Decision making like Jeff Bezos

First classify the decision by reversibility. A two-way door permits cheap correction, so delay can cost more than an imperfect first choice. A one-way door deserves greater scrutiny because retreat is difficult or expensive. For reversible decisions, the 70% rule supplies a stopping condition: act once you have enough of the relevant information, rather than waiting for certainty. In groups, disagree but commit separates debate from execution. Someone must own the final call; otherwise unanimity becomes a contest of stamina, and the person willing to argue longest effectively controls the system.

Speed is not the principle; calibrated speed is. Misclassifying an irreversible choice as an experiment can lock in serious costs, while treating every reversible choice as momentous creates delay and decision fatigue. The 70% rule is therefore not permission to ignore crucial unknowns—it applies only after you have judged the cost of being wrong and the feasibility of reversing course.