I published one idea-sourcing run unedited, with every rejection and its reason attached, and then counted the deaths.

Two numbers were not what I expected.

The run: 69 candidates, 36 dropped, 33 through the checks, 5 on the shortlist.

The first number. The leading cause of death was not competition, not pricing, not a crowded market. It was that there was no documented problem behind the idea at all — 21 of the 36 drops, more than twice the next cause.

That check is about evidence rather than about the market. It means the sweep went looking for somebody describing the problem in writing and did not find them. Two honest readings: the problem is real but nobody wrote it down, or the idea invented the problem. The check cannot tell those apart, and neither can I from the outside.

The second reading is the expensive one, because nothing about it feels wrong from the inside. An idea built on a plausible-sounding gap looks exactly like an idea built on a real one, right until launch.

The second number is the one that changed how I read every funnel chart since. Those 36 dead ideas carry 57 reasons between them, because 21 of them failed more than one check.

So the census does not partition the drops, and a weak idea is rarely weak in one direction. The idea with no evidence behind it is usually also the one that is hard to sell, because both come from the same missing piece of work.

The practical consequence for anyone holding an idea: answering the single objection you keep hearing usually does not rescue it. That objection is the one somebody bothered to say out loud, not the only thing wrong.

And the survival rate reads harsh until you notice the alternative. A sweep that returned forty promising candidates would have returned the 21 with no evidence behind them, and I would have spent my reading on those.

The full census: https://whittleos.com/guides/why-startup-ideas-fail