“Would You Use This?” - “Yes, I think so.”

Why Most Customer Interviews Fail — And How to Fix Them with Jobs to be Done

Photo by UX Indonesia on Unsplash

In the early stages of product development, founders and innovation teams are often told to “go talk to customers.” The advice is well-meaning. The execution? Frequently broken.

What follows are interviews laced with polite affirmations, abstract hopes, and an illusion of validation. Founders leave excited. They’ve “confirmed” their idea. But what they’ve really collected are false positives — nice words from nice people who don’t want to hurt their feelings.

“Would you use this?”

“Would you pay for this?”
“Does this sound useful?”

These questions feel like they de-risk the venture. They don’t. They invite people to imagine hypothetical futures — not describe their real lives.

And the answers?

A polite nod. A half-smile. A lukewarm “Yeah, I think so.”
It’s not discovery. It’s theatre.

False Certainty Is More Dangerous Than No Certainty

Innovation dies not because people don’t ask questions — but because they ask the wrong ones.

The core problem is methodological: predictive questions assume that intent equals action. But as any behavioral scientist or marketer knows, what people say they will do has almost no correlation to what they actually do.

People aren’t lying. They’re just guessing. And in guessing, they aim to be agreeable, aspirational, or simply helpful. That’s human nature — and it’s a trap for innovators.

“A customer’s politeness is the most expensive lie you can afford in innovation.”
— Yetvart Artinyan

From Solution-First to Problem-Led

The fix isn’t just better questions. It’s a different posture entirely.

Most discovery begins with a solution looking for applause. Instead, you should start with a struggle looking for resolution.

This is where Jobs to be Done (JTBD) provides a clearer lens. It flips the script:

People don’t buy products.
They hire them to make progress in a specific situation.

This shift is subtle but profound. It stops us from asking, “Do you like this?” and starts us asking, “What are you trying to get done — and what’s in your way?”

Let’s look at the contrast in real life:

• A parent doesn’t just want a “baby monitor with AI.” They want to sleep without guilt — knowing their newborn is safe, especially after a scare last week.

• A sales rep isn’t shopping for “deal analytics.” They want to walk into Monday’s pipeline review with confidence that they aren’t blindsided or behind.

These aren’t product features. They’re emotional missions wrapped in functional jobs.

You’ll never get these answers with:

“Would you use this feature?”

But you will get there by asking:

“What made you anxious last week?”
“When was the last time something slipped through the cracks?”

Don’t Validate Ideas. Surface Struggles.

Let’s reframe validation altogether. You’re not looking for yeses. You’re looking for struggles — because struggles create demand.

The most useful evidence lies in what people already tried to do, not what they claim they might try someday.

To operationalize this shift, we must redesign our discovery questions.

The following table contrasts ten commonly used — but misleading — validation questions with stronger alternatives rooted in behavioural evidence and JTBD principles: