The lag
The audit, the renewal and the resignation
Three of the five problems on this record arrived through events that happen on somebody else's schedule.
The standing slot for work with no ownerThe fastest of the five problems
week 5 · surfaced after 14 weeks
Reading supplier contracts before renewal
week 2 · surfaced after 22 weeks
The quarterly reconciliation of the project ledger
week 3 · surfaced after 31 weeks
Shadowing new starters through their first week
week 6 · surfaced after 47 weeks
Logging why a job was turned downThe slowest on the record
week 9 · surfaced after 64 weeks
Surfaced 5
Five 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, five surfaced as a problem. All five removals whose absence surfaced as a problem, in the order they surfaced.
If the lag of a removal is set by when somebody next stands in the place where its absence shows, then the practical question is what puts people in those places.
On this record the answer is unglamorous: dates in other people's calendars.
The three events
An audit. Somebody outside asks for a year of figures. This is the event that exposed the reconciliation, at week thirty-one. The date was known eighteen months in advance and appeared in nobody's thinking about the transition.
A renewal. A contract comes up and terms apply. This is the event that exposed the contract reading, at week twenty-two. The renewal dates were on file.
A departure. Somebody leaves and the organisation asks why. This is the event that exposed the shadowing, at week forty-seven, and it is the one of the three that no calendar can anticipate.
Two of the three were sitting in a calendar the whole time.
The general shape
External events that reveal removals share three properties:
- They are scheduled by somebody else, which is why they do not appear in an internal transition plan.
- They are infrequent — quarterly, annual, occasional — which is why the lag is long.
- They ask a question the organisation does not normally ask itself, which is why nothing internal has been testing the answer.
The third is the one that does the damage. An auditor does not ask whether the reconciliation is being done. They ask for the figures, and the figures are a thing the organisation believes it has.
A list almost everybody can write
The useful exercise takes twenty minutes and it is not about the removals at all. It is about the year.
| event | when | what it asks for |
|---|---|---|
| year end | annual | figures assembled across quarters |
| audit or review | annual or biennial | evidence of a process, not its output |
| contract renewals | various | that someone read the terms |
| statutory returns | fixed dates | records kept throughout |
| insurance renewal | annual | that the log exists |
| induction of a new starter | whenever hiring | that somebody can teach the job |
Write that list for your own organisation. Then read the removals against it, and the ones that touch an entry have a date rather than an unknown.
On this record, four of the five costly removals touch an entry on exactly such a list. The reconciliation touches year end and audit. The contract reading touches renewals. The refusal log touches nothing scheduled — which is why it took sixty-four weeks. The shadowing touches induction, which is scheduled by hiring rather than by the calendar but is just as knowable.
The one that cannot be diarised
Among the three, the shadowing is the exception, and the difference is worth being honest about.
Nobody can put "somebody resigns" in a calendar. What can be put in a calendar is the induction of the next person who joins, and on this record that is the moment at which the absence would have been visible: a new starter arriving to find that the week of sitting alongside somebody no longer existed.
The organisation did hire during the window. Four people arrived. At each arrival there was a moment when somebody could have asked what the first week was going to consist of, and the answer would have been the finding, eleven months before the finding arrived.
The event that reveals a removal is often not the event that punishes it. The induction revealed it. The resignations punished it. There were four inductions before the first resignation.
What this costs to act on
Nothing that is not already being done. The dates are known. The list of removals exists. The work is putting one against the other, once, at the end of a transition.
Against each of the thirty-eight entries, a single additional field: the next date on which this absence would be visible. For most of them the honest answer is "next week", and those can be ignored — they will report themselves. For a handful it is a date months away, and those are the ones to write down.
For one or two, there is no date at all. That is the refusal log, and it is the hardest case on the record, because there is no event to wait for. The only thing that works there is somebody periodically asking whether anybody has had to work something out that used to be written down — which is a strange question to ask and the reason it is strange is the reason it took sixty-four weeks.
The calendar that would have helped
It is worth being concrete about how small this is.
The organisation's year contained eleven dates on which somebody outside would ask it for something: a year end, an audit, two statutory returns, five contract renewals spread across the year, and an insurance review.
Eleven dates. All of them known at the start of the transition and none of them in the transition plan.
Set the thirty-eight removals against those eleven dates and four entries touch one. That is a five-minute exercise producing a list of four, of which three later became problems.
Nobody did it, and the obstacle was not difficulty. A transition plan is assembled from internal milestones — weeks one to eleven, the removals, the review — and the dates that matter belong to a different calendar that nobody thought to open.
Also on what stopped and when anyone noticed