UX analytics are analytics used to drive user experience design. But what exactly do we mean by “analytics”?
Analytics are made up of data that can be analyzed to draw logical conclusions, and for UX design specifically, that data contains information about the users of your app or website, such as their age, their location, their interests, or simply their behavior — that is, how they use your app or website.
With this information, you can draw logical conclusions about who your users are and what they’re looking for, and when you know what they’re looking for, you can deliver it.
Anything else is an educated guess at best, since data is objective — although even in this objective data there can lie many subjective truths and misconceptions. Analytics can be a mystery, and it requires a detective to unravel those mysteries. Over a series of articles, we’re going to teach you how to unravel them.
We’ll show you how to analyze data using various tools and methods, and find hidden clues in that data to help you make design decisions that boost conversions and improve UX.
Things We Can We Do with Analytics
Whole books have been written on the uses of analytics data in the field of UX, but they can be summarized under two headings.
1. Creating data-driven designs
With analytics, we can make more informed decisions about our approach to design. We call this data-driven design. Data-driven designs are those that have been created using the data collected from analytics, and with this data we can do the following:
- find out where users are leaving, and why
- optimize the customer journey to reduce exit rates
- rethink visual design to aid usability and accessibility
- find out where and why the user is “rage clicking”
- boost conversations and maximize sales
- rearrange and tailor content to fit the user user intent.
Long story short: we can find out what users actually want.
2. Driving other types of user research
Although the term “analytics” tends to make us think of charts, graphs and statistics, data can actually come in many different forms (for example, the answers returned from a customer survey, or the heatmap returned from a user test). It can even be in the form of feedback from an internal discussion with your design team. But where do you even start with, say, a customer survey?
What questions do you ask your users?
We can find that out with analytics. Let’s say, from your analytics, you can see that your red call-to-action button isn’t converting all that well with Chinese visitors. Since red doesn’t mean “error” in Chinese culture like it does in the West, you could draw the conclusion that the issue with the call-to-action is the color;. However, you wouldn’t be 100% certain about that conclusion.
So, you could then create a customer survey that asks visitors this very question. What I’m trying to say is that, while using analytics isn’t the only way drive your design decisions, analytics can be used as a foundation for any of those other methods.
In short, take an analytics first approach to design.
UX Analytics Tools: What’s on Offer
Analytics can drive UX design from start to finish, but there are different tools and methods available depending on the type of answers you’re seeking — be they suggestive or confirmative. These excellent tools help to extract and decipher data that can inform your design decisions every step of the way.
Now, when I say “analytics”, the first thing you might think of is the Google Analytics tool. Fair play to you — Google Analytics has been around for a veeeery long time, and it’s estimated that well over 50 million websites are using it to learn more about their traffic statistics, user demographics and user behaviors.
What about heatmaps, though? Are they still a thing?
Yes, totally, heatmaps are still a thing, and they’re usually built into smart apps like CrazyEgg, Hotjar and Fullstory as a way of combining analytics with user testing. Heatmaps are more sophisticated these days, where the observations collected by them are intelligently translated into numbers and statistics.
Customer surveys, feedback, they’re still useful?
Customer surveys, usability testing, lean UX workflows — all of the techniques that involve prototyping, iteration and feedback — also offer us useful insights and allow us to draw a broader circle around what constitutes “data”. But analytics lives at the very center of all that, and drives these methods. They’re next steps.
So let your mantra be this: Analytics first.
UX Analytics vs UX Theory
UX theories (more commonly known as best practices) are based on user studies backed by data. For example, mobile-first design is fueled by data that mobile web usage overtook desktop web usage in 2016. But UX theories are based on the data collected from generic user groups, meaning that they’re not necessarily your users, and thus may not apply in your case.
What does this mean?
It means that until you take a data-driven approach to design, you’re relying on generalist design principles that may not apply to your actual audience. Data that 60% of users over the age of 50 are now active online may be misleading if you’ve built a trendy app like Snapchat, which most likely will mostly appeal to a younger demographic. You can’t safely draw usage conclusions from data that’s not relevant to your audience.
So although UX theory is still useful to know, the user experiences that you design will be far better when driven by data that you’ve collected, from the users and customers that you’re designing for.
Great, What Now?
Over a series of articles, we’re going to dive deep into UX analytics. We’ll start with the basics, to alleviate a host of common misconceptions about data-driven design.
After that we’ll dig into Google Analytics, covering key metrics, user research, how A/B testing can be used to confirm any conclusions, heatmaps, and of course the tools used for A/B testing — through tutorials and a case study. We’ll also look at how to reduce abandoned checkout rates, boost email conversions, and identify UX flaws with specific user groups — all through using analytics data.
To kick things off, we cover 5 Myths About Data-driven Design.