Disagree and commit — reduce the decision drag of endless debate
Decisions that are made and executed at 80% confidence are almost always better than decisions endlessly refined but never taken.
Why it works
Decision perfectionism — refusing to commit until all stakeholders are aligned or all information is in — creates decision drag that slows execution, demoralizes teams, and paradoxically often produces worse outcomes than a 80%-confidence decision taken quickly and adjusted based on feedback. The "disagree and commit" norm (popularized independently at Amazon) allows productive disagreement without blocking action.
How to do it
- Establish a norm: anyone can raise a concern, but once a decision is made, all parties execute it even if they disagreed.
- Set a decision deadline for significant choices — if no decision by the deadline, the most senior person decides.
- Debrief decisions explicitly after 30–60 days, not as a retrospective blame session but as a learning loop.
- Distinguish "this decision is wrong and we should reverse it" (legitimate) from "I don’t like this" (commit and execute).
Evidence
Decision-making research supports the value of committed execution over perpetual deliberation for most reversible decisions; the cost of under-execution of an 80%-right decision typically exceeds the cost of a promptly corrected wrong decision. Experimental work on precommitment and deadlines shows that externally imposed deadlines improve performance over open-ended self-set ones, supporting the discipline of setting a firm decision deadline rather than deliberating indefinitely. (mechanistic)
Disagree-and-commit works for reversible, fast-feedback decisions; it is a poor default for irreversible high-stakes decisions where more deliberation genuinely reduces expected cost.
Sources
Common mistake
Using "disagree and commit" as a way to silence genuine dissent rather than as a norm that allows productive disagreement to be heard and then set aside — the disagreement must actually be registered, not suppressed.
Practice this with IX Coach
7 days free, then $40/month (~$1.30/day).
More practices for Designing a Calm Company (Fried & Hansson)
- Default to asynchronous communication
Treat real-time communication as the exception, not the default — most things do not require an immediate response.
- Protect mornings for deep work — no internal meetings before noon
Morning is when cognitive capacity is highest for most people; spending it on reactive coordination is a predictable waste.
- Set realistic goals and chip away at them — resist "crush it" culture
Aspirational goals that require heroic effort normalize crisis; realistic goals with steady progress normalize calm.
- Maintain at most three active priorities at one time
More than three active priorities is the same as no priorities — attention cannot be shared four ways.
- Make "no" the default — saying yes should require active justification
Every yes costs future time and focus; treat it as a commitment that requires real justification.
- Protect the advantages of small — resist headcount as a measure of success
Adding people to a team adds communication overhead faster than it adds output — protect the productivity of small.