How to Run Customer Discovery Interviews (Without Wasting Your Time)

I once spent six weeks interviewing thirty-two people for a software idea I was certain would be a hit. After the last call, I sat staring at a spreadsheet full of notes that said almost nothing useful. People had been polite, agreeable, even enthusiastic. But when I finally launched, nobody bought. The problem wasn't my product—it was my interviewing. I had asked the wrong questions, in the wrong way, and walked away with a pile of well-meaning lies instead of real insight. That experience taught me that customer discovery interviews are either a goldmine or a time-suck, and the difference comes down to how you run them. Here's what I learned the hard way, so you don't have to.

Why Most Customer Discovery Interviews Fail (And How to Avoid the Same Trap)

Most founders I meet treat customer interviews like a formality—something to check off before building. They schedule a few chats, ask a handful of canned questions, and call it validation. But the real reason interviews fail is that they're designed to collect confirmations, not truths. We instinctively want people to like our idea, so we ask leading questions like "Would you use a product that does X?" The answer is almost always yes, because saying no feels rude.

Another common mistake is talking too much. I've sat in interviews where the founder spent the first ten minutes pitching their solution before asking a single question. By then, the interviewee is already primed to agree. Worse, many interviewers never dig past surface-level complaints. Someone says they "hate managing expenses," and the interviewer nods and moves on. But the real insight lives in the specifics: When was the last time it caused a problem? What did you do? How did you feel?

The third trap is confirmation bias. We hear what we want to hear. If someone says "I might pay for that," we interpret it as a buying signal. But in reality, "might" is a polite no. The smartest thing I ever did was to go into interviews expecting to be wrong. That shift in mindset—from proving my idea to disproving it—made every conversation more honest.

Pro tip: If you leave an interview feeling great about your idea, you probably didn't learn anything new. Real discovery should feel uncomfortable.

To avoid these traps, you need a system. That starts before you ever dial a number.

The 5-Part Prep Checklist That Turns Interviews into Gold

Walking into an interview without preparation is like sailing without a map—you'll move, but you won't get anywhere. Here's the checklist I use before every round of discovery, refined by trial and error.

1. Define Your Assumptions (Not Your Questions)

Before writing a single interview question, list your core assumptions about the customer, their problem, and your solution. For example: "Small business owners spend more than 5 hours a week on invoicing." Each interview then becomes a test of that assumption, not a pitch for your product. This keeps you focused on learning rather than selling.

2. Choose the Right Interviewees

Don't talk to friends or family unless they genuinely fit your target persona. I once interviewed my cousin because she was available, but she had zero experience with the problem I was solving. Her feedback was useless and misleading. Instead, use LinkedIn or industry forums to find people who live the problem daily. Offer a small incentive—a $5 coffee card or a charitable donation—but avoid making it transactional. The goal is a conversation, not a survey.

3. Write a Discussion Guide (Not a Script)

A rigid script kills the natural flow of a good interview. Instead, create a loose guide with 5-7 open-ended topics. Start with broad context questions like "Tell me about a typical day at work," then narrow down to specific problem areas. Leave room for tangents—they often reveal the most unexpected insights.

4. Prep Your Recording Setup

I learned this one the hard way: recording is non-negotiable. Even with the best notes, you'll miss tone, hesitation, and offhand comments that matter later. Use a tool like Otter.ai or just your phone's voice memo app. Always ask for permission upfront: "Do you mind if I record this so I can focus on our conversation? It'll only be used for my notes and I'll delete it after." I've never had anyone say no.

5. Set a Time Limit—and Stick to It

Thirty minutes is the sweet spot. Any longer and both you and the interviewee lose focus. I schedule 45-minute slots but aim to wrap at 30, leaving buffer for follow-up. Respecting their time builds goodwill and makes them more willing to help again.

With this prep done, you're ready for the conversation itself. But the questions you ask will make or break the whole exercise.

How to Ask Questions That Uncover Real Problems (Not Polite Lies)

People want to be helpful, which means they'll tell you what they think you want to hear. Your job is to make it safe for them to tell the truth. That starts with the questions you ask.

Avoid Leading Questions at All Costs

Never ask "Would you use a tool that does X?" or "Do you find Y frustrating?" Instead, ask about specific past behaviors: "Tell me about the last time you had to track expenses for a project. What happened?" Real stories reveal real patterns. Hypotheticals reveal wishful thinking.

