What is customer discovery (and how to run your first 10 interviews)

You’ve probably felt it already. You have an idea that feels right, a few friends say it sounds interesting, and you keep tweaking a product in your head late at night. But when you try to explain it to someone outside your bubble, the conversation goes fuzzy. They nod, ask polite questions, and nothing concrete comes out of it. That tension is exactly where customer discovery lives. It’s the uncomfortable but necessary work of replacing assumptions with evidence before you spend months building the wrong thing.

To put this guide together, we reviewed founder talks and interviews from Y Combinator, First Round Review, and early startup podcasts, then cross checked those lessons against documented outcomes from companies like Airbnb, Dropbox, Intercom, and Superhuman. We focused on what founders actually did in their earliest days, not the polished frameworks they describe years later, and translated those behaviors into a process you can realistically follow as a first-time founder.

In this article, we’ll define what customer discovery actually is, why it matters so much early on, and walk through a clear, repeatable way to run your first 10 customer interviews without turning them into awkward sales calls.

What customer discovery actually is

Customer discovery is the process of learning how people experience a problem today, before you decide what to build to solve it. It’s not pitching. It’s not validating your solution. And it’s definitely not asking strangers if they like your idea.

At its core, customer discovery is about understanding three things:

First, understand what problem someone recently experienced in real life, not hypothetically. Second, consider how they tried to solve it using their own time, money, or workarounds. Third, find out why that problem mattered enough to cause stress, delay, or cost.

Steve Blank popularized the term during his work with early startups and later through the Lean Startup movement, but the practice predates the label. Long before the phrase “customer discovery” was common, founders were already doing it out of necessity. They just called it talking to users, knocking on doors, or sitting next to customers while they worked.

The key distinction is that discovery is about learning, not convincing. If you leave a conversation feeling validated but no smarter about how someone behaves when the problem actually hits, you didn’t do discovery.

Why customer discovery matters so much early

At the pre-seed and seed stage, your biggest constraint isn’t ideas. It’s conviction. Every feature you build, every hire you make, and every pitch you give is based on what you believe to be true about your customer.

Founders who skip discovery usually don’t fail immediately. They fail slowly. They build features no one uses, struggle to explain value clearly, and burn months chasing traction that never quite arrives.

There’s a reason so many successful founders talk about early conversations as a turning point. Brian Chesky has described how Airbnb only started to grow after he and Joe Gebbia went to New York in 2009 to meet hosts in person. They didn’t ask if people liked Airbnb as an idea. They watched how hosts struggled to create listings and realized photos were the bottleneck. After personally photographing around 40 apartments, revenue in New York roughly doubled within a month. The insight didn’t come from a survey. It came from observing a real problem up close.

Customer discovery compresses time. Instead of guessing for six months, you can often learn in a few weeks what actually matters and what doesn’t. For an early stage founder, that difference is everything.

What customer discovery is not

Before getting tactical, it helps to be clear about the common traps.

Customer discovery is not asking, “Would you use this?” People are bad at predicting their future behavior, especially when there’s no cost to saying yes.

It’s not demoing a prototype and asking for feedback. That shifts the conversation into politeness and opinion instead of facts.

And it’s not talking only to friends, investors, or other founders. Those conversations can be encouraging, but they rarely reflect real buying behavior.

Discovery is about past behavior, not future intent. What someone did last time the problem showed up is far more reliable than what they say they might do someday.

Who you should interview first

For your first 10 interviews, the goal is not statistical confidence. It’s pattern recognition. That means you want people who are as similar to each other as possible in the way that matters.

Start by writing a very plain description of who you think your customer is. Avoid broad labels. “Small business owners” or “creators” are too vague to be useful. A tighter description might be something like, “US-based e-commerce operators doing 200 to 1,000 orders per month on Shopify who handle fulfillment themselves.”

This kind of specificity makes it easier to recruit and much easier to compare answers. When everyone has roughly the same context, differences in behavior actually mean something.

Rahul Vohra took a similar approach at Superhuman. In the company’s early days, he focused on a narrow segment of email power users and asked a simple question about disappointment if the product disappeared. By narrowing the audience before measuring intensity, Superhuman was able to sharpen both product direction and messaging. For a first time founder, the takeaway is simple: pick a narrow slice, not everyone who might someday use your product.

How to recruit your first 10 interviewees

Speed matters more than perfection here. You should aim to get your first 10 interviews scheduled within a few days, not weeks.

Most founders combine three channels.

