One page, written before anything is announced. Its value is not that the plan survives contact, which it will not. It is that dependencies become visible, and that the stopping conditions get set by a version of you with no investment in continuing.
Section 1 — What changes, in behavioural terms
Not the objective. The specific action somebody performs differently, named by role.
Crew leaders enter completed work in the field app before leaving site, rather than submitting paper at the end of the week. Office staff stop re-keying paper tickets. Nothing else changes.
If this section cannot be written concretely, the change is not yet designed and the announcement will produce confusion that gets misread as resistance.
Section 2 — Why, in two sentences
The reasoning that supervisors will have to reproduce in their own words. Two sentences, because anything longer will not be repeated accurately.
We invoice eleven days after work is completed, and that gap is most of our cash problem. Entering at the site closes it to two.
Section 3 — What stops
What is being removed to make room. If the answer is nothing, say what the additional load is and who carries it.
Weekly paper submission ends. The Friday reconciliation meeting ends. Net effect on crew leaders is roughly ten minutes a day added; on office staff, about four hours a week removed.
Note that the cost lands on a different group than the benefit. That asymmetry should be stated rather than discovered.
Section 4 — Sequence
Steps in order, each with an owner and a date, and each with the condition that must hold before the next begins.
1. Talk to the three crew leaders others watch. Understand what will not work. — M. Reyes, by 14 Feb.
2. Adjust the design based on what they say. — M. Reyes, by 28 Feb.
3. Devices and access issued and tested on site. — D. Whitmore, by 7 Mar.
Condition: nobody starts until every crew has working access.
4. One crew runs it for three weeks. Ordinary crew, not the easiest. — J. Okafor, from 10 Mar.
5. Review with that crew. Fix what surfaced. — 31 Mar.
6. All crews. Paper accepted in parallel for two weeks only. — from 14 Apr.
7. Paper stops. — 28 Apr, no extension.
8. Encode: added to onboarding, in the supervisor checklist, reviewed each January. — D. Whitmore, by 31 May.
Section 5 — Checkpoints
Fixed dates for looking at it, including one deliberately placed in the difficult middle.
31 Mar — pilot review. 30 Jun — is it holding. 30 Sep — month-six check, expect this to be the low point. 31 Jan — annual review, decide whether to keep it.
The September entry matters most. Scheduling a checkpoint into the period where the change will feel like a mistake means the difficulty is met as an expected milestone rather than as evidence.
Section 6 — What would make us stop
Written now, while you have no stake in the answer. This is the most important section and the one most often omitted.
We stop if: entry compliance is below 70 percent at the 30 Sep check and we cannot identify a fixable cause; or the days-to-invoice figure has not improved at all by then; or two of the three crew leaders we consulted say it is making the work materially worse.
Stated in advance, halting is the plan operating rather than a reversal, which is the argument made throughout this Library for pre-committed exit criteria.
Section 7 — Who owns it after the launch
One name, and a person for whom this is part of their actual job in month five.
D. Whitmore owns this through 31 Jan. Reports at the monthly operations meeting whether or not there is anything to report.
Three rules for the page
Write it before the announcement. Written afterward, section six will be shaped by the investment already made.
Keep it to a page. If it will not fit, the change is too large to carry as one thing and should be split into two sequenced changes.
Update it when the sequence slips. A step that moves usually moves others, and the page is where that becomes visible rather than being discovered at step six.
The example above is illustrative. The organization, names, and figures are invented.
Edited by Patrick J. Wolf, PhD