3 Powerful Segmentation Tricks to Hone in on Website Weak Spots
Jack Maden | June 2, 2016
Jack Maden | June 2, 2016
When confronted by unfathomable amounts of website data, it can seem difficult to know where to start. How on earth, you think to yourself, am I supposed to find anything meaningful amongst all this?
That’s where segmentation comes in.
I don’t mean the standard slicing and dicing of user data by device, browser and general demographics. Such segments are important, sure; but as website data gets larger and more complicated, it’s important to build on them with more granular filters.
I’m talking about hardcore segmentation.
Unpacking insights with hardcore segmentation helps inform hypotheses for A/B testing, and also sets up routes for personalizing your website for your visitors. It starts by asking the right questions of your website.
For example, some top-level questions to inform segmentation for an ecommerce store might include: which product pages have the lowest conversion rate? Which forms have the highest abandonment rate? Is the basket to checkout transition loading fast enough across all devices and browsers?
It’s then that you need to drill down and get more specific.
Working alongside clients using the Decibel Insight platform, I’ve encountered three innovative ways to get more out of your segmentation.
1. Hone in with form analytics
Say there’s a form that’s underperforming on your website, like the one below.
Only 6.28% of users engage with the form when they see it. Of those, a third fail to complete it and 12.5% experience errors. What’s going on here?
You can instantly create a segment of users who interacted with the form, and already you’re more focused on a website weak spot. But we can get even more granular, as the image below demonstrates.
Say we want to just focus on users who abandoned this form, or who experienced errors on this form: no worries. Simply load up their session statistics and watch back recordings of their journeys to see what’s going wrong.
In this case, you’ve identified a weakness with your website – an underperforming form – and honed in on it by creating and applying segments of users who interacted with, abandoned, or experienced errors on that form. This has significantly reduced the number of sessions you need to analyze or watch back to come up with a hypothesis or discover an actionable fix.
2. Segment stages of your conversion funnels
Conversion funnels, in depicting users with a specific journey, already count as particular segments of users. But let’s take it a step further. Say you’ve created a conversion funnel for visitors going from the homepage, to the features page, to the sign up page, like this.
The funnel above seems relatively healthy; but what if we were to filter it for only mobile users, as in the image below?
We can now see that it’s not performing so well. To find out why, we can click on a particular stage of the funnel to load up the session of users who left at that point in the journey.
The ways you can slice up the data are endless: you just need to ask the right questions of your website to identify and hone in on its weak spots.
3. Combine segmentation with visualization
As well as focusing on specific users by layering filters through tools like form analytics and conversion funnels, you can bolster the power of segmentation by combining it with visualization. This combination can be remarkably compelling.
For example, the above image is the control of an A/B test we ran on the Decibel Insight homepage, overlaid with a heatmap displaying the number of clicks the page received and where they took place. We pulled in the segment via our integration with an A/B testing tool.
Now, this is the homepage with some refreshed copy.
We can instantly see, via the combined power of segmentation and visualization, that users who saw the new version of our homepage were more likely to hit our call to action button to sign up for an account.
It’s not just segments from A/B tests that benefit from this kind of visualization, however. Loading up a scroll heatmap on a particular web page and filtering it for different devices, for example, can be exceptionally useful for seeing the aggregated behavior of users who access the site in different ways.
Moreover, comparing the attention heatmaps of, say, new users versus returning users can be invaluable for informing the visual hierarchy of your on-page elements.
As user data gets more complicated, getting granular with your segmentation is the way to go when it comes to unpacking actionable insights from your analytics.
A recent webinar with WiderFunnel Founder & CEO Chris Goward went into great detail about how utilizing segmentation can boost conversion rate. If you’re interested in learning more, I recommend you start there!