The most expensive mistake in building products is building the wrong thing.

Not building it poorly. Not building it slowly. Building something nobody wants.

I've made this mistake. Most builders have. You spend weeks or months on something, launch it, and hear... nothing. Crickets. The rare comment that makes clear you fundamentally misunderstood the problem.

There's a way to avoid this. It's not complicated. It just feels uncomfortable.

Talk to people before you build.

The Expensive Mistake

Here's the pattern:

  1. You have an idea
  2. You get excited about the idea
  3. You start building
  4. Weeks later, you launch
  5. Nobody cares
  6. You wonder what went wrong

What went wrong happened at step 3. You skipped validation.

Validation isn't asking your friends if your idea sounds good. It isn't a Twitter poll. It isn't assuming that because you have the problem, others do too.

Validation is finding people who have the problem and understanding their world deeply enough to build something they'll actually use.

Finding People to Talk To

The first challenge: who do you talk to?

Find where they hang out online. Reddit communities. Twitter conversations. LinkedIn groups. Slack communities. Discord servers. Wherever people with your target problem gather.

Look for people actively complaining. Search for frustration. "I hate when..." "Why isn't there a..." "Does anyone else struggle with..."

Reach out directly. Yes, cold outreach. Something like: "I noticed you mentioned struggling with X. I'm researching this problem—would you be open to a 15-minute call to share your experience?"

Most people will say no. That's fine. You only need 5-10 yeses.

The Conversation Framework

When you get someone on a call, resist the urge to pitch.

You're not there to sell your idea. You're there to understand their problem. The moment you start pitching, you stop learning.

Here's the framework:

Start broad. "Tell me about your role and what you're working on." Get context.

Go to the problem. "Tell me about the last time you dealt with [the problem area]." Specific and recent is key.

Dig into behavior. "What did you try? What worked? What didn't?" You want to understand existing solutions and workarounds.

Understand stakes. "What happens if this problem doesn't get solved?" This tells you how much they care.

Explore value. "What would solving this be worth to you?" Be direct. If they hesitate, the problem isn't that painful.

Signals That Matter

After a few conversations, you'll start seeing patterns. Here's what to look for:

They've already tried solutions. If they've paid money or spent significant time on workarounds, the problem is real.

They get emotional. Frustration, anger, resignation—emotion means the problem matters to them.

They ask when they can use it. If they're leaning forward, asking about timelines, trying to give you their email—that's signal.

They offer to pay now. The strongest signal. "Can I prepay?" means you've found something.

Signals That Don't Matter

Some responses feel positive but mean nothing:

"That sounds cool." Polite enthusiasm. They're being nice.

"I'd probably use that." Hypothetical commitment. Worth nothing.

"You should definitely build that." Easy to say when it's not their money or time.

"Let me know when it's ready." Translation: "I'm not interested enough to actually engage."

Words are cheap. Look for actions: have they spent money, time, or effort on this problem already? Are they willing to commit something now?

How Many Conversations?

You don't need a hundred interviews. Five to ten usually reveals the pattern.

You'll know you've talked to enough people when:

  • You can predict their answers before they say them
  • The same pain points keep coming up
  • You're hearing the same workarounds and failed solutions

If every conversation surprises you, keep going. If you're nodding along because you've heard it before, you have enough.

What To Do With What You Learn

Three possible outcomes:

Kill the idea. Nobody cares. The problem isn't painful enough. The market is too small. This is valuable—you saved months of building the wrong thing.

Pivot. The real problem is adjacent to what you thought. The users are different than expected. The solution looks different. Adjust and validate again.

Build. Consistent pain, real attempts at solutions, willingness to pay. Now you can build with confidence.

The Discomfort Is the Point

Validation feels awkward. Cold outreach feels weird. Asking strangers about their problems feels intrusive.

That's exactly why most people skip it.

They retreat to the comfortable work: building. At least building feels productive.

But building without validation isn't productive. It's expensive procrastination. You're avoiding the discomfort of learning that your idea might not work.

Better to learn in a week of conversations than in a month of building.

Related Reading