The first is warm introductions. Ask friends, former colleagues, or advisors if they know one or two people who fit your specific description. Be clear about who you want to talk to and that you’re not selling anything.

The second is targeted outbound. Short, respectful emails or LinkedIn messages can work surprisingly well if they’re specific. Mention why you’re reaching out, acknowledge their role, and ask for 20 minutes to learn about how they handle a particular problem. You don’t need fancy copy. You need relevance.

The third is in product or community intercepts. If you already have a landing page, newsletter, or small user base, ask directly. A simple note offering a coffee gift card in exchange for honest feedback can be enough.

What matters most is recency and authority. You want people who have experienced the problem recently and who are involved in deciding how it gets solved.

How to structure the interview itself

A good customer discovery interview feels more like investigative journalism than a brainstorm. Plan for about 30 minutes and structure it intentionally.

At the start, set the frame. Explain that you’re not selling anything and that your goal is to understand how they currently handle a specific situation. This lowers defenses and reduces the urge to be polite.

Spend most of the conversation on a single recent episode. Ask them to walk you through the last time the problem occurred, step by step. Where were they? What triggered it? What tools did they use? How long did it take? Who else was involved?

This focus on the past is critical. When Drew Houston was building Dropbox, early conversations centered on how people actually moved files between computers and devices. By observing behavior rather than preferences, the team was able to design around real friction instead of imagined features.

Only toward the end should you ask about impact. What did this cost in time or money? What broke as a result? What did they do instead of their ideal work?

Throughout the interview, your job is to talk less than you want to. Silence is often where the most useful details emerge.

Questions that tend to work well

Rather than a rigid script, think in terms of an arc.

1. Start with context. “Tell me about your role and what a normal week looks like.”
2. Move to a concrete incident. “Can you walk me through the last time this problem came up?”
3. Then dig deeper. “What happened next?” “What made that frustrating?” “What did you try before this solution?”
4. End with reflection. “If this problem went away tomorrow, what would that change for you?”

Avoid hypothetical questions and avoid pitching your idea. If they ask what you’re working on, give a short, neutral answer and redirect back to their experience.

How to capture and synthesize what you learn

After each interview, write down what you heard the same day. Memory fades quickly, and patterns only emerge when notes are consistent.

A simple structure works well. Note the trigger, the steps they took, the tools involved, the time or cost, and any exact phrases they used to describe frustration or relief.

After your first 5 to 10 interviews, look for repetition. Are the same workarounds showing up? Are the same moments causing stress? When you hear the same story multiple times from people who don’t know each other, you’re onto something real.

This is how Intercom’s early team shaped their product. Des Traynor has talked about how hundreds of conversations revealed recurring support and engagement jobs that became the core of the product. The insight didn’t come from one loud user. It came from counting patterns.

What “good” looks like after 10 interviews

You don’t need certainty after 10 interviews. You need clarity. A good outcome is being able to clearly describe a specific problem in the customer’s own words, explain how they currently solve it, and articulate why that solution is inadequate.

You should also be able to say who this problem matters to and who it doesn’t. Disqualifying a segment is just as valuable as confirming one. If after 10 interviews everything still feels vague, that’s a signal. Either the problem isn’t painful enough, or the segment is too broad.

Common mistakes to watch for

The most common mistake is selling too early. The moment you start explaining your solution, the conversation shifts and honesty drops. Another mistake is talking only to people who can’t buy or decide. Their feedback may be interesting, but it won’t translate into a viable business. Finally, don’t overreact to a single enthusiastic response. Look for frequency, not excitement.

Do this week

If you’re ready to start, here’s a simple plan.

1. Write one clear decision you need to make in the next two weeks.
2. Define one narrow customer segment in two sentences.
3. Draft a short interview outline focused on a recent experience.
4. Reach out to at least 20 people across warm intros and outbound.
5. Schedule your first five interviews, even if the calendar feels intimidating.
6. Run the calls, focusing on listening more than talking.
7. Write up notes the same day using a consistent format.
8. After five interviews, review notes and look for repetition.
9. Adjust your questions if needed and complete five more.
10. Summarize what you learned in a one-page memo for yourself or your team.

Final thoughts

Customer discovery is uncomfortable because it forces you to confront reality early. But that discomfort is a gift. Founders who learn to do this well aren’t magically better at ideas. They’re better at listening, noticing patterns, and changing course before it’s too late.

Start small. Ten conversations can change how you see your idea entirely. The goal isn’t to be right. It’s to be less wrong, faster.