From Gut to Product: The Startup Hunch

It starts with a “maybe”—that flicker of curiosity that nudges you in a crowded coffee line, on a support call, or reading through bug reports at midnight. A moment when your brain feels there’s something worth chasing—even if you can’t quite name it yet.

That’s the hunch. And Hunch by Bernadette Jiwa is a blueprint for what happens if you take that feeling seriously.

Not with validation frameworks. Not with market sizing. But by paying closer attention.

Because according to Jiwa, most breakout ideas don’t start with data. They start with noticing.

Hunches Aren’t Luck

We think of them as flashes—random, sudden, lucky. But Jiwa makes a sharper case: hunches are earned. They’re built.

Not through effort. Through attention.

Every time you feel a tiny itch—why is this like that?—you’re collecting dots. And if you keep collecting, the connection eventually shows up.

That’s not magic. That’s pattern recognition.

Jiwa breaks it down into three habits:

• Curiosity: Not passive scrolling. Active, persistent interest. Like Nick Disabato building Draft after seeing the same problem pop up with every freelance client—nobody knew how to handle proposal edits. He didn’t survey them. He noticed them.

• Empathy: Not just user interviews. Actually feeling what someone else feels. Spanx didn’t start in a focus group—it started with Sara Blakely solving her own wardrobe problem. One frustration, deeply felt, widely shared.

• Imagination: Seeing across domains. Remixing insights. Like a boy in Kenya rigging up a moving flashlight system to scare off lions and protect his cattle. Observation turned into invention.

These aren’t rare traits. But they require space. And most founders are too busy optimizing to slow down enough to see.

Solve Pain, Not Features

A lot of builders chase features. Jiwa says: chase tension. Look for the thing that feels off—even if no one’s talking about it.

GoldieBlox wasn’t data-backed. It was born from one quiet question: Why aren’t there engineering toys designed for girls?

That wasn’t a market insight. It was a gut pull.

And here’s the part indie hackers forget: your lack of expertise might be the edge. You’re not boxed in by how things “should” work. You’re just asking why they don’t.

How to Train the Instinct

Jiwa offers a system. Not rigid. Just repeatable.

1. Protect space to think. You won’t find a hunch mid-scroll. But you might find it in the shower. Or on a walk. Or staring at the ceiling. Let boredom do its thing.

2. Spot the weird stuff. Not just errors. Patterns. Why do users click the second CTA? Why do reviewers say “complicated” even after you simplified it?

3. Ask sharper questions. Not “what do people want?” Try “what’s missing that no one’s naming?” That’s how Nadia Eghbal approached open source: by listening to what wasn’t said.

4. Test tiny. Brett Kelly wrote a PDF—Evernote Essentials. No platform. No dashboard. Just a simple way to prove people would pay to make sense of clutter.

5. Reuse the insight. One hunch worked? Good. Now look sideways. Where else does that same pattern show up?

6. Ship small, ship now. Don’t roadmap it to death. Andrew S. Kaplan built Copy.ai by tossing out a quick demo. No splash, just signal.

It doesn’t have to be “right.” It just has to be real enough to react to.

Don’t Dismiss the Tug

If you’re only listening to dashboards, you’ll miss what makes something resonate.

The best products? The ones that feel inevitable in hindsight? They usually start with one person noticing what everyone else ignored.

So if something’s bugging you—some offhand comment, some repeated tension, some feature request that feels bigger than the feature—don’t file it away.

That might be the start.

Not of a business.

Of a better question.

And maybe, if you let it sit long enough… a better product.