Skip to main content

Trends, tech news, and tool comparisons.

How to run customer interviews as a solo founder: a 15-minute method

· 7 min read
Dealpuffin
Proof points and live coupons for early adopters

To run a customer interview as a solo founder, spend 10 to 15 minutes asking a person about a specific time they hit the problem, and say almost nothing about your product. Ask what they did about it, what they disliked about that fix, and write down their exact words. You are collecting facts about their past.

This method follows Y Combinator partner Eric Migicovsky's talk How to Talk to Users (YC Startup School, July 2019), summarized in YC's Startup School recap. Where a step is our own suggestion, it says so. We read both on 15 November 2026.

The three rules​

Migicovsky's talk reduces interviewing to three rules.

  1. Talk about their life, not your idea. Founders love pitching, and he says a user interview is not the time to pitch. The goal is to extract information that helps you improve the product, your marketing or your positioning.
  2. Talk about specifics, not hypotheticals. Instead of asking whether they would use or pay for a feature, ask about things that already happened. In his words, that gives you stronger information for decisions.
  3. Listen. Founders spend their days pitching investors and candidates, so restrain the urge to talk and take notes.

He frames the stakes simply: YC's core teaching is that you must build your product and talk to users.

The five questions​

The talk lists five questions anyone can use in early interviews. Here they are, with a note on what each one is for.

  1. What is the hardest part about [doing this thing]? It tests whether the problem is a real pain point for them. Fill in the bracket with their job.
  2. When was the last time you ran into this problem? This pulls out a concrete incident and its context, which you can later compare your solution against.
  3. Why was that hard? The answer tells you how to explain the value in their words.
  4. What, if anything, have you done to solve this problem? That is your real competition, including spreadsheets and doing nothing.
  5. What don't you love about the solution you already tried? The gaps here are the start of your feature list.

Migicovsky also says users are not good at naming the next feature they want, so do not ask them to design your roadmap. Our suggestion is to treat their answer to question 5 as a problem statement and design the fix yourself.

How to run one, step by step​

  1. Practice on yourself. Migicovsky suggests testing your interview strategy on yourself by walking through a time you hit the problem.
  2. Start with warm contacts. He recommends friends, coworkers and warm introductions for the first interviews, and says you do not need to talk to thousands of people. Our caution: friends tend to be kind, so ask them about the problem and skip "is my idea good?".
  3. Keep it short. He says a 10 to 15 minute first interview can give you the best information, and that you should be mindful of the other person's time.
  4. Record or take notes. Alone, you cannot do both well. He suggests bringing a friend or cofounder, or asking permission to record. As a solo founder, ask permission and use a recorder, then review the audio within a day.
  5. Do five before you judge. He says you learn a lot from the first five or ten interviews and your process improves. Start with one, then three, then five.
  6. Steer away from hypotheticals. When the person drifts into "I would probably...", he says to bring them back to what actually happened.

A note template​

Use one row per interview. The columns are our suggestion.

DateWho and roleLast time it happenedWhat they didWhat they dislikedExact quoteFollow-up
11-15Freelance designerLast Tuesday, invoice paid 3 weeks lateChased by email, then a reminder in NotionFelt awkward and kept forgetting"I just avoid sending the second reminder"Ask about how many late invoices per month

A worked example​

Imagine you are building invoicing software for freelancers. This is a made-up scenario to show the method.

A bad interview: "I'm building a tool that sends automatic payment reminders. Would you use it?" Almost everyone says yes, and you learn nothing.

A good interview starts with question 1: "What's the hardest part of getting paid as a freelancer?" She mentions late payments. You ask question 2: "When did that last happen?" She describes a specific invoice, three weeks late. Questions 3 to 5 reveal that she chases by email, finds it awkward, and has tried a calendar reminder that she ignores. Now you have a specific problem, her workaround and its flaw. You never mentioned your product.

Use the data​

Migicovsky describes a way to find your best first customers from interviews. His criteria include people who hit the problem frequently, and people who have the ability to act, because an operator may feel the pain daily but lack the budget. You can add these as columns to your notes.

When you have users, he points to the Superhuman CEO's survey, which asks how users would feel if the product went away. Migicovsky relays the analysis that 40% or more answering "very disappointed" is a signal of product-market fit. For small user counts, he says even 30 or 40 users is enough to ask.

He also suggests testing demand with action instead of opinion: rather than asking whether users would want a feature, show an upgrade flow that asks for a card. See how to price a SaaS product early stage before you do that, and tell testers what is real.

Where interviews fit in your launch​

Interviews feed your positioning and your list of first prospects. They pair well with how to get your first 10 paying customers and with how to get your first 100 users for a SaaS. Ask each person at the end whether they know anyone else who has the problem. Our suggestion: ask for one introduction per interview.

Frequently asked questions​

How many customer interviews should I do?​

Start with one, then three, then five. Migicovsky says the first five or ten teach you the most and improve your process. Keep going until the answers repeat.

How long should a customer interview be?​

Aim for 10 to 15 minutes for a first interview. Migicovsky says that length can produce the best information, and a short slot is easier to get from a busy person.

Should I pitch my product during the interview?​

No. The talk says the goal is to extract information, not to sell them on using your product. If they ask, answer briefly and return to their experience.

Can I interview friends and family?​

You can start there, and the talk suggests warm contacts. They tend to be supportive, so ask about concrete past events rather than whether they like your idea.