For reversible decisions, act at 50-60% information — experiential learning from doing exceeds deliberation gains

Why This Is a Rule

Information has diminishing returns in decision-making. The first 50% of information dramatically narrows the option space and clarifies the key trade-offs. The next 20% refines your estimate marginally. The final 30% costs disproportionate time and effort while barely shifting the decision. For reversible decisions, the crossover point — where acting and learning from outcomes produces more information than continued deliberation — occurs around 50-60%.

Jeff Bezos frames this as the 70% rule for high-velocity decisions (he uses 70% for all decisions; for purely reversible ones, 50-60% is appropriate because reversal cost is low). The key insight is that information from deliberation and information from experience are not equivalent. Deliberation produces hypothetical understanding. Experience produces ground-truth feedback. For reversible decisions where you can iterate, ground-truth feedback from one attempt often exceeds weeks of hypothetical analysis.

The 50-60% threshold also accounts for information availability bias: the information you haven't gathered yet may not exist in accessible form. You might spend weeks seeking data that turns out to be unknowable until you try. Acting at 50-60% converts "unknowable in advance" into "learnable through experience."

When This Fires

• For any decision classified as a two-way door (Classify every decision as one-way or two-way door before deliberating — minutes for reversible, days for irreversible) where you're deliberating beyond the initial information gathering
• When you feel "not quite ready to decide" on a reversible choice — check if you're past 60%
• When analysis paralysis hits on a low-stakes decision despite having substantial information
• When the remaining information would take significantly longer to gather than acting and iterating

Common Failure Mode

Waiting for certainty on reversible decisions: "I want to be 90% confident before choosing." For a reversible decision, the cost of waiting from 60% to 90% confidence is weeks of deliberation. The cost of being wrong at 60% is the reversal effort — which, by definition, is low for two-way doors. The math strongly favors acting at 60%: low cost of being wrong × high probability of learning from experience > high cost of waiting × marginal confidence gain.

The Protocol

(1) For a reversible decision, assess: roughly what percentage of the information you'd ideally want do you currently have? (2) If >50-60% → act. You have enough to make a reasonable choice, and the remaining uncertainty will resolve faster through experience than through analysis. (3) If <50% → gather more, but set a time-boxed limit on additional research (Classify every decision as one-way or two-way door before deliberating — minutes for reversible, days for irreversible). Don't let "just a bit more research" extend indefinitely. (4) After acting, pay attention: what did you learn that you couldn't have learned through deliberation? This calibrates your threshold over time. (5) If the decision proves wrong → reverse it. That's what two-way doors are for. The reversal itself produces information about what matters, refining future decisions.