Taking things out
The one-page version
Everything on this site, compressed to what somebody at week two of their own transition could act on in an afternoon.
The standing item for any other business
week 8 · still not missed
A written agenda for every one-to-one
week 6 · still not missed
The suggestion box
week 6 · still not missed
Weekly timesheets broken down by task
week 3 · still not missed
Not missed 4
Four removals were written down with the week each of them stopped. Four of them have not been asked for in the 78 weeks since. Four of the fourteen nobody has asked for, seventy-eight weeks on.
This is the whole of it, for somebody who has thirty-eight decisions in front of them and no time to read a site about one organisation's spreadsheet.
The finding
A shorter week works by subtraction. Something has to stop. On one record of thirty-eight removals across eleven weeks, watched for seventy-eight weeks afterwards:
| what happened | n | median lag |
|---|---|---|
| nobody has asked for it | 14 | — |
| crept back without a decision | 11 | 7 weeks |
| somebody asked for it back | 8 | 3 weeks |
| its absence became a problem | 5 | 31 weeks |
Of the nine removals somebody argued about at the time, none became a problem. Of the seventeen nobody remarked on, four did.
The twenty-four that came back came back at a median of just under half their old load, so the reduction was larger than fourteen out of thirty-eight suggests.
The mechanism
A contested removal has a watcher. Somebody objected, so somebody is looking, and when the consequence arrives they say so and are believed. Information arrives in a fortnight.
A quiet removal has nobody. It stops, nothing happens, and the nothing is read as evidence that it was not needed. "Nothing happened" and "nobody is reporting" look identical from inside.
The lag is not set by how important the thing was. It is set by how often somebody stands in the place where its absence shows.
Four things to do
Write the list. Five columns: what stopped, the week it stopped, whether anybody said anything (argued / asked / nothing), what became of it, and a sentence. The first three during the transition, about an hour in total.
Ask who has done it. Before agreeing any removal: "who here has done this?" If nobody, the silence means nothing. Mark it and ask somebody who does.
Ask what lived in the day that went. A month afterwards, of every team. This catches the removals nobody proposed, which on this record were the two slowest failures.
Make an appointment with the list. A named person, a calendar entry at twelve months, attached to the file. If you do only one thing on this page, do this one: on this record it would have caught two of the five problems about nine months early.
Three things not to do
Do not keep the quiet removals. Thirteen of the seventeen were fine and several are the biggest savings on the list.
Do not override objections on the strength of this. The mechanism depends on objections being worth making. In a place where objecting is pointless, the nine fast reports become nine silences.
Do not read silence as success. At week twenty-six, nineteen entries had produced nothing. Fourteen of those were genuine savings and three were problems that had not arrived. Nothing distinguished them.
The sentence for the report
This covers twenty-six weeks. Thirty-eight procedures were removed; nineteen have produced no event of any kind, and we cannot distinguish those that were unnecessary from those whose consequences have not arrived. [Name] will review the list at twelve and eighteen months.
Four lines. It weakens no claim in the report and it is the difference between a document that says what it knows and one that says what it has seen.
The caution on all of it
One organisation, one transition, thirty-eight rows, one list kept by one person who was not keeping it to prove anything. Five problems is a small number and nine-against-seventeen would not survive any statistical test.
What travels is the practice and the mechanism. The numbers stay here.
The two numbers to carry
If everything else on this page is forgotten, two figures are worth remembering because they are the ones that change a decision.
Just under half. That is what came back, by weight, across twenty-four returns. A shorter week recovers substantially more than the count of permanent removals suggests, because what returns returns trimmed. Anybody building a case for the arrangement should use this number rather than the headline.
Thirty-one weeks. That is the median lag before a removal's absence became a problem — longer than the pilot that would have reported on it. Anybody reading a six-month evaluation, their own or somebody else's, should hold this number next to it.
The first is the strongest argument for attempting a shorter week that this record contains. The second is the strongest argument for not believing the report at the end of it. Both came out of the same thirty-eight rows, and an account that gives one without the other is describing half of what happened.