One-Way Doors vs. Two-Way Doors
Jeff Bezos popularized a useful mental model for decision-making: classify decisions as one-way doors (irreversible or very costly to reverse) or two-way doors (easily reversible).
-
One-way door (Type 1): Choosing your primary database for a system that will store billions of records. Migrating later is possible but enormously expensive and risky. This decision deserves careful analysis, broad input, and deliberate trade-off evaluation.
-
Two-way door (Type 2): Choosing which JavaScript testing framework to use. If you pick Jest and later prefer Vitest, the migration is straightforward. This decision should be made quickly — don't waste a week debating it.
The mistake most organizations make is treating every decision like a one-way door. They form committees, write RFCs (Request for Comments — formal design documents), hold three rounds of review, and take six weeks to choose a logging library. This is paralyzing and wastes the scarcest resource in engineering: focused attention.
The opposite mistake — treating one-way doors as two-way doors — is equally dangerous. "Let's just pick DynamoDB and we can always switch later" ignores the reality that migrating a petabyte of data from DynamoDB to PostgreSQL is a multi-quarter project that will consume the team's entire capacity.
How to assess reversibility:
| Factor | Easy to Reverse | Hard to Reverse |
|---|---|---|
| Data migration required? | No or trivial | Large-scale migration |
| API contracts affected? | Internal only | External/public APIs |
| Team retraining needed? | Minimal | Significant |
| Cost of being wrong | Days of rework | Months of rework |
| Contracts/commitments | None | Multi-year vendor agreements |
Before spending significant time analyzing a trade-off, first ask: "Is this a one-way door or a two-way door?" If it's a two-way door, make a reasonable choice quickly and move on. Save your deep analysis for the decisions that are genuinely hard to undo.
Key Point: The reversibility of a decision should determine how much time you invest in analyzing it. Spending three weeks evaluating two-way-door decisions is waste. Spending three hours on one-way-door decisions is reckless. Calibrating your analysis effort to the reversibility of the decision is itself a critical meta-skill — one that distinguishes engineers who move fast and make good decisions from those who are either paralyzed by process or blindsided by consequences.