A founder I worked with came back from ten customer interviews beaming. “Nine out of ten people said they would use my product,” she reported. “The market is there.”
I asked her to read me her first question.
“So I’m building a productivity app that helps freelancers manage their time better. What do you think — would you find something like that useful?”
Every word of that question was wrong. Not because it was impolite or unclear. Because it was designed — unintentionally — to produce a yes.
She told people her idea before asking for their opinion. She used the word “useful,” which implies the product has utility. She framed the question as a yes/no, where “no” requires the person to reject her to her face. She was essentially asking “Would you please validate me?” and then interpreting the polite responses as market data.
Nine out of ten people would say yes to almost any question framed that way. The data was worthless. The interviews were a waste of ten people’s time.
Here is how to do it right.
The Bias Catalog
Before learning what to do, you need to understand what goes wrong. There are six common biases in customer interviews, and most founders commit at least three of them simultaneously.
Leading bias. Your question implies the answer. “Don’t you think it would be helpful to have a better way to track expenses?” The only socially acceptable answer is yes.
Confirmation bias. You interpret ambiguous responses as positive. “They said ‘interesting’ — that means they’re interested!” No. “Interesting” is what people say when they have no strong opinion and want to be polite.
Social desirability bias. People answer with who they want to be, not who they are. “Would you pay for a fitness app?” “Yes!” (They have paid for three fitness apps and used none of them.)
Selection bias. You interview people who are easy to reach — friends, family, existing network — rather than people who represent your actual target market.
Order bias. The first thing you say shapes everything that follows. If you open with “I’m building something exciting,” the person is primed to be supportive. If you open with “I’m researching a problem,” the person is primed to be analytical.
Hindsight bias. People rewrite their past to match the present conversation. “Oh yeah, I’ve always struggled with expense tracking.” Maybe. Or maybe they are constructing a narrative that aligns with the question you asked.
The Bias-Free Interview Structure
Here is the interview structure I teach every founder at Startup Burgenland. It takes twenty minutes and produces reliable data.
Opening (2 minutes): “Thanks for taking the time. I’m researching [topic], not selling anything. There are no right or wrong answers. I’m trying to understand your experience, and honest answers — including ‘that’s not really a problem for me’ — are the most helpful thing you can give me.”
This opening does four things: it removes sales pressure, it normalizes negative responses, it frames the conversation as research, and it gives explicit permission to disagree.
Block 1: Current behavior (8 minutes)
“Walk me through how you currently handle [the activity related to your product]. What does a typical week look like?”
Then follow-up questions based on what they say. Not based on what you want to hear — based on what they actually describe.
“You mentioned you use a spreadsheet. How long have you been using it?” “What happens when you fall behind on updating it?” “Have you ever tried anything else?”
These questions follow the Mom Test principles: they ask about life, not your idea. They ask about specifics in the past, not generalities about the future.
Block 2: Pain exploration (5 minutes)
“What’s the most frustrating part of how you currently handle this?”
Then dig. Not one layer deep. Three layers.
“Why is that frustrating?” (Layer 1) “What happens when that occurs?” (Layer 2) “How does that affect your work/income/time/stress?” (Layer 3)
Three layers of “why” gets you from surface complaints to root causes. The surface complaint is “updating my spreadsheet is tedious.” The root cause is “I lose an hour every week that I could be spending on billable work, which costs me EUR 75.”
The root cause is where your product’s value proposition lives.
Block 3: Hypothetical solution (3 minutes)
Only now — after you understand their current behavior and their pain — do you introduce any concept related to your product. And even now, keep it abstract.
“If something could solve [the specific pain they described], how much would that be worth to you?”
Note: you are not describing your product. You are asking about the value of solving the problem they just told you about. Their answer tells you the maximum price your product could command.
“Honestly? If I could get that hour back every week, I’d pay EUR 20-30 a month for that easily.”
That is pricing data. Real pricing data. Not “what would you pay for an app?” but “what is the solution to your specific pain worth in your specific context?”
Close (2 minutes): “Is there anything about this topic that I didn’t ask but should have?” This catches blind spots. Sometimes the most important insight comes from a question you did not think to ask.
The Physical Setup
Small details change the quality of responses.
Location: Neutral territory is best. A coffee shop, a video call, their office. Your office makes them a guest, which increases politeness. Their home is too intimate.
Recording: Do not record. I know this is controversial. Recording produces more accurate transcripts but less honest answers. People become more guarded when they know they are being recorded. A notebook and pen produce better data because the person feels like they are having a conversation, not giving testimony.
Body language: Open posture. Genuine curiosity. Do not cross your arms. Do not check your phone. Maintain eye contact but do not stare. Nod to show you are listening but do not nod enthusiastically — enthusiasm is a form of leading.
Note-taking: Write direct quotes when possible. Their exact words are more valuable than your summary. “I dread the last week of every quarter” is data. “They dislike quarterly tasks” is your interpretation.
Five Questions You Should Never Ask
“Would you buy this?” Hypothetical purchase intent is fiction. Replace with: “How much time/money do you currently spend dealing with this problem?”
“What features would you want?” Customers are bad at designing products. Replace with: “What does your current workaround not do that you wish it did?”
“Do you like this idea?” Liking is irrelevant. Paying is relevant. Replace with: “How are you currently solving this?”
“How much would you pay?” Stated willingness to pay is unreliable. Replace with: “What have you paid for similar solutions in the past?”
“Is this a big problem for you?” Yes/no questions produce yes answers. Replace with: “On a scale of 1-10, how frustrated are you with this, and what makes it that number?”
Processing Your Data
After five to ten interviews, spread your notes on a table. Look for patterns.
Consistent pain: Did three or more people describe the same frustration? That is your primary feature.
Consistent workaround: Are most people solving this the same way? That workaround is your competitor — even if it is “I just ignore it.”
Consistent language: What words do people use to describe the problem? Those words become your marketing copy. Not your words — theirs.
Surprises: Did anyone say something that contradicted your assumptions? These surprises are the most valuable data points because they reveal what you did not know you did not know.
If the patterns are clear after eight interviews, stop. You have enough. Move to testing demand with real money.
If the patterns are unclear after ten interviews, your customer segment might be too broad. Narrow your target and interview again within the narrower group.
The Discipline of Neutrality
The hardest part of customer interviews is not the questions. It is the discipline of remaining neutral while someone evaluates your idea.
You will want to explain when they seem confused. Do not. You will want to persuade when they seem skeptical. Do not. You will want to agree when they say something positive. Do not.
Your job is to observe, not to influence. The moment you start selling, the interview is over and the data is contaminated.
This discipline is uncomfortable. It requires sitting with uncertainty and potential rejection. But the data it produces — honest, unbiased, grounded in real behavior — is the foundation of every product decision you will make.
Learn to interview without biasing. Everything else — your product design, your pricing, your marketing, your positioning — becomes clearer when the data underneath it is clean.
