How to Validate Your Idea Properly

The golden rule: never talk about your idea

It is also the rule founders break most often.

Your conversation is about their world. Their problems, their routines, their frustrations, their spending. The moment you describe your product, everything changes. They stop telling you what's true and start telling you what they think you want to hear. Most people are polite. Polite feedback is useless feedback.

Keep your idea out of the conversation entirely. If they ask what you're building, deflect: "I'm still figuring that out. Right now I'm just trying to understand how people deal with this problem."

Ask about behavior, not hypotheticals

"Would you pay $20 a month for this?" is a bad question. People are terrible at predicting their own future behavior. They'll say yes to be encouraging, then never open their wallet.

Ask about what they've actually done. "How do you handle expense tracking today?" "When was the last time this problem cost you real money?" "What have you tried so far to fix this?"

Past behavior is the best predictor of future behavior. Current spending is the best predictor of future spending. Stick to things that have already happened.

Listen more than you talk

Aim for an 80/20 split: they talk 80% of the time, you talk 20%. Your job is to ask a question and then shut up.

This is harder than it sounds. You'll want to explain, correct, add context, or jump in when they pause. Don't. Silence is one of the best tools you have. People fill silence with the stuff they actually think, not the stuff they think you want to hear.

When something interesting comes up, follow it. "Tell me more about that" and "What happened next?" are often more useful than your prepared questions.

Go where it hurts

When someone says something with emotion (frustration, excitement, annoyance), stay there. Don't move on to your next question.

"You mentioned that drives you crazy. Why?" "What happens when that goes wrong?" "How much time does that cost you?"

The most useful data when validating an idea is usually behind the first answer, not in it. The first answer is the polished version. The follow-up is where people get specific.

Structure your notes as you go

Don't wait until after the conversation to organize your notes. Tag insights as they happen, while the context is fresh.

Five categories cover most of what you'll hear:

• Pain point: A problem or frustration they described
• Solution: Something they want or a workaround they use
• Pricing: Anything about money, budgets, or spending
• Objection: A reason they wouldn't use or pay for something
• Other: Anything useful that doesn't fit the above

Tagging as you go means your data is structured from day one. After five or six conversations, you'll start seeing which tags cluster together. That's when patterns form.

Yes, you can talk to friends and family

There's common advice that says never to talk to anyone you know. The logic: friends and family want you to succeed, so they'll sugarcoat their answers.

But if you're following the golden rule, that problem doesn't exist. They can't be biased about your idea if they don't know your idea. The bias comes from pitching, not from the relationship.

Anyone who fits your target customer profile is worth talking to. Your cousin who runs a small business has real invoicing frustrations whether or not she's your cousin. Her answers about how she tracks expenses today are just as real as a stranger's, as long as you never bring up what you're building.

Two caveats. It's harder to stay disciplined with people you're close to. You'll want to share your idea over coffee. Don't. And don't limit yourself to your inner circle. Your personal network is a starting point, not the whole sample. Use friends and family as people you talk to and as bridges. "Do you know anyone who runs a small business and handles their own invoicing?" is how one conversation turns into two.

End every conversation the same way

Close with a referral question: "Is there one person you could introduce me to who might have thoughts on this problem?"

This one question turns each conversation into two. By the time you've done five conversations, your list of people worth talking to has grown well beyond whoever you started with. The people who get referred to you are often better fits than your initial contacts, because they were recommended by someone who understands the problem.

What a good conversation looks like

A solid conversation runs about 20 minutes and follows a rough arc:

Start warm. Ask about their role or background (2-3 minutes). Move into context questions about how they currently handle the problem area (5 minutes). Then dig into specific experiences, pain points, and workarounds (8-10 minutes). Ask about spending and what they've tried (3 minutes). Close with the referral question.

By the end, you should have tagged several insights across at least two categories. If you walk out with just a general sense that "they seemed interested," the conversation didn't go deep enough.

Describe your idea on Evidnt and get tailored validation questions in minutes. Free, no account needed.

Next step

Knowing what to avoid is the other half of the job: 7 Mistakes Founders Make When Validating Their Idea.

Related

• What Is Customer Discovery?: The foundations
• 7 Mistakes Founders Make: What to avoid
• How Many Conversations Do You Need?: When you have enough data