Idea — a hypothesis, not a plan

You don't have an idea problem. You have a decision problem: nine ideas, zero comparisons, and a background hum of guilt about all of them. The IDEA stage exists to fix that — cheaply.

You have 9 ideas. Which one gets the next 6 months?

The failure mode isn't lacking ideas — it's that every idea lives at the same foggy altitude: "AI thing for recruiters" isn't a hypothesis, it's a vibe. An idea becomes a candidate the moment you can write three sentences about it:

1. Who has the problem, and how do they suffer today?
2. What would you build that they'd pay for?
3. What would have to be true for this to work?

If you can't write #3, you don't have an idea — you have a mood. That's fine; write it down anyway. The point of the IDEA stage isn't commitment, it's comparison. Every idea gets a row. Rows are cheap; months aren't.

Idea debt is decision debt

The folder of "someday" apps feels like an asset library. It's a liability: every undecided idea taxes attention, blurs the numbers, and lets you feel busy without choosing. Enter every idea — then let the portfolio do what your memory can't: rank them honestly, side by side, on the same rubric.

Stress-test before you commit

Before anything earns research time, put the three sentences in front of someone who isn't you. Other founders will find the holes your optimism papered over — and occasionally tell you the idea is better than you thought. Both outcomes are cheaper than building.

Where the OS enforces this

• The portfolio is the idea shelf: every idea is a row, every row gets scored, and nothing lives in your head.
• The Opportunity Score ranks ideas on the same 12 signals — gut feel loses to arithmetic.
• The playbook makes you write the hypothesis down: who hurts, what you'd build, what must be true.
• The demo product (ChurnPilot) shows what a fully worked idea looks like before you commit your own.

Next: Research the market before yourself.