A plan for a week you cannot predict

Nearly every planning method quietly assumes you own your hours: block the morning, protect the deep work, batch the calls. Then there are the weeks where a child is ill, the rota changes on Tuesday, or the job is answering whatever happened overnight.

People in those weeks usually conclude that planning is not for them. The opposite is true — but the unit has to change.

Plan the order, not the clock

A fixed schedule breaks on the first surprise and takes your confidence in the plan with it. A ranked list does not: when four hours appear unexpectedly, you take the top of the list; when forty minutes appear, you take the top of the short list.

So keep two lists, not a timetable. The things that need a long run, and the things that fit in a gap. The skill is having both ready — because the gap never announces itself in advance, and deciding what to do with it while standing in it costs more than the gap is worth.

Anchor the week to the one thing that is fixed

Even chaotic weeks have something stable: a shift pattern, a nursery pickup, a Monday meeting. Hang the few non-negotiable things off those anchors — the admin after the pickup, the difficult call before the shift — and they stop depending on the week behaving.

This is the same mechanism as an if-then plan: naming when and where beats naming what, and it beats it by a wide margin in the research (1).

Expect the week to be shorter than it looks

We predict our own timings optimistically, and reliably so: in the classic study, people finished in 56 days what they had predicted at 34, and even their deliberate worst case came in short (2). In an unpredictable week, the honest default is to plan for about half of what seems possible and treat the rest as a bonus.

That is not pessimism. A week that ends with two things done out of three planned feels like failure; the same week with two out of two feels like a week that worked. The work was identical.