The change is working. People are doing it, the early difficulty has passed, and the results are beginning to appear.
It is still not permanent. It is currently being sustained by the leader continuing to attend to it, and that is a resource with a limit and an eventual end date.
The test
Would this survive the leader leaving?
If a successor arriving in eighteen months, knowing nothing about why the change was made, would find it operating and would have to make a deliberate decision to undo it, the change is encoded.
If it would quietly lapse because nobody was asking about it, the change is not a change. It is a practice being held in place by attention, and it will end when the attention does.
The four places it has to live
Hiring. What is asked in interviews and what a job description requires. An organization that changed how it works and did not change who it recruits will regress within two hiring cycles, without anybody deciding to.
Promotion. The strongest signal available. Whoever gets promoted is the organization’s real statement about what it values, and it is believed over anything said. A change that has never affected a promotion decision has not been adopted.
Budget. A change requiring resources that were never allocated is dependent on somebody choosing to fund it each year against competing claims. Written into the base, it persists by default rather than by advocacy.
Calendar and process. A step in an existing form, an item on a standing agenda, a scheduled review with a named owner. This is the cheapest and most reliable of the four, and it is the one that operates without anybody remembering why.
Remove the alternative entirely
The most common reason a working change reverts is that the old way was left available for a small number of exceptional cases.
The exceptional cases grow. Under pressure people revert to the familiar, and each reversion is individually justified by the circumstances of the day. Within a year the exception is the practice and the new system handles the easy cases.
Where the old system genuinely must be retained for something, name what and put a boundary on it. The paper form is used only for the two accounts that cannot be moved until their contract ends in September. Anything vaguer becomes an open door.
Record why it was made
An encoded change with no attached reasoning creates a problem some years out.
The step exists, nobody knows why, and it looks like inefficiency. Eventually a capable person removes it, and the failure it was preventing returns. This is how organizations lose things they had genuinely learned.
One line attached to the practice is sufficient. This review was added in 2027 after the notification failure. A successor can then assess whether the reason still holds, which is a legitimate decision, rather than removing something they assume is habit.
Withdraw attention deliberately
At some point the leader has to stop asking about it, and how that happens matters.
Attention that simply fades is read as the change having been abandoned, even when it is fully embedded. Attention withdrawn explicitly is read as completion.
Say it. This is now how we do things, it is in the process and the review schedule, and I am going to stop raising it in this meeting. Somebody owns it and here is who.
Then leave one scheduled check, once or twice a year, which is the difference between a practice that is maintained and one that decays quietly while everybody assumes it is fine.
The honest timeline
Leaders consistently underestimate this and the underestimate causes real damage, because programmes are resourced and staffed against it.
For anything altering how people actually work, eighteen months to two years is a realistic period from announcement to a state where the change would survive your departure. The difficult middle runs from about month three to month twelve.
A leader who has budgeted six months will withdraw attention at the point of maximum fragility, conclude that the organization did not adopt it, and have manufactured exactly the pattern that makes the next initiative harder.
Which is an argument for attempting fewer changes rather than for working harder at them. One change carried to encoding is worth considerably more than three abandoned at month five, and it is also the only version that leaves the organization more able to change next time.
Edited by Patrick J. Wolf, PhD