Before you give up a year of your life, pressure-test the idea.
Frame the problem in a sentence. Name the customer. List what has to be true for this to work — and what would kill it. Fifteen minutes, not fifteen pages.
No card · Spark plan · Your data never trains a model
Most founder ideas die from love, not from flaws.
The flaws are knowable on day one. Founders skip the work because it's boring. Six months later the capital is gone, the co-founder relationship is stressed, and nobody wrote down what we were actually trying to prove.
Of failed startups cite "no market need" as a primary cause — a signal available at idea stage, not month 18.
How long it takes to fill the Idea frame. Less time than a coffee with a generic advisor.
Every idea is forced to list three facts that would kill it. Most founders skip this. The ones who don't, ship.
A three-part frame you finish in one sitting.
One sentence. Who feels it, when, why it hurts.
Not a market. Not a category. A specific human in a specific moment doing a specific thing badly. If you can't write this sentence, you don't have an idea yet — you have a topic.
List every assumption, mark the load-bearing ones.
Every idea rests on a handful of beliefs — willingness to pay, channel, frequency, retention. OwnershipFlow surfaces them, flags the ones you can test cheaply, and routes them into Validation when you move forward.
Three facts that would end the idea today.
If the biggest customer already has this in-house. If the regulation changed last month. If the unit economics only work at millions of users. Write them down. The ideas that survive this are the ones worth a year of your life.
You start building with a thesis, not with a topic.
After one evening:
A one-sentence problem statement you can defend.
Every founder conversation after this compounds on the same foundation. Nobody argues the premise twice.
A list of testable assumptions, ranked by risk.
Validation picks these up. You start experiments on what matters, not on what's easy.
Three kill switches you check monthly.
The unromantic part of founder work: watching for the fact that would end the thing. Better to catch it at month three than month thirty.
A frame your co-founder and first hire can read in five minutes.
Onboarding the fifth person gets cheaper. The thesis stays the thesis.