Decisions.

The framework

1. 01

   Give the decision the four properties that make it bind.

   A commitment holds when it is written by you, effortful, visibly your own choice, and heard by someone specific. A note in your own diary fails the last two. So does the team update that buries the decision inside a paragraph of hedges, because "we're likely to start moving away from" is a weather forecast. Morton Deutsch and Harold Gerard showed in 1955, in the Journal of Abnormal and Social Psychology, that people who wrote their judgement down and committed to it publicly were markedly harder to shift under group pressure than those who kept it private. Write the sentence. Send it to a person with a name. Put a date on it.

2. 02

   Choose the audience who will chase you, not the one who will clap.

   Peter Gollwitzer, professor of psychology at New York University, published work in Psychological Science in 2009 showing that announcing an intention to an audience that acknowledges it can reduce the effort people then put in — the recognition lands early and quietly does some of the work the achievement was meant to do. The audience matters more than the announcement. Avoid the applause room: the all-hands, the post, the encouraging friend who says it sounds brave. Choose the witness with standing to ask about it in three weeks and no incentive to be gentle — the person the decision costs, the customer who will plan around it, the board member who writes things down.

3. 03

   Commit to the observable, not the intention.

   State what will be visibly true, by when, to whom — a commitment nobody could check is a mood with a deadline attached. "We are getting more disciplined about discounting" is unverifiable; "from the first of next month no discount above ten per cent leaves this company without my signature, and Ana holds the log" can be checked by a stranger. Then take the step that makes reversal cost something real: send the notice, publish the date, move the budget line, tell the people whose plans now depend on it. If backing out would cost you nothing but a shrug, you have forecast a decision rather than made one.

4. 04

   Publish the reversal rule in the same breath.

   The same consistency pressure that gets decisions executed also gets dead decisions defended, which is how a capable operator ends up arguing in public for a position they privately abandoned weeks ago. Disarm it in advance. In the message that announces the decision, name the condition under which you would reverse it: the number, the date, the signal you would treat as evidence you were wrong. Stated beforehand, a reversal reads as the plan working exactly as designed and costs you nothing. Left unstated, it reads as losing an argument you staged yourself — and that quiet fear is the real reason most decisions never get said out loud at all. So settle the ninety-day version before you send anything: what would have to be true for you to reverse this without embarrassment, and where can another person read it?