Classify every decision as one-way or two-way before starting

Before applying any decision process, ask: can I reverse this if I am wrong?

Why it works

The costs of deliberation — time, cognitive energy, delay — are justified only in proportion to the cost of being wrong. For reversible decisions, the cost of being wrong is low (you reverse course), so the optimal investment in deliberation is also low. Misclassifying a two-way door as a one-way door inflates the deliberation cost proportional to an overestimated downside.

How to do it

1. Before any significant decision, ask: "If this turns out badly, can I undo it or substantially correct it within [reasonable time horizon]?"
2. If yes: two-way door. Apply a faster, lighter process.
3. If no, or with significant cost: one-way door. Apply the heavier, slower process.
4. Do this classification explicitly and quickly — the goal is not to deliberate about the classification.

Evidence

Decision theory supports matching deliberation depth to decision reversibility and stakes. The specific two-way/one-way framing is practitioner advice from Bezos; the underlying principle — that deliberation cost should be proportional to the cost of error — is consistent with decision-theoretic reasoning. Bezos’s 2015 shareholder letter is the primary source for the classification itself, framing reversible decisions as ones that can be reopened and therefore need not carry the weight of a heavyweight process. (mechanistic)

The framework is widely cited in business contexts but has not been studied as a decision-improvement intervention. Its value is logical, not empirically established.

Sources

• Bezos, J. (2016). 2015 Letter to Shareholders. Amazon.com, Inc. (filed with the SEC as Exhibit 99.1 to the 2015 Annual Report).

Common mistake

Spending significant time classifying a decision that is obviously reversible — the classification should take seconds, not minutes.