21
November 2024

Not all user interview data is good data

But it’s not the user's fault. The truth is, as humans, we are utterly terrible at predicting and even describing our behavior accurately. So, how can you conduct user interviews with this in mind?
Sara Fazzini
Design Manager, Belka

During interviews, we tend to answer with what we believe is socially desirable or acceptable, rather than being fully honest about how we actually behave. This is called social desirability bias. And, as a result, the data you collect can be unreliable and not truly reflect the users’ habits and needs.

Here’s an example. 

When asked if we’ll buy more vegetables or reduce our plastic use, we almost always say yes. But will we really? Probably not. When it’s time to make the actual purchase, other factors take over: convenience, emotions, or ingrained habits. It’s not that we’re intentionally lying, it’s just that we project an image of ourselves that aligns with what’s ‘right’ or desirable.

If you ask me if I eat vegetables, I’ll say yes. But if you ask if I have them in my fridge or when I last bought them, well, that’s a different story. We say one thing but do another.

How can you manage this behavior in a user interview?

Don’t look for this in a user interview

As we’re not great at predicting the future or describing the present in a generic way, these data points won't be very useful in a user interview:

Predictions of future behavior

Be wary of statements like “If I had this feature, I’d buy the premium plan.” Generally, generic promises of future behavior are unreliable. There are smarter and more reliable ways to gather this insight without directly asking, “Would you buy it?” For example, “Have you ever bought something similar in the past?”

Descriptions of solutions

“You should add a button here” or “You could make a document merging feature” might be good suggestions, but they’re not the main focus. If the user suggests solutions, dig deeper to understand the problem behind their suggestion. What matters is the problem they’re trying to solve, because, ultimately, we can’t be certain that their solution will work.

Instead, look for this

To gather useful data (the really good stuff) from your user interview, you need stories about past experiences - both positive and negative.

When asked to recount specific moments of real experiences, users are compelled to share their behaviors and provide context for them in a more realistic way. Stories reveal the circumstances, reduce the risk of idealized responses, expose emotions and frustrations, and uncover real needs. They provide far more insight than simple statements or theoretical answers.

Our favorite questions

You need questions that uncover concrete, practical details about how users interact with your product. Some of our favorite questions to get the good data include:

  • "Can you tell me about the last time you…?"
  • "Can you describe an episode when…?"
  • "What did you find frustrating in that situation?"
  • "What alternatives did you consider in that situation?"

First, you need to set the environment for a story to come out. You can’t simply ask, “Tell me about your yesterdays’ dinner” – you’d get a reply like “I had my quinoa and zucchini rice” and that’s not a story. You need to guide users to rebuild the story with you: help them vocalize the story by asking about the setting and other people with them. 

At Belka, we’ve had success with questions like, “Imagine I was a fly in the room during dinner, what would I have seen?” or “Describe your dinner to me like I’m blind.”  

Getting to the good stuff

The unique value of a user interview lies in accessing their stories and problems. Focus on uncovering these, on understanding how they work and behave. Once you’ve got the good stuff, you can design everything else.

Want more good advice from Belka?