User testing is a tricky area of expertise. It requires a lot of empathy, thinking on the fly, preparation, and the ability to extract yourself from bias. You may want users to behave a certain way, with user testing you’ll get the truth.
“Truth” is also tricky. Because it’s subjective, different people have different approaches to using a product. Designers “know” how a product should work, giving them a certain bias towards how people should behave. But people may disagree, taking unexpected paths that make perfect sense to them. And if that path isn’t friction-free, they’re less inclined to use your product.
For example, users who visit an insurance site might want to review their benefits. Or maybe they want to find a provider. Or to look up symptoms of a particular ailment. Each represents a different user flow with multiple routes. They could search. They could click icons. They could fill a the contact form. They might ignore the login link. There’s no way to know … until you test with real people.
User testing uncovers assumptions, helping you tailor user flows to how people behave. The result is a more intuitive product with less friction. But getting there requires UX designers to learn some new skills.
Listening is a skill you can practice
Listen: Start by talking less and asking open-ended questions. Initial questions help define “what’s up.” Follow-up questions should focus on “why” and “how.” And always follow up with more questions. If a person shows enthusiasm for a topic, follow it to understand their passions.
Timing: Agree upon a time limit with your interviewee, and don’t go past. Your own attention span is limited, and interviews take concentration. 45 minutes is usually enough, though some can go as long as an hour and a half. If an interview starts to run long, then set up a follow-up meeting.
Team up: If you can, bring in a co-worker to take notes while you talk to the interviewee. That frees you to keep up a casual chat — and maintain eye contact — to keep the conversation real.
Say it back: Repeat and paraphrase what the interviewee says so they know you’re paying attention. This both builds trust and helps you confirm that you understand them. It also gives you an opportunity to follow up if you want more information on a vague or interesting answer.
Come prepared: Have questions ready, but be ready to abandon them if the conversation takes an interesting turn. Questions provide a framework for talks, not hard-and-fast goals. For example, relevant stories are often the most interesting points in conversations. They can reveal how a person feels about the topic on hand.
Stay on track, mostly: While people tend to stray from the questions you ask, especially during follow-ups, that’s not always bad. Dig deeper if people start to tell stories.
Encourage them to think aloud: You want that narrative of the interviewee’s thoughts. What someone says, and what they do, aren’t always consistent. Understanding what they want to accomplish and how they assume is the best way to do that will help you find your own assumptions.
Share control: Conversations are interactive. If you keep directing the conversation — say, by prompting them with questions — then you’re probably asking closed-ended questions. Always ask open-ended questions.
Asking smart, open-ended questions
Open-ended questions are those that lead to more conversation, which is the goal of any user interview.
Ask yourself or a coworker each question. Scrap anything that can be answered with a yes or no. “How would you log in to this service?” is better than “Do you see the login link?”
Ask specific questions. “Based on this site’s appearance, would you trust the organization with your money?” is better than “What do you think of this site?”
Ask “why” and “how” several times. For example, follow up “Why did you avoid that button?” with “Why doesn’t its offer sound appealing?” And ask for examples. “When would you want to do that?”
Phrase questions to elicit emotional responses. “How do you feel about …?” or “Tell me about ….” For example, “What happened at your job today?” is less powerful than “Tell me about the projects you’re working on.”
Don’t lead them on. “Isn’t this design powerful?” will earn a yes-or-no answer partially based on what they think of your own bias.
Where to from here?
You’ve interviewed someone about your product. You took a ton of notes. What comes next? Start by organizing your notes, clarifying your thoughts as you go.
Look for follow-up questions. What doesn’t make sense, or might lead to more insight about the project? This is where lo-fi prototypes shine: you can use a mockup of your product to refresh your interviewee’s memory.
Don’t keep insights to yourself. If you work on a team, share your findings with everyone who needs to know what users actually do with your product. Since the point of user testing is to discover assumptions and problems in user flow, fixing those problems is equally important for everyone involved.
User testing will uncover your assumptions about how a product should work. You can use that knowledge to tailor user flows to how people behave, not how you want them to behave. The result is a more intuitive product with less friction.
