = time on page 2 recorded as 3 minutes
- = time on page 3 recorded as 0 as they left without triggering a hit
The time on site will also be recorded as 8 minutes (which is the difference between the timestamps from the first hit to the last hit).
But what if your users were staying on page 3 for 20 minutes, reading all the content before leaving?
What if most people are landing on your blog post and spending 10 minutes reading it in full, obtaining everything they needed from your post, before leaving with their information needs satisfied?
With the default metrics, you’d see a very low time on site and a very high bounce rate which might make you think the content on this page was low quality, when in actual fact, your visitors might be highly engaged with this page.
If you only use the default metrics, your report could be misleading.
Here’s how to configure Google Analytics to track two additional engagement metrics for more accurate reporting
So what are these two mystical metrics of which I speak?
The first is scroll depth.
So how far down the page are users scrolling?
Do very few people scroll after landing on the page or do they scroll all the way to the bottom of the page?
Do most people scroll to a certain point then drop off?
With this data, if you identify a page with a high bounce rate and low time on page recorded by the default metrics, you can also check the scroll depth.
If users are also not scrolling, you know there is likely a problem with the content on the page or at least the headline or value proposition.
The second additional metric is…. Time on page!!
Yep, you heard that right.
“But how is that different to the time on page currently reported?”
It’s all in the configuration.
Remember how I said that a page view OR an event can trigger the hit required to record the timestamp?
Well for much more accurate time on page reporting, we set up event tracking to trigger an event to fire every 15 seconds to see if the user is still on the page. Every time the event fires, a hit (and therefore a timestamp) will be recorded.
If you factor in scroll depth, you can now see:
- How long they spent on the blog post
- How far down the post they scrolled
So for example, let’s say you have a very in depth, high quality blog post and Google Analytics is saying people are bouncing when they land and don’t spend any time on the post.
You can check your custom configured time on site and scroll depth metrics to see how long people are actually spending on the post and how far down the post they are scrolling.
Often, what you might find is that people are highly engaged with a post you initially thought was poor performing. Therefore changing the content of this post could actually make things worse and reduce the user experience.
Now… if they are not clicking through to other pages or taking actions after engaging with the post, you might have some work to do… but that’s for another video.
OK, for step by step instructions showing you how to set up the scroll depth tracking and event based time on page tracking that we’ve talked about in this video, please visit our blog post here.
So get it configured, give it a bit of time to collect data, then include it with the engagement metrics in your next report for greater accuracy and insights from which to base decisions.
Helping you navigate and leverage the world's most powerful digital marketing channels.