The Customer Interview Questions Nobody Asks (But Should)
"Would you use this?" is the most dangerous question in startup land. It sounds reasonable. It feels productive. And the answer is almost always "yeah, totally!" — which tells you absolutely nothing.
Here's why: people are nice. They don't want to crush your dream over coffee. So they tell you what you want to hear, you walk away validated, and six months later you're staring at a product nobody signed up for. Rob Fitzpatrick, who literally wrote the book on this (The Mom Test), puts it bluntly: it's not your customer's job to tell you the truth. It's your job to ask questions they can't lie about.
Why most interview questions produce garbage
The problem isn't that founders skip interviews. Plenty of founders do them. The problem is they ask the wrong things and come back with three types of bad data:
Compliments. "That sounds awesome!" Cool. Compliments are the fool's gold of customer discovery. They feel great, they mean nothing, and they've cost startups millions in misallocated resources.
Hypothetical fluff. "Would you pay $20/month for this?" People are terrible at predicting their own behavior. They'll say yes to anything that sounds vaguely useful when there's no credit card in front of them. Hypotheticals test imagination, not intent.
Wishlists. "It would be great if it also did X, Y, and Z." Now you're taking feature requests from someone who hasn't even experienced the core problem. You'll end up building a Swiss Army knife that does everything poorly.
The common thread: all three happen when you talk about your idea instead of their life.
The questions that actually work
The Mom Test's core rule is deceptively simple: talk about their life, not your idea. Ask about specifics in the past, not opinions about the future. Here are the questions that follow those rules — and what each one actually reveals.
"Walk me through the last time this happened."
This is the single best question in customer discovery. It forces specifics. You'll hear what tools they used, who they talked to, what went wrong, and where they gave up. One walkthrough answers a dozen questions you haven't thought to ask yet.
What it reveals: Their actual workflow, existing solutions, and friction points — not what they think they do, but what they actually did.
"What have you tried so far to fix this?"
If they haven't tried anything, the problem isn't painful enough. If they've tried three things and all of them suck, you've found a gap worth filling. The effort someone has already invested in solving a problem is the single strongest signal that they'll pay for a better solution.
What it reveals: Problem severity and willingness to pay. No effort = no urgency = no market.
"What's the hardest part about [process they described]?"
Don't ask "what's your biggest problem?" — that's too abstract and invites rehearsed answers. Instead, anchor it to something specific they just walked you through. The hardest part of their workflow is where your opportunity lives.
What it reveals: Where the real pain concentrates. Often it's not where you assumed.
"Why haven't you solved this already?"
This question is uncomfortable to ask and uncomfortable to answer — which is exactly why it works. The answer usually falls into a few buckets: they didn't know a solution existed, they tried and failed, it's too expensive, or — critically — it's actually not that important to them. Each answer tells you something different about your market.
What it reveals: Whether this is a real blocker or just a mild annoyance they've learned to live with.
"Who else should I talk to about this?"
Most founders forget this one. It does two things: it expands your interview pipeline organically, and it tells you whether the person considers this problem serious enough to refer you to someone else who shares it. A warm intro is a stronger signal than any verbal commitment.
What it reveals: How connected this problem is to a community — and whether your interviewee takes it seriously enough to vouch for the conversation.
How to know you're getting real signal
After 15-20 interviews using these questions, patterns emerge. Here's how to separate signal from noise:
Behavior over opinions. Trust what people did, not what they say they'd do. "I spent three weekends trying to fix this" beats "I'd definitely use that" every time.
Consistency across strangers. If five unrelated people describe the same problem in similar language, you're onto something. If every interview surfaces a different pain point, your problem space is too broad.
Emotional weight. Pay attention to sighs, frustration, and the moments when someone leans forward. The problems people get animated about are the ones they'll pay to solve. As we explored in Building Software Has Never Been Cheaper, knowing what to build matters more than the ability to build it.
Set a kill threshold. Before your first interview, write down: "If fewer than 8 out of 15 people describe this as a top-3 problem, I move on." Decide the bar before you start, because you will move it afterward if you don't.
The bottom line
Your customers will never tell you your idea is bad — not because they're dishonest, but because you're asking the wrong questions. Stop asking about your solution. Start asking about their life. The five questions above aren't clever tricks — they're just what happens when you care more about the truth than about being right.
This is the exact muscle StartSah's Discovery phase is designed to build. Structured interview guides, persona mapping, and assumption tracking that keep you focused on what customers actually said — not what you hoped they meant. Because the gap between a good interview and a wasted one is usually just the questions you asked. Try it out →
