A validation checklist is a claim about what you would reject. I went and checked mine against a real run.

Two numbers changed how I use it.

First, order beats content. Every list I have seen is good questions in no particular order, so people answer the interesting ones first. Sort instead by what an honest answer costs: an afternoon settles whether a problem is documented; it takes months to find out you cannot reach the buyer.

Most ideas die in the afternoon tier and never cost you the rest. That is the entire benefit, and it is invisible if you answer in any other order.

The first number. On one published run, 69 candidates in, 36 dropped. The single most common cause of death was the absence of a documented problem: 21 of the 36, more than twice the next reason at 9. That is the cheapest question on the list.

The second number is the uncomfortable one. Those drops carry 57 reasons between them. The rows do not divide up the dead ideas — most of them failed several checks at once.

Which kills a hope I definitely had: that if I answered the one objection I keep hearing about my idea, it would come back. Usually a weak idea is weak in several directions, and the objection you have heard is just the one somebody bothered to say out loud.

The other thing I got wrong for a while: I had one list. There should be two. A candidate the system sourced and an idea a person brings cannot be judged identically — there is no corpus to anchor a pasted idea against, so demanding documented evidence of it would kill every idea for arriving from outside.

And a checklist cannot tell you an idea is good. Every item is a way of failing. Passing them all is permission to run a real test, nothing more.

Both lists and the counted census: https://whittleos.com/guides/startup-idea-validation-checklist