Tregg Frank
May 20, 2024
Guides
Discovery research crash course

You may be here because you've heard that it's good to talk to users. Maybe you want to know how to do it better. Or... you've been tasked conducting research for the first time and don't know where to start.

The good news is that while the practice of conducting user research is as complex as you want to make it, it's not as hard as you'd think to get started.

We're not aiming for pristine, academic research right now. Progress over perfection. Our goal here is to get you started on the path to being good enough to learn what you need to make decisions for your product.

Decide what you want to learn

Spontaneous customer conversations are great, but it's much more efficient to know what you want to learn up front.

Start with a list of things you want to learn in question format. That will allow you to tailor the questions you ask users to serve answer what you're really wondering about.

Since you'll generally want to keep the research sessions to 45 minutes or less, you can also use these questions as a filter. Does someone have a question they want answered that doesn't level up to what you're tryng to learn? Skip it for now.

Decide if your goal is primarily to better understand your customer segment or if your goal is to understand how your customers are using your product. Discovery and usability research are different. Both are helpful, but the questions you'll ask and activities you do will be different.

You can get discovery questions out at the start of usability research, but a focused discovery research session shouldn't have much in the way of usability questions or observation.

Not sure of the difference? If you're trying to figure out what usability issues you have, you're primarily doing usability research. If you're trying to validate the concept of a product, to learn people's expectations, to understand their behaviors outside of your product you're primarily doing discovery research.

Collect questions from your stakeholders ahead of time

One of the worst feelings in the world is scheduling 8 interviews, going through all of them, writing a report, and then having a stakeholder ask if you learned about something that was never asked. Avoid that headache and have everyone send you questions they're curious about before hand.

Have a plan, write a script

It's cool to go with the flow, but having a general script to follow will be very helpful in making sure you hit all the desired points.

Pro tip: Start with 5 minutes of small talk to grease the wheel a bit and get the participant talking. This is a great time to do simple stuff like ask them about themselves. For B2B products I often ask about their team and company.

Establish that your ego isn't tied up in what you're discussing. We want to get honest answers to our questions, so make sure the participant understands that. The entire goal of product research is to learn enough to make better product decisions, not to make ourselves feel better.

Be comfortable with silence

Shhhhh. Stop. Stop talking.

Do not fill the gaps unless you're encouraging the participant to think out loud.

Silence during an interview is powerful. Resist the urge to fill every gap in the conversation. Often, giving participants a moment to think will lead to more thoughtful, detailed responses, and yes, they will eventually respond. It’s surprising how much more information people will share if given an extra second or two to reflect. People hate silence. They'll fill it and you'll learn valuable insights.

Answer questions with questions

You almost never want to answer the participants questions unless you're entirely blocked from moving forward with the conversation. Most questions about features, positioning, the interface, and what things mean should be asked back to the participant.

"What does this icon mean?" should be met with "what do you think it means?". Once they answer, you can tell them what you had intended. After that, follow up by asking what might have been better for them. "What icon would you have used?"

Don't ask leading questions

A leading question is one that assumes a portion of the answer. For example, asking someone what they like about something. What if they didn't like anything about it? You're leading them to positive output. We don't want to validate our opinions, we want useful insights.

In my opinion "how much would you use this?" is as much a leading question as "what do you like about this?". If you're going to lead them anywhere, lead them somewhere negative. "What's the worst thing about this?"

Record the conversation

If you're meeting remotely, ask the participant if it's cool to record the call for internal purposes (I've personally never had someone say no) and then capture a full audio and video recording.

If you're meeting in person, ask the participant if it's okay to record audio on your phone or laptop. Video can be helpful, but in these situations framing the shot can be more hassle than it's worth. If you're coordinating research through an agency that sources participants and provides facilities, or if your office is really rad, you might have a really great set up for this already. Use it!

Conducting user research might seem daunting at first, but you shouldn't let that stop you. Learn enough to make informed decisions for your product. That's the goal. As you gain more experience, you'll refine your techniques and develop a deeper understanding of your users and your product. Good luck!

More Content

Research is mapmaking