Use the "Five Whys" Technique

When someone mentions a problem, keep digging. If they say "I hate reconciling receipts," ask "Why is that a pain?" Then ask again: "And why does that matter?" By the third or fourth why, you'll often uncover a deeper pain they hadn't articulated themselves. I once discovered that a founder's real frustration wasn't the accounting software—it was the fear of an audit. That changed my entire product focus.

Ask About Workarounds

One of the most powerful questions is "What do you do today to solve this problem?" If someone has built a messy spreadsheet or cobbled together duct-tape solutions, that's a strong signal the problem is real. If they shrug and say "It's fine, I just deal with it," the pain probably isn't acute enough to pay for a fix.

Embrace Silence

This took me years to learn. After someone answers, count to three in your head before responding. Most people will fill the silence with more detail—often the most revealing part. I've had interviewees say something like "Well, honestly, the real issue is..." after a pause. If you jump in too fast, you cut off that gold.

One more rule: Never pitch your solution during the interview. Even if they ask, say "I'm still learning—tell me more about your situation." Once you pitch, the conversation stops being honest.

When the call ends, the real work begins. Notes alone won't help you build a better product. You need a system to extract the signal from the noise.

After the Interview: The 3-Step Process to Extract Insights (Without the Noise)

Raw interview notes are chaos. I used to highlight what seemed important, but that just reinforced my own biases. Now I use a three-step process that forces me to stay objective.

Step 1: Transcribe and Tag Within 24 Hours

Your memory fades fast. Use your recording to transcribe the interview (automated tools work fine), then tag each quote with a theme: pain point, workflow, workaround, emotional reaction, budget. I use a simple spreadsheet with columns for the quote, the theme, and the interviewee's role. This takes about 30 minutes per interview but pays off tenfold when you look for patterns.

Step 2: Look for Patterns, Not Anecdotes

A single passionate story can be misleading. If three out of ten people mention a specific pain, that's worth investigating. If only one person mentions it, it might be a niche issue. I draw a simple table with each problem listed and a count of how many interviewees mentioned it. The ones with the highest counts become my priority.

Step 3: Update Your Assumptions—Not Your Roadmap

Resist the urge to start coding or building after the first few interviews. Instead, update your list of assumptions. What did you confirm? What did you disprove? What new questions emerged? I keep a living document called "What I Know vs. What I Think I Know" and update it after every three interviews. This prevents me from falling in love with an idea too early.

One practical tip: after each batch of interviews, write a one-page summary that answers three questions: (1) What is the core problem we heard most often? (2) How are people solving it today? (3) What would need to be true for them to switch? Share this with a teammate or mentor for a sanity check—it's easy to miss your own blind spots.

Done right, customer discovery interviews aren't a chore. They're the fastest way to learn what actually matters—and save yourself from building something nobody wants.

Frequently Asked Questions

How many customer discovery interviews do I need to run before I can validate an idea?

Aim for 10–15 focused interviews per customer segment. Quality matters more than quantity—listen for patterns, not isolated anecdotes.

Should I record customer discovery interviews?

Yes, but always ask for explicit permission first. Recording frees you to be present in the conversation and catch subtle details later.

What's the best way to find people to interview for customer discovery?

Start with your personal network, then use LinkedIn, industry forums, or even cold outreach with a low-pressure ask. Offer a small incentive (e.g., a $5 coffee card) if needed.

How do I avoid leading the interviewee with my questions?

Use open-ended questions like 'Tell me about the last time you experienced...' instead of 'Don't you think X is a problem?' Let them describe their experience without injecting your assumptions.

What if an interviewee gives me very vague or short answers?

Probe gently with 'Can you tell me more about that?' or 'What happened next?' Silence can also be powerful—wait a few seconds after they answer to encourage elaboration.

Your Takeaway

Customer discovery interviews are the single best way to de-risk your idea before you invest time and money. But only if you run them with a clear system: prepare with assumptions, ask about past behavior, and extract patterns after each session. The next time you schedule an interview, go in expecting to be wrong. You'll walk out with the kind of truth that actually helps you build something people need.

Worth bookmarking before your next round of interviews—this guide will save you hours of wasted calls.