What came back
Creeping back is not a decision
Eleven procedures reappeared without anyone agreeing that they should.
The daily stand-upBack twice a week instead of five, standing and timed
week 0 · crept back after 3 weeks
Chasing unpaid invoices every weekBack in full
week 3 · crept back after 2 weeks
The spare key logReplaced by something else doing the same job
week 10 · crept back after 14 weeks
Two people on every delivery
week 4 · crept back after 9 weeks
Crept back 4
Four removals were written down with the week each of them stopped. Every one of them produced an event inside the 78 weeks watched. Of these, four crept back. Four of the eleven that crept back without anyone deciding.
Eleven of the thirty-eight came back without a decision. Nobody proposed it, nobody approved it, no meeting discussed it. The thing simply started happening again, at a median of seven weeks.
For eight of the eleven, nobody on the record can say who restarted it.
How it happens
The mechanism is individual and entirely reasonable at every step.
A procedure stops. Somebody who relied on it finds a small gap. They fill the gap themselves, privately, because that is less effort than raising it — a note to themselves instead of the handover, a glance at the diary instead of the formal check, a quick count instead of the monthly one.
Then a second person does the same. Then the two of them coordinate, informally, because coordinating is easier than duplicating. Within a couple of months the procedure exists again, assembled out of individual habit, usually at a lower weight and with no documentation at all.
No one did anything wrong. No one did anything, in the sense of taking an action anybody would describe as a decision.
Why this is mostly good
The eleven that crept back came back at a median of forty-five per cent of their old load. They are, with the substitutions, the cheapest returns on the record.
That is not an accident. What reassembles is only the part somebody actually needed, because each person rebuilt only the bit that was costing them something. The procedure that comes back is a filtered version of the one that went, filtered by actual use.
An organisation could not design that reduction. It emerged, and it emerged accurately, which is more than most deliberate reductions manage.
Why it is nonetheless a problem
Three reasons, and they are cumulative.
The week contains things nobody knows about. After eleven quiet returns, the organisation's own account of what it does is wrong in eleven places. The list says these things stopped. They did not.
The returns are undocumented. A procedure that was written down somewhere is now a set of private habits. It will not survive the person who holds it, and nobody will know when it stops a second time.
The saving is overstated. If a report says thirty-eight things were removed, and eleven of them are quietly running again at half weight, the hours do not add up — and the discrepancy shows up as unexplained pressure on the week rather than as a line in a report.
A procedure that crept back is doing its job and is invisible to every system the organisation has for knowing what it does. That is a better outcome than it failing, and it is not a good outcome.
The specific case worth dwelling on
The induction folder came back at twenty-nine weeks — the slowest return on the record — and nobody can say who restarted it or when it started being current again.
It is now maintained. It is not maintained by anybody whose job it is. When that person leaves, it will stop being maintained, and there will be no entry on any list, because the first time it stopped it was recorded and the second time will not be.
The record has caught one cycle of a thing that will now cycle invisibly.
What to do about it
Nothing during the transition — there is no way to prevent a quiet return and no reason to want to.
What works is a question at the read-back, six months in, asked of each team about their own entries:
"Which of these are we actually doing again?"
Ten minutes, and nothing else makes the eleven visible at all. On this record it was asked, informally and late, and it produced seven of the eleven immediately. The other four surfaced at the eighteen-month pass.
Having found them, the useful follow-up is not to stop them. It is to write down what came back, at what weight, and who is doing it — which converts eleven private habits into eleven ordinary entries that the organisation knows about and that will survive their current holder.
That is twenty minutes of work and it is the difference between a transition that produced a shorter week and one that produced a shorter week plus eleven undocumented procedures.
Why eight of eleven have no author
The most striking number in this group is that nobody can say who restarted eight of the eleven, and the explanation is not that people forgot.
A quiet return has no single moment. One person fills a gap privately. A second does the same. At some point the two coordinate. There is no step at which anybody restarted anything; there is a gradient from nobody doing it to several people doing it, and asking who restarted it is a question the history does not contain an answer to.
This has a practical consequence that nobody anticipates. When the organisation later wants to change the restored procedure — resize it, document it, hand it to somebody else — there is no owner to talk to and no decision to reopen. It has to be rediscovered from scratch, usually by asking around.
Three of the eleven were formalised at the eighteen-month read-back for exactly this reason: not to legitimise them, which nobody cared about, but so that somebody's name was against them.