Published on January 24, 2022
Maybe the two scariest words in the English language when it comes to UX. Because the great unknown in the creative realm of web design is what exactly these “real people” are going to do when encountering your website for the first time.
Real people often do all kinds of odd things and react in ways few can predict. The popularity of Nickelback, for example. The Tide pod challenge. You get the idea. That’s why usability testing is a no-brainer when you’re launching a website. You have to know how real people operate, react to, scroll through, and feel about being on your web page before you take it live.
So let’s walk through the basics of usability testing. Let’s wade into the world of real people.
Finish or Not?
The whole point of usability testing is to see how easy your website is to use, which often means figuring out whether “real people” can complete the main task you want done. This is probably the main reason you’ve devoted the time and money into designing the perfect website.
But it’s only perfect if real people do what you want them to. So what you’re looking for is whether users can finish a task on their own. Moreover, users need to finish a task in a timely manner without getting pissed off about it.
The corollary to this goal is that you need to begin usability testing early, in the prototype stage. Once you get validation that your concept has legs, you’ve established a strong foundation to build on. If users react poorly, then you know you need to switch gears. Designing the best product requires a certain amount of failure you can learn from and grow.
We Have Ways of Making You Talk
Usability testing can come in a few flavors. The most vanilla is an in-person setting, where a moderator walks users through the design–but without too much coaching. Often these sessions are filmed for further analysis. It’s imperative that the feelings and thoughts of the users find a voice, either in exit interviews or surveys. If something is confusing them, you can actually quantify the intervals and see if it’s the same element.
A remote setting, maybe on Zoom, also is advantageous because now you can create conditions as they exist in nature. This session might be best conducted without a moderator so that you get some unfiltered reactions.
If you’re feeling very brave, you can go guerilla and just ambush friends and colleagues to see how they do, though this method might strain relationships and it’s also not scientific and prone to bias, but sometimes you have to take risks.
It goes without saying that you’ll need to incentivize these testers so that they’ll go the extra mile for you. But that will be money well spent.
Look Before You Leap
But before you start hiring real people to test your website, you need to know precisely what you want them to test, and that means you need to plan ahead. You should come up with a checklist to work from and define the scope of what you want to test so you can compare oranges to oranges. Otherwise, you might end up with a random assortment of data points that have no context.
Usually tests last about an hour, and in this hour, you’ll want users to conduct no more than five different tasks, such as ordering a product and checking out. You don’t want to overload the test subjects with a long and winding to-do list. Know what you’re looking for ahead of time.
And let them come to your website in the most natural way possible. The moderator should not be auditioning to be the next Alex Trebek. Don’t provide too many instructions. You want flaws to become known, and if you guide too much, the flaws might get masked. Navigation is a huge issue and if users can’t move through your site, chances are they’ll bounce and you won’t get conversions.
How many real people does it take to expose the cracks and fissures of your website? Not as many as you might think. A coterie of five–yes, five–folks can reveal over 80% of the problems in web design from a UX perspective.
Not five random people. You might want testers who do their shopping online or who come from a certain profession or belong to a certain age group. You also might want to know their attitudes toward your product or service. There are many demographic considerations to account for, and having a wide range of experiences will be the most helpful. Digital natives will respond to your website differently than a Boomer.
Usability testing is NOT A/B testing, so don’t get those concepts confused. A/B testing is how you can determine which version of a website works best, but the key to usability is in understanding why users might be struggling with a CTA element. A/B testing doesn’t drill down to the ultimate causation of a problem.
While surveys can be an important component of usability testing, surveys are NOT synonymous with usability testing. Just giving users some kind of Google form is not the same as observing their behavior. By the same token, heat maps that capture where actual eyeballs are spending the most time on a website aren’t a replacement for usability testing. Again, heat maps can be very helpful but they can leave much unsaid about what users are feeling. Rage-scrolling is a thing, after all.
And you might be thinking: don’t web design companies have in-house testing in the form of robust QA? Yes! No doubt, and we’re very proud of how rigorous our testing department is, but in-house testers can’t stand in for actual users who encounter a website for the first time.
The Goal, The Goose
Providing the best possible UX is the goal. Seeing those fabled “real people” in action with your website is one very smart way to glean useful data that directly speaks to your website’s performance in an area that can be subjective. But if “real people” can’t figure out how to navigate or use your website, your goose is cooked. Usability testing makes a lot of sense for a lot of good reasons.