Set the evidence threshold before you make the call
An Inc. column argues that leaders should decide in advance how much proof is enough, then structure the bet so it can be undone. The four questions behind it were not published in the summary available.
Most founding decisions are made on incomplete evidence. The question is not whether the evidence is complete, but how much of it you agreed to accept before you started looking.
That is the argument set out by Soren Kaplan writing in Inc. Certainty, he holds, is a trap; the stronger habit is to fix a threshold for action and then make bets that can be reversed.
The column offers four questions to do that work. It describes them as a way to weigh incomplete evidence, set the threshold, and revise a decision responsibly once new facts arrive.
The wording of those four questions has not been published beyond the headline and summary of the piece.
The practice underneath is still usable. Deciding in advance what would count as enough proof stops the goalposts moving later, when the founder has grown attached to an answer.
So does the second half of it. A bet you can unwind is a cheaper bet, and hiring and pricing are the two places where founders most often make one they cannot.
A pricing change can be run on a cohort before it touches the whole book. A senior hire can start as defined paid work before it becomes a contract.
Neither is free. Both convert an irreversible decision into an experiment with a stopping rule, which is what the threshold is for.
The revision clause matters as much as the threshold. A decision you refuse to revisit when the facts change is not conviction; it is an unmanaged bet.