Doing UX Research? Check These 6 Analytics Measures First.
Sometimes people ask whether web analytics can substitute for User Research, and my answer is always a flat “no”. Web analytics can tell us “what” people are doing on your site, but it’s only through UX Research (direct observation) that we can really understand why people are doing what they are doing. And if we don’t know why people are behaving the way they are, we have no roadmap to improving the interface.
All that said, web analytics are still super-important when planning user research. By telling us what people are doing (and not doing) on a site, web analytics tell us which “whys” we need to be investigating. To put it plainly, “Why Questions What” – and if we don’t have an idea of the “What”, we might miss the most important “Whys”.
Where people sometimes get stuck though is knowing which web analytics to look at. Analytics are.. abundant, to say the least, and looking at everything can lead to analysis-paralysis.
So here’s a brief overview of the 6 critical web analytics measures that you must look at when planning user research.
1. Search logs
Have a look at your site’s search logs. Most users are what is called “navigation dominant”, meaning that they prefer to use your site’s navigation links to find what they’re looking for. For most users, the search function is the approach of last resort – it’s what they use when they haven’t been successful with navigation. So looking at the search queries can show you what people are having difficulty finding. Then, in your user research, you can explore why that’s the case.
2. Page Flows
Looking at the sequence of pages users follow when using your site can provide a wealth of data for research exploration. Where do people enter the site? Do people go somewhere and then backtrack because they didn’’t find what they wanted? What are the most common links people click on from various “critical” pages in your site, and what does it tell you about the efficiency (or lack of efficiency) users are experiencing? If there are unexpected pages that get a lot of traffic, or pages you consider critical that don’t, these are all good issues to explore in research.
3. Top and Bottom Pages
Seeing which pages get the most (and least) traffic can help you identify critical issues for investigation. If the pages that are the most robust in terms of meeting users’ needs and supporting your business objectives are not well trafficked, you’ll want to find out why. And if you are hosting pages that are never seen, it may be worth knowing why. For example, is it because people misinterpret the nav labels that lead to those pages? Or something else?
4. Conversion Funnels
For e-commerce sites, understanding where people enter and exit your conversion funnels is must-have information for user research. But just about every site provides users with SOME action they can take to “take a step toward the brand”- whether it’s signing up for a newsletter or a freemium or going to the contact-us page. Analytics can help any website see where there’s room for “conversion” optimization – and research can then help identify the specific changes that will support increased conversion.
5. User Devices
Web analytics can tell you the percentage of your traffic that is mobile vs. desktop. This alone may shape the kind of user research you do, because it tells you a lot about the context of use. Maybe, if all the traffic that converts is mobile, you research needs to be on mobile too.
6. Time on site
Looking at how long it takes users to successfully do something on your site can highlight areas of inefficiency to explore. If it takes someone an unexpectedly-long time to complete a critical task, it’s worth knowing why.
Summary:
Qualitative UX Research will help you understand WHY users behave as they do on your site, but your investigation needs to be informed by the (quantitative) “what” from web analytics. Doing UX research without looking at analytics may lead you to miss the most important questions – or start in the wrong place. Combine web analytics with user research for best results!
What Next?
