70% Rule
The 70% Rule treats incomplete information as a cost of timely action: decide when the evidence is sufficient to support a choice, before collecting the remainder costs more than it clarifies.
Jeff Bezos’s decision threshold is startlingly low: act with roughly 70% of the data, because waiting until you have 90% already makes you too slow.
E1The missing 30% has a price
A decision consumes two scarce resources: information and time. Early evidence moves you beyond guesswork, but later evidence may arrive only after additional delay. The rule sets a stopping point: once the available facts make one option defensible, stop treating complete certainty as the standard. This resembles the Secretary Problem in one narrow respect—search itself needs a stopping rule—and works best when paired with Estimate, Then Reconcile, so the choice remains revisable as new facts arrive.
E1Seventy percent of what?
The number is not proof that your evidence is representative or predictive. A confident estimate built from poor signals can still be an illusion of validity. Nor does speed automatically dominate: the right threshold depends on the real alternative—including the cost of delay, the cost of error, and whether the decision can be repaired.
Name the decision-changing unknown
For tomorrow’s delayed decision, list what you already know and identify the single missing fact most likely to reverse your choice. If you cannot name one—or cannot obtain it before delay becomes costly—make the best-supported choice now and schedule a specific point to reconcile it with new evidence.
Episodes that teach this
-
Jeff Bezos Decision Making Framework | How to Make Fast Decisions?
· explained at 4:11
2,345 views
“Take the decision when you have around 70 percent of the data”; waiting for 90% is “too slow.”