What a procedure is actually doing now

Two questions that are not the same

When you pick up a procedure and ask whether it can stop, there are two different questions in front of you and they are easy to conflate.

1. What was this introduced to prevent?
2. What is it doing now?

The first has an answer in most cases, if anybody has been there long enough to remember. The second almost never has an answer, because nobody is measuring it.

The gap between the two is where the whole difficulty lives. A procedure introduced to prevent an error that has not occurred in six years may be preventing it successfully, or the conditions may have changed so that the error is no longer possible, or the error may be occurring and going unrecorded. From inside, those three states look identical.

What the record shows about the gap

On this list, the original reason was known for twenty-six of the thirty-eight. For the other twelve nobody could say, and in four cases the person who introduced it had left years earlier.

Sorting the outcomes by whether the original reason was known gives a result worth stating carefully:

Those numbers are too small to carry weight on their own and the difference between them is not large. What is worth noting is the direction, because it is the opposite of what the folklore predicts.

The folklore says: if nobody remembers why we do this, it is safe to stop. On this record, the removals where somebody could say exactly why the thing had been introduced were not noticeably safer. Knowing the original reason told you about the past, and the question in front of you was about the future.

Why the original reason misleads

A remembered reason is a story, and a story is persuasive in proportion to how vivid it is rather than how current.

Somebody remembers the incident that led to the second reviewer. They remember it because it was memorable, which usually means it was bad and it was unusual. The memory makes the procedure feel load-bearing. What it actually establishes is that the error happened once, in conditions that may or may not still obtain.

Meanwhile the procedure with no remembered origin feels weightless. It cannot defend itself in the meeting. Nobody will argue for it, which — as the rest of this site is about — means nobody will be watching when it is gone.

A better question

The question that behaves well is neither of the two above. It is:

If this stopped today, who would be the first person to find out, and roughly when?

It is a better question because it has an answer that is checkable, and because the answer tells you what to do rather than what to think.

• If the answer is "the person who does it, next Tuesday" — stop it. The feedback is fast enough that you will find out.
• If the answer is "the finance team, at year end" — stop it, and put a note in the calendar at year end.
• If the answer is "nobody, until something goes wrong" — that is the dangerous category, and it is not a reason to keep the procedure. It is a reason to decide, in advance, who is going to look and when.

The third case is where the five costly removals on this record all sat. None of them was kept. None of them should necessarily have been kept. What was missing was the second half of the sentence.

What this costs

Asking that question of thirty-eight items takes an afternoon. On this record it was not asked of any of them, and the list has still been more useful than most of the material published about shorter weeks, for the single reason that it was written down.

The sequence worth copying is small: write the removal down, ask who would be the first to find out, and write that down too. Three columns. The third one is the one nobody fills in, and it is the only one that does any work after the first month.

The procedure that defends itself

One pattern recurs often enough on this record to be worth isolating.

Four of the thirty-eight entries had, at some point, prevented something. Not hypothetically: there was an identifiable incident, and the procedure had been introduced immediately afterwards and had been in place ever since.

All four were argued about at the time of removal, and in each case the argument took the same form — somebody told the story of the incident. The story is what makes the procedure defensible, and the story is a fact about the past.

Three of the four were removed anyway and none has caused trouble in seventy-eight weeks. The fourth was kept.

The pattern is not that the stories were wrong. It is that a procedure with a story attached gets a hearing and a procedure without one does not, and whether a procedure has a story is mostly a question of whether anybody who was there is still in the building.