March 2nd, 2017

Usability Testing: design with data, not guesswork.

graphic of screen showing usability testing icons

Do you know how your visitors use your site?

Do you make design decisions about your website in a bubble? We all have opinions and preferences. They’re based on what we think looks nice, which words tickle our fancy, private aversions or attractions we may not even be aware of. Maybe we catch ourself saying things like “That red button is too bright and too sales-y. Can we make it a different color? Maybe not so in-your-face?” And it sounds like a good idea, so the designer makes the change. But was it a good change? You don’t know. Because you didn’t test it.

This is what happens when you design in a bubble: you make decisions not on information, but on preference. And because preferences differ among people, what’s appealing or intuitive for you may not be for another user. If you don’t take time to challenge those preferences, your site is likely to drastically underperform.

The good news is that you can do usability testing to observe how your visitors interact with your site. It’s far easier than you think it is, and once you have that information you can improve your site to create a more harmonious user experience.

How do you do usability testing?

Usability testing isn’t hard. The most basic usability test of all is simply to sit someone down in front of your website and watch how they use it. Grab a few people from your office who haven’t been involved in the project, sit them in front of your dummy site, and watch what they do. The closer the person is to your desired visitor, the better. If you can get your site in front of a dozen or so people to test it out and give you feedback, that will give you a small idea of how people use your site.

That said, this is by no means as rigorous a test as you could do with more people and more data. A dozen people is still a dozen opinions—better than one or two, but not enough to provide any strong conclusions. At build/create, we use Inspectlet to track and monitor user behavior on our websites. This has the benefit of allowing for hundreds (if not thousands) of user sessions, allowing for a wealth of analytics, but it also allows us to monitor behavior remotely and without users aware of being observed.

Inspectlet lets you watch user sessions, allowing you to follow their cursor movements, see what they click on and how they fill out forms, how long they stay on your page, and what page they’re on when they leave. No guesswork, just data.

What to look for as you test for usability

Once you start using a session tracker to monitor user behavior, there are several things you should keep an eye on as you seek to improve user experience on your website. For instance:

  • Can they find your site navigation and use it to find the right page?
    • Can your users find your navigation when it’s hidden behind a hamburger menu?
    • Does having your navigation on the side or along the top perform better?
    • When they navigate to a page, do they stay there or immediately leave?
  • How far down do they scroll on any given page?
    • Would adding more images help?
    • Should you add more copy, or edit what you have?
    • At what point do you lose their interest?
  • Do they click your CTA buttons, or do they slide right by them?
    • Should your CTAs be larger and bolder?
    • Is the copy ambiguous?
    • How does the position of the CTA on the page affect click-through rate?
  • What are they clicking on?
    • Are they clicking on things they think are links but aren’t?
    • OR, are they missing links entirely because they can’t see them?
  • When do they decide to leave your website?
    • Is there a specific page that seems to be losing visitors?
    • What path do they follow on your website before they leave?

Between heat maps and recorded sessions, you should be able to answer these questions with a high degree of certainty and implement them on your site. But even then, your work isn’t done: there’s still more to learn.

Usability testing isn’t a once-and-done deal.

As people grow used to different design functions on the Internet, their expectations and behaviors change. Many micro actions which originated on smart phones—such as the small pull-down gesture to refresh a feed, tap-to-like, or the right/left swipe have caught on enough to be expected in certain situations. And as more people grow in computer literacy and proficiency, they update how they interact with your site.

Websites aren’t static. They should be a reflection not only of what your users are looking for, but what you want them to see. Any time you make changes to your site, you should pay attention to how it affects user behavior. Furthermore, your ability to watch how users interact with your site should motivate you to spend more time adjusting and fine-tuning your site.

User behavior provides an essential reference point as you design your site. But it shouldn’t dictate everything you do. Your job, as you interpret user data, is decide which things you want to be easier and more efficient for the user, and which you want to discourage.

For instance, you may notice your users quickly leave your home page and navigate to some other part of your site. You now have a choice: you can either improve that section of your site and hope to convert more leads from there, or you can take some of the information from that part of the site and promote it more strongly on your home page. Or maybe that specific feature or product isn’t what you most want to promote about your business, so you decide to phase it out of your site or try to upsell visitors on a similar but more profitable product. Either way, usability testing puts you in control.

Related Articles