Tregg Frank
May 20, 2024
Guides
Leading questions card floating in space

You should be trying to get unbiased feedback when talking to customers. That's the entire point of discovery research. But how do you do that? Ask better questions.

Okay, that is definitely easier said than done. I've heard even great researchers and PMs biasing their research participants with bad questions. The good news is that there are a few simple tips to follow that will improve your results.

Bad research questions often stem from wanting validation rather than truth.

The most painful, and simultaneously important goal of research is to gain a realistic understanding of your customers and your own product. That means you need to be genuinely curious and not defensive. At all. If you're defensive it will come through in your questions. Being defensive will impact the quality of your responses.

So, what can you do?

  1. Tell your participant that you're not emotionally tied to what you're showing them. Ask them to be as honest as possible.
  2. Ask them to think out loud at all times.
  3. Ask them about past experiences.
  4. Ask questions that sound negative. "What's the worst thing about this?", "how would you remake this to better suit yourself?"
  5. Leave questions open ended.

But what if I want to know if they'd use the thing I'm making?

Build and launch an MVP. Measure the usage. Are people paying for it? Are people using it? What are people saying once it's real?

Human beings are bad at predicting future behavior. They may tell you they would use your to-do app every single day. In reality, forming a behavior like that is hard. You need to focus on how to make the product itself so good that it promotes returning. Don't rely on someone telling you they will be back.

Let's get practical

Here are some example questions that are leading, and good alternatives to consider.

Instead of "Would you use this?"
Try
"What does this do for you?"

Instead of "Is this better than our competitor?"
Try
"Have you used anything like this in the past? How are they different from one another?"

Instead of "Do you find our new feature useful?"
Try
"What would make this new feature more useful?"

Instead of "Is this UI intuitive?"
Try
"Can you navigate to XYZ for me."

Instead of "Is the app faster now?"
Try
"Would you describe our app as slow or fast?"

Instead of "Would you recommend our product to a friend?"
Try
"How would you describe our product to someone who has never used it?"

Instead of "Is this layout clear to you?"
Try
"How would you arrange this page if you had the ability?"

Instead of "How do you like our colors?"
Try
"Do the new colors remind you of anything you've seen before? What do they make you think of?"

Instead of "Do you find the checkout process easy?"
Try
"Go ahead and check out and talk me through the process."

Don't be dismayed

Here's the truth; at some point you will bias your users when doing discovery research. It'll happen. That's not the end of the world, and it's a great rationale for why you should talk to many customers and why you should do so often.

You can argue that there's no such thing as truly unbiased research unless you're running a massive quantitative study. And even then, the way data is cut and sliced can imply things other than reality.

Bias itself isn't something to scoff at at all times. It's just really important to be mindful of the ways in which we're directing the conversation. Because the last thing we want is to hear someone tell us something convenient that we wanted to hear when the reality is we're about to build something no one wants.

Don't let the fear of biasing people a little bit stop you from talking to customers. Discovery should be a regular part of your product teams' workflow.

Alright. Get out there and learn from your customers. You won't regret it.

More Content

AI will never replace user research