How to conduct better user interviews with the hypothesis-driven design process


Share on LinkedIn

User interviews are the easiest way to learn what your customers really want so you can build products they’ll gladly give you money for. Yet, despite their clear benefits, too many product leaders and companies ignore the value of talking to users.

Part of the problem is arrogance. When you reach a level of success, it’s easy to start believing that you know better than the customer. Just think of the famous Henry Ford quote:

“If I had asked people what they wanted, they would have said faster horses.”

However, no matter how innovative or forward-thinking you are, there’s really no downside to talking to your customers. Instead, many companies skip out on user interviews due to some combination of fear, uncertainty, and the lack of a clear process.

While overcoming the fear of talking to customers is something you’ll have to get over on your own, there’s a process you can borrow from the design world that will make all your user interviews more insightful, valuable, and actionable.

What is hypothesis-driven design?

Designers are often faced with high-level goals with no clear solution path. But building a product with lots of unanswered questions means taking on unnecessary risks.

The hypothesis-driven design process follows five steps to take you from a sea of unknowns to a safe harbor of well-tested ideas and valuable insights.

Hypothesis-driven design process - InVision
Image source: InVision Inside Design

First, write down all the unanswered questions and assumptions you have. Designers often use the How Might We technique to phrase these questions. For example, if you’re building the next Uber but are concerned about safety, you might ask ‘How might we ensure riders feel safe when taking a trip?’

Then, prioritize those questions and assumptions with your team based on what you know.

After, turn those prioritized questions into hypotheses. In other words, answer your ‘how might we’ question with multiple solutions. For example, you could make riders feel safe by automatically sharing their location with a loved one or sharing driver feedback.

A hypothesis can be presented in the framework of

We believe that [this solution]
For [your persona]
Whill achieve [the desired result]

Now, use that framework to develop an experiment to test your hypothesis (UX designers have a whole toolbelt full of user research methods they use).

Finally, with the results of your test, you either continue building with confidence or return to a different assumption and solution set to test.

How to apply it to your user interview process

The hypothesis-driven design process helps you test assumptions and get clarity on your product before spending time and money on building new features. However, it’s still time-consuming and requires a ton of resources to be successful.

But what if you took that same process and adapted it to a low-barrier tool like a user interview instead?

Here’s what that looks like in practice:

Step 1. Start with questions and assumptions

Every great user interview starts with a goal. What do you want to learn? What assumptions do you want to test? Where are you feeling uncertain and want to reduce risk?

Here’s an example.

Let’s say you’re building a tool for UX designers and you want to learn how and why people transition from graphic design to UX.

You might start with some assumptions like: Graphic designers want to make more money. Or, overly subjective clients make graphic designers want to switch careers. Or even, graphic designers don’t think they have transferrable skills.

You can even rephrase those questions and assumptions using the same ‘how might we’ technique. For example, ‘how might we educate graphic designers on the benefits of switching to UX design?’

Step 2. Design your discussion guide

Now, you want to prioritize those questions and assumptions and work them into a discussion guide.

A discussion guide lists your objectives, specific questions, and which topics you want to discuss in a way that flows naturally. Instead of a designer who might choose one hypothesis to test, you can ask all of them throughout your interview.

Organize your discussion guide with a clear hierarchy from topic to assumption to question.

Specific questions

For example:

Topic: Dashboard usage
Assumption: Users want more ways to customize their dashboard
Specific questions:

  • How do you use your dashboard on a daily basis?
  • What specific features do you use the most?
  • How would you customize the dashboard to make those features easier to use?

Not only does this give a logical flow to your questions, but it also builds in a format for post-interview analysis as each answer is connected to its hypothesis or assumption.

Step 3. Master the art of ‘semi-structured’ interviews

Now that you have a list of prioritized assumptions and questions, it can be tempting to call up a customer and go through them one-by-one. Don’t do this.

Instead, you need to allow for the natural flow of conversation while still keeping it focused on your goal.

In Interviewing Users: How to Uncover Compelling Insights, author Steve Portigal explains how interviewing is a skill that is fundamentally different from what you do in a normal conversation:

“Great interviewers leverage their natural style of interacting with people but make deliberate, specific choices about what to say, when to say it, how to say it, and when to say nothing.”

A ‘semi-structured’ interview uses your questions as a guide but also includes more general questions to gain high-level insights and space for follow-ups.

Step 4. Recognize the 7 stages of a user interview

Once you’re talking to a customer, you’re essentially in the ‘test’ portion of your interview process. This can be unnerving for those who aren’t comfortable with user interviews.

However, almost all interviews follow 7 clear stages that you can use as your compass.

  1. Crossing the threshold. This is the beginning of the interview. Set up a proper environment and project a positive attitude.
  2. Restating the objectives. Give a high-level overview and ask if they have any questions.
  3. Kick-off questions. Transition into the actual interview and ask a few basic questions. Try to use a transitional phrase like ‘So, to get started…’ and then ask an open-ended question.
  4. Accept the awkwardness. Your interview subject is probably feeling awkward as well. If a certain path of questioning is resulting in shorter answers, move on to something else and circle back later.
  5. The tipping point. This is when you’ve successfully built rapport and your subject starts to give longer and more detailed answers. Try to stay in this phase as long as possible.
  6. Reflection and projection. If you’ve done your job up to this point, the interview should switch gears from tactical responses to reflections on the potential benefits. This is a great place to look for the emotional side of the interview. Why would they want to use your product? What’s the larger benefit for them?
  7. The soft close. Keep recording even once you’ve signaled that the formal interview is over. Many of the best insights come after a subject feels they’re ‘off the clock’ and can speak candidly.

Use this flow to take stock of where you are in the interview and make sure you’re asking the right questions at the right time.

Step 5. Turn your insights into actionable next steps

At this point, you’ve gathered valuable insights that either confirm your hypotheses or gives you more questions. Use what you’ve learned to adjust the discussion guide and head into your next interview.

User interviews can be your crystal ball

A user interview is more than just talking to customers. When you use a process like hypothesis-driven design, it becomes an easy and low-cost way to engage with users, test assumptions, and understand their problems, habits, skills, and desires.

Jory MacKay
Jory MacKay is the cybersecurity editor of the Aura blog. He’s written for and been quoted in Fast Company, The New York Times, Quartz, Inc. Magazine and more.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here