Work expands so as to fill the time available for its completion. Cyril Northcote Parkinson wrote that line in 1955 as satire, a joke about bureaucracies inventing work for themselves. Seventy years on, it's the most reliably true sentence in productivity, and everyone who's ever been given a month for a two-week job knows it. The report due Friday takes until Friday. The same report, due tomorrow, somehow takes until tomorrow.
I watched this law operate at industrial scale for twenty years in programme management. Give a workstream six months and it will consume six months. Not through laziness, but through a hundred small expansions: extra review cycles, broader scope, polish nobody asked for, meetings about the meetings. The work inflates to fit its container. Always.
The usual advice, and why it half-works
The standard response is artificial deadlines: lie to yourself about when things are due, and let the false urgency compress the work. It does work, briefly, and then it stops, for the obvious reason. You know you set the deadline. The part of your brain that inflates work to fit containers is the same part that knows Thursday's "deadline" is really next Tuesday. Self-imposed urgency is a currency that hyperinflates: the more you use it, the less each deadline buys.
The subtler problem is that most people apply deadline pressure to individual tasks while leaving the container itself unbounded. Your task list stretches to the horizon; your planning covers whatever it covers; and inside that shapeless expanse, every task quietly inflates. The true container, the one Parkinson's Law actually reads, is not the task's deadline. It's your planning horizon.
My twist: don't fake deadlines. Shrink the container.
Here's what twenty years of watching the law operate taught me: you can't beat Parkinson's Law with willpower, but you can beat it with architecture. The law fills whatever container exists. So build a small container, permanently, and let the law work for you.
That means a fixed, short planning horizon. Not per-task fake deadlines you'll learn to ignore, but a structural boundary on how far ahead your working view extends. Inside a one-to-two-week window, work has nowhere to sprawl. "I'll get to it eventually" isn't an available thought, because eventually isn't on the screen. The container does the compressing, silently, without you having to manufacture urgency you know is fake. And unlike a self-imposed deadline, you can't call the container's bluff, because it isn't bluffing. Next week genuinely is the edge of the plan.
How ka-do handles it
The 1–2 week planning window is ka-do's core opinion, and Parkinson's Law is a large part of why it exists. List view or board view, your working world is at most two weeks deep. Everything further out lives in Unscheduled: stored, safe, but outside the container, where it can't quietly inflate into "in progress."
What ka-do won't do is fake urgency on top of that. No countdown timers, no artificial red alerts, no gamified streaks manufacturing pressure. The window is the entire intervention. A permanently small container, applied by the shape of the tool rather than enforced through nagging. You feel it as a strange lightness at first: the plan ends. That's not a limitation. That's the law, finally working in your favour.
And in the interest of the honesty this series runs on: the window constrains me too, and I still find work expanding to fill a fortnight that should have taken a week. The law doesn't switch off. It just operates inside a container two weeks wide instead of a year wide — and the difference in what escapes is everything.
The one-line version
Work fills the container. You can't stop the filling, so stop building big containers. It's the same principle underneath why timeboxing actually works when the tool is built for it.
Ka-do's planning window is two weeks, on purpose. Try it free →