A founder I mentored at Startup Burgenland spent five months building a meal-planning app for busy professionals. She had the features mapped. She had a designer. She had written a detailed specification. She was three weeks from launch.
Then she talked to her first potential customer.
The conversation lasted eighteen minutes. In those eighteen minutes, she learned that busy professionals did not want meal-planning apps. They had tried three of them already. What they actually wanted was someone to handle the grocery shopping. Not the planning. The shopping.
Five months of development. Eighteen minutes of conversation. The conversation was worth more.
This is the problem interview, and it is the single most valuable skill a founder can develop. Twenty minutes of structured conversation that can redirect your entire business before you waste half a year building something nobody asked for.
Why Most Founders Skip This Step
The reason is uncomfortable but simple: talking to potential customers is scary. Building things is safe.
When you are building, you are in control. You decide the features. You set the timeline. Nobody can reject you because nobody has seen what you are making yet. The product exists in a state of infinite potential, where it might be brilliant.
The moment you talk to a customer, that comfortable ambiguity collapses. They might tell you your idea is wrong. They might be indifferent, which is worse. They might reveal that the problem you planned to solve does not actually exist.
So founders invent reasons to delay. “I need a prototype first.” “I want to have something to show them.” “I should finish my research before I start interviewing.”
All of these are reasonable-sounding excuses for the same underlying fear: what if they say no?
Here is the thing. If they are going to say no, you need to know that now. Not in five months. Now. The problem interview is not about getting validation. It is about getting truth. And truth, even painful truth, is always cheaper than building the wrong thing.
The Structure: Five Questions in Twenty Minutes
A problem interview is not a conversation. It is a structured inquiry. You are not there to chat, brainstorm, or pitch. You are there to extract specific information about how someone currently experiences the problem you plan to solve.
Here are the five questions I use. Each one has a specific purpose.
Question 1: “Tell me about the last time you experienced [problem].”
This is the most important question. It forces your interviewee to recall a specific, real event rather than speculate about hypothetical behavior. “Tell me about the last time you struggled with managing your finances” produces real data. “Would you use a financial management tool?” produces fantasy.
Listen for specifics. Dates. Amounts. Emotions. Tools they used. If they cannot recall a specific instance, the problem might not be significant enough to build a business around.
Question 2: “What did you do about it?”
This reveals their current workaround. Every real problem has a workaround — even if the workaround is “nothing” or “I just dealt with it.” The workaround tells you three things: how much they care about solving the problem, what solutions they have already tried, and what bar you need to clear.
If their workaround is a spreadsheet they update religiously every Sunday, you need to build something noticeably better than that spreadsheet. If their workaround is “I ignore it and hope for the best,” you need to build something easy enough to replace doing nothing.
Question 3: “What’s the hardest part about that?”
This is where you find the real pain. Not the surface problem, but the specific friction point within the problem. The meal-planning founder discovered that the hard part was not planning meals — it was carrying groceries up four flights of stairs after a twelve-hour shift.
The hardest part often becomes your product’s core value proposition. If you can eliminate the specific pain point, everything else is secondary.
Question 4: “How often does this come up?”
Frequency determines business viability. A problem that occurs daily can support a subscription product. A problem that occurs yearly might support a one-time purchase. A problem that occurred once is not a business opportunity — it is an anecdote.
You want problems that are frequent, painful, and currently unsolved. Two out of three is not enough.
Question 5: “If you could wave a magic wand and change one thing about how you handle this, what would it be?”
This is a creative question that reveals their ideal outcome. Do not take their answer literally — customers are better at describing problems than designing solutions. But their wish tells you what “better” looks like from their perspective.
If ten people give you the same magic-wand answer, you have a product specification written by your market.
The Rules of Engagement
The questions matter. But how you ask them matters more. Here are the rules I teach every founder I work with.
Do not pitch. The moment you start describing your product, the interview is over. They switch from “person sharing their experience” to “person being polite to someone who wants to sell them something.” This is exactly the bias you need to avoid.
Do not lead. “Don’t you think it would be great if there was an app that…” is not a question. It is a prompt for agreement. Ask open questions. Let them fill the space with their reality, not your assumptions.
Take notes, not recordings. Recording changes behavior. People become more guarded, more performative, more polite. A notebook and pen keep things conversational. Write down direct quotes — their exact words are more valuable than your paraphrased summary.
Be comfortable with silence. When you ask a question, wait. Count to five in your head before saying anything else. The best answers come after the pause, when people stop giving you the easy answer and start giving you the real one.
Talk to strangers. Friends and family will lie to you. Not maliciously — they just want to be supportive. Interview people who have no reason to protect your feelings.
How Many Interviews Is Enough?
The magic number is not a number. It is a pattern.
Interview people until you start hearing the same things. In my experience, this happens somewhere between eight and fifteen interviews. By interview eight, you will notice themes. By interview twelve, new interviews will mostly confirm what you already know. By fifteen, you are hearing the same stories with different names.
When two consecutive interviews teach you nothing new, stop. You have enough signal.
But — and this is important — if every interview tells you something different, you either have a diffuse problem (many people experience it differently) or you are talking to the wrong people. Go back to your customer profile and narrow it down.
Finding Interview Subjects
This is where most people stall. “I don’t know anyone who fits my customer profile.” Here is how to find them.
Online communities. Reddit, Facebook groups, LinkedIn groups, forums. Find the communities where your potential customers already gather, map your watering holes, and post a simple request: “I’m researching [topic] and would love to hear about your experience. Anyone willing to chat for 15 minutes?”
Warm introductions. Ask everyone you know: “Do you know someone who deals with [problem]?” Two degrees of separation usually gets you to the right people.
Cold outreach. LinkedIn messages work better than you think, especially if your message is short, specific, and makes it clear you are researching — not selling. “I’m building something for [role] and trying to understand [problem] better. Would you have 15 minutes this week?” converts at about 15-20% in my experience.
Coffee shops and co-working spaces. If your target customer is a small business owner or freelancer, they are probably sitting in a coffee shop right now. Walk up. Introduce yourself. Ask if they have five minutes.
This feels awkward. Do it anyway. The awkwardness fades after the first two conversations. The insights do not.
What to Do With What You Learn
After ten to fifteen interviews, you will have a stack of notes. Here is how to process them.
List every problem mentioned, ranked by frequency. The problem that came up in eight of twelve interviews is your primary opportunity. The one that came up twice is interesting but secondary.
List every workaround mentioned. These are your competitors — not the companies you found on Google, but the actual behaviors your product needs to replace.
List every “hardest part” answer. This is where your product’s value lives. Build for the hardest part first. Everything else can wait.
Then write one sentence: “My customer is [who], and their biggest frustration with [problem] is [specific pain point], which they currently deal with by [workaround].”
That sentence, built from real interviews rather than assumptions, is worth more than any business plan you could write. It is the foundation of a product that people actually need.
The Twenty-Minute Investment
Twenty minutes per interview. Ten to fifteen interviews. That is three to five hours of total work.
Three to five hours to learn whether your idea solves a real problem. Three to five hours to discover the specific pain points you should build for. Three to five hours to hear, in your customers’ own words, what they wish existed.
Or you can skip it. Spend five months building. Launch to silence. Wonder what went wrong.
The problem interview is not glamorous. It does not feel like progress the way building does. But it is the most efficient use of time in the entire startup process. Twenty minutes that save six months.
The meal-planning founder pivoted to a grocery delivery service for shift workers. She found her first ten paying customers from her interview list. People she had already talked to, who already trusted her, who had already told her exactly what they needed.
That is what happens when you ask before you build.
