Time Blocking for People Whose Day Gets Taken Away From Them

Every guide to time blocking I read during my first attempt assumed a day that belongs to me. Block nine to eleven for deep work, protect it, decline meetings. On a ward, the day does not belong to me — it belongs to whoever deteriorates at ten past nine — and the first time I tried this properly I abandoned it inside a week, feeling like the failure was mine.

It was not mine. The method has an unstated precondition, and if the precondition is false the method inverts: a schedule you cannot keep does not organise your day, it generates a running tally of your failures. That tally is demoralising in a specific way, because at the end of a day where you worked hard and helped people, the artefact says you completed two of your seven blocks.

What I do now has three differences from the standard version, and I have taught it to enough colleagues to be reasonably confident it generalises beyond healthcare to anyone in support, operations, or client work.

The first is that I block half the available time, not most of it. On a ten-hour shift I will plan four to five hours. The other half is not “buffer” in the sense of contingency at the end — it is distributed, unlabelled, and expected to be consumed. When I planned eighty percent, the interruptions ate the blocks. When I plan fifty, the interruptions have somewhere to go and the blocks survive. The uncomfortable part is that fifty percent feels lazy when you write it down on a Monday. It is not; it is an honest estimate, and honest estimates look lazy compared to fantasies.

The second is that blocks name a stopping point rather than a duration. Not “90 minutes on the audit” but “audit through to bed 12.” The reason is that an interrupted duration is a failure and an interrupted checkpoint is a position. If I get pulled away at bed 7, I know exactly where to resume, and — this matters more than it sounds — I have completed something. Keeping that checkpoint in a daily plain-text log gives it somewhere to survive the interruption. Duration-based blocks have no partial credit, and a system with no partial credit will punish you every single time reality intervenes.

The third is a triage rule written before the shift, not during it. Mine is a single line: anything involving a patient interrupts anything involving a screen. It sounds obvious. Its value is not in the content but in having decided in advance, because deciding in the moment costs attention I do not have and, worse, produces inconsistent choices that I then second-guess for the rest of the day. Everyone I know who does interrupt-heavy work well has a rule like this, and most of them have never written it down.

The piece that took me longest to accept is that rebuilding focus is a block. After a genuine interruption — not a question, an actual event — returning to what I was doing takes me somewhere between five and fifteen minutes, and for two years I did not account for that anywhere. Now the fifty percent I leave unplanned is explicitly for interruptions and their wake, and once I named the wake I stopped being surprised by where my afternoons went.

What this method does not do is make me more productive in the sense that word is usually used. I do not get more done. What it does is make the plan match the day closely enough that I trust it, and trusting it means I actually look at it, and looking at it is the entire mechanism by which any planning system works.

If you have tried time blocking and it collapsed, I would ask one question before you try again: on the day it collapsed, was the plan wrong, or was the day wrong? I spent a long time answering that the second way.