The situation is over. Everybody is exhausted, the ordinary work has piled up, and the last thing anyone wants is a meeting about what just happened.
This is the moment the most valuable information the organization will acquire all year is available, and it is available for about three weeks.
Why the window closes
Two processes run simultaneously and both destroy the material.
Memory converts what happened into a coherent story. The confusion of the second day, the decision that was nearly made differently, the twenty minutes when nobody knew who was calling the insurer: all of that resolves into a clean narrative in which events followed sensibly from one another. This is not dishonesty. It is what memory does, and it is covered in the Judgment subchapter.
Meanwhile positions harden. As people describe their part in it, they commit to an account, and after a few tellings that account is what they remember and will defend.
Which is why a review held at six weeks produces a discussion of the story rather than of the event, and why the log matters so much: it is the only surviving record of what people actually knew at the time.
The two failures that make it worthless
It becomes an inquest. One instance of a finding being used against somebody converts every future review into a defensive exercise, permanently. People will attend, contribute, and say nothing true.
It becomes reassurance. A review concluding that everybody performed well under difficult circumstances is more common and equally useless. It happens when the senior person speaks first and sets a tone of appreciation, and the appreciation may be entirely deserved and it terminates the inquiry.
Both are avoided by the same structure, which is the subject of the next entry: the review examines the system rather than the people, and the senior person does not open it.
What is actually being looked for
Not who did what wrong. Four things, and none of them is about performance.
Where information failed to travel. Somebody knew something and it did not reach the person who needed it, and the reason is nearly always structural.
Where the arrangement did not survive contact. The card said one thing and the situation required another.
Where something was missing. A number, an access, a document, an authority nobody held.
And where the outcome was good and the process was not. The most valuable finding of all, because success conceals it and nothing else will ever surface it. A decision that worked out because somebody happened to be in the building is not a functioning arrangement.
Judge the decisions, not the results
The discipline that makes a review honest is the same one that governs any examination of past judgment: assess what was decided against what could have been known at the time.
Hindsight makes every outcome look foreseeable and every decision look obvious, which produces two errors. Sound decisions that went badly get treated as mistakes, teaching people to avoid anything uncertain. And poor decisions that went well get no examination at all, which is the more dangerous case because the method survives to be used again.
The log is what makes this possible. Without a contemporaneous record of what was known, everybody in the room is reasoning from an account that has already been adjusted.
The organization also needs an ending
The review has a second function that is rarely stated and matters to the people who lived through the thing.
A crisis that stops without being closed leaves people carrying it. Saying out loud what happened, what it cost, what was done well, and that it is finished performs something no memo accomplishes. Where somebody was harmed, that acknowledgement is owed rather than optional.
This is a different meeting from the findings review and should not be merged with it, because the analytical exercise and the human one interfere with each other. Both are worth holding.
The only measure that matters
Whether anything changed.
A review producing a document and no altered practice has consumed the window and returned nothing, and the next situation will surface the identical gaps. Three specific changes with a name and a date attached is a successful review. Fourteen recommendations is a report.
The entries that follow deal with how the review is run, how accountability is handled without turning it into an inquest, what rebuilding trust actually requires, and why organizations implement some findings and file the rest.
Edited by Patrick J. Wolf, PhD