The unknown that would change the decision

The unknown column has one job, and most people never ask it to do that job. It exists to identify the single missing piece of information that would change what you do.

Not the information that would make you more comfortable. Not the information that would make the memo look thorough. The information that would change the decision.

The test takes about four minutes and it works like this. Take an item from your unknown column. Write the two or three answers it could return. Under each answer, write what you would do. If the action is the same under every answer, the item is comfort, and it belongs in a separate list called things I would like to know. If the action differs, the item is decision-relevant, and it should be at the top of the column with a name and a date next to it.

This is also the sentence that most improves a professional document. Naming the single piece of evidence that would most change your conclusion pre-empts the obvious challenge, demonstrates that you looked, and is frequently the most respected paragraph in the report. It reads as command of the material rather than as hedging, provided it is specific. "Further data would be helpful" is hedging. "If the two missing months of payroll data show the same pattern as the other ten, the conclusion holds; if they show the reverse, the finding on overtime does not survive" is a position.

There is a second move available once the decision-relevant unknown is named, and it is the one that separates people who are good at this from people who are merely careful. Ask whether the decision actually depends on the answer.

Frequently it does not. Two remaining explanations for a delay may both be addressed by the same intervention. Two possible reasons a customer left may both point to the same change. Where that is true, saying so converts an apparent gap into a demonstration of judgement, and it releases the decision immediately.

The corresponding failure is to name an unknown, declare it decisive, and use it to delay. Delay dressed as inquiry is rarely challenged, because a request for more information reads as care whoever makes it. The test above is what makes the challenge possible, including from you to yourself: what would each answer make you do.

Be careful not to run that argument in only one direction. Over-search is a real failure mode and it is not established as the more common one. People also avoid decision-relevant information at high rates, for reasons that have nothing to do with rigour: to protect a belief, to keep a hope alive, to stay motivated. Both failures are common, so which one you are running on this decision is a question about this decision, and the two-answer test settles it in four minutes either way.