Before you write a single line of code or spend on incorporation, you need to know whether the problem you are solving is real and painful enough that people will pay to make it go away. The trouble is that most customer conversations lie to you. Friends, family, and even polite strangers want to encourage you, so they say your idea is great. Rob Fitzpatrick named this problem in his book The Mom Test: if you ask your mother whether your business idea is good, she will say yes, because she loves you. A good interview is designed so that even your mom cannot mislead you.
Why "would you buy this?" gives false positives
A false positive is when someone signals interest they do not actually feel. It is the single most dangerous outcome of a bad interview, because it convinces you to build something nobody wants. Fitzpatrick sorts the misleading answers you will hear into three kinds of bad data: compliments ("that's a cool idea"), fluff (generic claims, hypotheticals, and future promises like "I would definitely buy that"), and ideas (the feature requests and suggestions people toss at you). Pitching your own solution makes all three worse: the moment you describe what you are building, the other person switches into compliment mode and starts inventing polite feedback. Questions like "Do you think this is a good idea?" or "Would you pay 499 rupees a month for this?" are broken by design, because they ask people to predict their own future behaviour, and people are terrible at that.
The three rules of the Mom Test
Fitzpatrick reduces good interviewing to three rules:
- Talk about their life, not your idea. Your job in a problem interview is to understand their world, not to sell. Ideally, do not mention your idea at all.
- Ask about specifics in the past, not generics or hypotheticals about the future. "Do you usually struggle with GST filing?" invites a vague guess. "Walk me through the last time you filed your GST return. What went wrong?" pulls out real, checkable facts.
- Talk less, listen more. Aim to speak roughly 20 percent of the time and let them speak the other 80 percent. You already know your idea. You are here to learn theirs.
Good questions versus leading questions
A leading question smuggles your desired answer into the wording. Compare the two columns below and notice how the good version stays anchored in real, past behaviour.
- Bad: "Would this app be useful for your kirana store?" Good: "Tell me about the last time you ran out of stock. How did you find out, and what did you do?"
- Bad: "Would you pay for a tool that automates invoicing?" Good: "What are you using to send invoices today, and how much time or money does it cost you?"
- Bad: "Don't you find hiring developers frustrating?" Good: "When did you last hire a developer? Take me through how it actually went."
Also chase the money and the workaround. If a problem is real, people are already spending time, effort, or cash to work around it. Ask "What have you already tried to fix this?" and "Where does the budget for that come from?" No workaround usually means no real pain.
Look for commitment, not applause
Praise is worthless. What matters is whether the person will part with something valuable. Fitzpatrick describes three currencies of commitment, in rising order of seriousness: time (agreeing to another meeting or a trial), reputation (introducing you to their boss or peers, which risks their own credibility), and money (a deposit, pre-order, or signed letter of intent). A conversation that ends with a concrete next step is a signal. A conversation that ends with "keep me posted" is a polite rejection. Always know the specific ask you want to make before the meeting, and make it before the meeting ends.
A simple practice plan
Aim to speak with 10 to 15 people who genuinely have the problem, not friends being kind. Keep a short list of open questions, record with permission, and write your notes immediately afterwards using their exact words. After every few interviews, look for patterns: the same painful moment described again and again is your signal to build. Silence, vague enthusiasm, and unpaid problems are your signal to keep digging or to move on.