You have to decide now and the facts will change tomorrow. This is the ordinary condition of a crisis rather than an unusual difficulty within one.
The instinct is to wait for the picture to settle. It rarely settles in the time available, and the cost of waiting is not visible while you are paying it.
Say that it is provisional
The whole technique is a single sentence attached to the decision.
We are closing the north building for the rest of the week. That is based on what we know this morning, and if the engineer’s assessment tomorrow says otherwise we will reopen it.
This costs nothing and it changes what a later reversal means. A decision announced as provisional and then revised is the plan working. The same decision announced flatly and then revised is a leader who was wrong, and the organization will read the second version as instability even though the underlying judgment was identical.
It also solves the problem of the leader who cannot reverse. Provisional framing removes the standing cost of changing course, which is the thing that traps people into defending positions they no longer hold.
Name the trigger
Provisional without a condition is just vague. The useful version specifies what would change it and when that will be known.
If the sample results are above the threshold, we extend to the second building. Those arrive Thursday morning and we will decide by noon.
Three things follow. Everybody knows what to watch for. The revision, when it comes, is legible as the plan operating rather than as a change of mind. And you have committed to a decision point, which prevents the drift in which a provisional decision quietly becomes permanent because nobody revisited it.
Reversibility governs the pace
The general principle covered in the Decision-Making subchapter applies with more force here, because the information is worse and the pressure to act is higher.
Reversible decisions should be made fast and adjusted. Closing a facility for a day, holding a shipment, standing up extra staffing: all cheap to undo, and delaying them to gain certainty is nearly always the wrong trade.
Irreversible decisions should be resisted until they cannot be. Terminating somebody, making a public admission of cause, committing to a remedy, signing a settlement. Each is permanent, each is being pressed for by somebody who wants resolution, and each is much better made in week three than in hour six.
A large part of crisis judgment is simply keeping those two categories separate while everything feels equally urgent.
Prefer decisions that preserve options
Where two responses are roughly comparable on the merits, take the one that leaves more available afterward.
Suspending a person with pay preserves every subsequent option. Terminating them forecloses all of them and cannot be undone if the investigation turns. Halting a process is recoverable; announcing that it was defective is not.
This is not indecision. It is recognising that in a situation where information is arriving, the value of a decision includes what it does to the decisions that follow, and early foreclosure is the most common way a manageable situation becomes an unmanageable one.
The pressure to be certain
People under stress want confidence from whoever is in charge, and the demand is real. Meeting it by manufacturing certainty is the single most expensive error available.
An assurance given to calm a room becomes a commitment. Everybody will be back in the building Monday, said to end a difficult meeting, is now a promise that will either be kept at cost or broken visibly.
What people actually need is not certainty about outcomes. It is certainty that somebody is in charge, that there is a process, and that they will be told the truth. Those can be supplied honestly.
I do not know when the building reopens. I know we will have the assessment Thursday and I will tell you the same day either way. That answer holds a room, and it does not create a debt.
Write down why
Every decision taken during a crisis will be examined afterward against information that was not available when it was made.
A short note in the log of what was known and what drove the choice is the only defence against hindsight, and it takes under a minute. Without it, a reasonable decision made on partial information is indistinguishable from a careless one, including to the person who made it.
That record also protects the people below you, whose decisions will be reviewed by somebody who was not there and now knows how it turned out.
Edited by Patrick J. Wolf, PhD