How I decide if a product is even worth building

Curiosity is a wonderful research assistant and a terrible CEO. Left unsupervised, it turns every interesting problem into a roadmap item, then a prototype, then a quiet product that never quite finds a market. Deciding what is worth building is less about inspiration and more about refusing to fund hobbies that have learned to wear a pitch deck.

My filter is simple on purpose: is the pain urgent, will someone pay, are we differentiated, how fast is first value, and what evidence would kill the idea early? If those answers are soft, the build should stay small or stop. Soft answers are not encouragement. They are fog.

Curiosity builds vs evidence builds

**Bad founder pattern:** build for months because friends said “cool,” then discover the problem was mild and the alternatives were free. **Good founder pattern:** interview for urgency, test a thin offer, and kill politely when the market shrugs. YC-style user obsession is not a vibe. It is a filter.

**Bad consumer idea:** lifestyle brand first, job second. **Good consumer idea:** a sharp job people already fail at weekly, then brand as expression of that job.

Run the filter before the sprint

1. Write the job in one sentence and the workaround people use today.
2. Talk to ten people with that job. Score urgency and frequency honestly.
3. Test willingness to pay with a real ask: deposit, preorder, or paid pilot.
4. Name your wedge and why a bigger player will ignore it for a while.
5. Prototype only the path to first valuable outcome. Time it.
6. Write kill criteria before you fall in love. Review them on a calendar, not vibes.

Y Combinator’s Library remains useful on talking to users before scale. Related: why great products fail before product-market fit.