How to Run Customer Discovery Interviews Before Building Your SaaS

import DefinitionBlock from ‘@validation/ui/seo/definition-block.astro’; import AnswerBlock from ‘@validation/ui/seo/answer-block.astro’; import InlineSignup from ‘@validation/ui/components/inline-signup.astro’;

Why Most Founders Skip This Step

Discovery interviews feel slow. You have an idea, you’re excited, and sitting on Zoom asking questions when you could be building feels like a delay.

It isn’t. It’s the fastest way to find out whether what you’re about to build is something people will actually pay for.

The typical skip-discovery outcome: a founder builds a product based on assumptions, launches, and discovers the people they built it for either don’t exist in the numbers expected, don’t feel the pain acutely enough to pay, or already have a solution that’s good enough. That costs months. Interviews cost days.

Step 1: Find 10 People Who Have the Problem

Your ICP needs to be specific enough that you can point to real people. If you can’t find 10 people who match it in your network or in public communities, that tells you something important before you write a line of code.

Look in: your LinkedIn first-degree connections, Slack communities for your target vertical, subreddits where your ICP hangs out, Twitter/X following your problem space, and former colleagues.

A quick DM works: “I’m researching how [role] handles [specific problem]. Would you do a 20-minute call? Genuinely no pitch, just trying to understand the problem better.” Most people say yes when you’re asking for their expertise, not their wallet.

Step 2: Write an Interview Script Around Past Behavior

The Mom Test is the foundation: ask about past behavior, not future intent.

Bad questions:

• “Would you pay for a tool that did X?”
• “How important is this problem to you on a scale of 1-10?”
• “Do you think this idea would help you?”

Good questions:

• “Walk me through the last time you dealt with this problem.”
• “How do you currently handle it?”
• “What did you try before your current approach?”
• “What does it cost you, in time, money, or frustration?”

Write 5-7 questions. Leave space for follow-ups. The best question you can ask is “tell me more about that” when someone says something interesting.

Step 3: Run the Interview. Your Job Is to Listen

Set a 30-minute block. Spend the first 2 minutes on context, then ask your first question and be quiet.

Most founders talk too much in discovery interviews. They hear something that connects to their idea and start explaining the solution. Stop. The moment you start pitching, the interview is over. Now they’re evaluating your idea instead of describing their reality.

If someone says something surprising or contradicts your assumption, follow it. “That’s interesting. I expected you’d say X. Can you tell me more about why Y?” Surprises are where the valuable information lives.

Step 4: Find the Patterns

After 5 interviews, stop and read your notes before doing the next 5. Look for:

• Repeated phrases (people describing the problem with the same words)
• Consistent workarounds (the same DIY solution showing up across interviews)
• Emotional spikes (where people got animated or frustrated)
• Time and money numbers (what the problem actually costs them)

If you hear the same complaint in 3 out of 5 interviews, you have a pattern worth building around. If every interview surfaces a different primary pain, either your ICP is too broad or the problem isn’t their biggest one.

Step 5: Turn Patterns Into Decisions

Discovery data has two outputs: product decisions and messaging decisions.

Product: the consistent pain points tell you which features matter. If 8 out of 10 people mention the same frustration with their current solution, that’s your core feature. Everything else is secondary.

Messaging: the exact phrases people use should appear on your landing page. When your copy mirrors how buyers describe the problem, they recognize themselves immediately. That recognition is the fastest path to email capture. Don’t paraphrase. Use their words.

What Good Discovery Looks Like

You’ve done useful discovery when: you can predict what an interviewee will say before they say it, your landing page copy comes mostly from interview quotes, and you can name the five most common objections before a buyer raises them.

That’s the point. You’re not collecting opinions. You’re building a mental model of your buyer that’s accurate enough to make product and marketing decisions without guessing.