There's a book called The 12 Week Year with a genuinely good idea in it: annual planning is broken, so compress it. Treat twelve weeks as your "year" (plan it, execute it, review it) and the deadline stops being an abstraction eleven months away. The urgency that normally arrives in December arrives every quarter instead. No mid-year complacency, no "we'll catch up in Q4."
I agree with every part of the diagnosis. Annualised thinking is where goals go to hibernate; anyone who's watched a January plan meet a September reality knows the year is a fiction we plan in and a mess we live in. Compression is the right instinct.
My problem is that twelve weeks is still long enough to hide in.
The hiding-space test
Here's the question to ask of any planning horizon: can I comfortably defer something inside it? If yes, the horizon has hiding space, and Parkinson's Law will find it. A twelve-week container is smaller than a year, but it's still roomy. Week two of a twelve-week cycle feels exactly like February used to. There's loads of runway. The reckoning is ten weeks off. The plan is compressed; the procrastination has simply been rescaled to fit.
You can see this in how 12-Week-Year practitioners actually behave: the system needs weekly scorecards and accountability meetings bolted on precisely because twelve weeks doesn't self-enforce. The compression created urgency at the cycle level and none at the week level, so the method has to manufacture weekly pressure with process. That's not a flaw in the execution. It's the horizon telling you it's still too long.
My twist: compress until you can't hide, then stop
Follow the logic of the book to its honest conclusion. If compressing a year to twelve weeks kills annual complacency, what kills twelve-week complacency? Compress again. Where do you stop? At the point where the horizon has no hiding space left but still has room to plan. For actual working life, that's one to two weeks.
One to two weeks is the shortest horizon where "later this week" is still a real plan rather than a dodge, and the longest one where nothing can comfortably hibernate. It doesn't need scorecards bolted on, because the deadline pressure is ambient: the edge of the plan is always visible. And it matches how work actually arrives. Nobody's real week respects a quarterly theme, but almost everything genuinely respects a fortnight.
Keep the longer cycles for what they're good at — direction, goals, the why — but stop pretending they're planning horizons. They're aiming horizons. The plan itself lives where the hiding space doesn't.
How ka-do handles it
This is ka-do's entire architecture stated as a planning philosophy. The working view is a one-to-two-week window. Not configurable up to a quarter, not extendable to a roadmap. A window. Longer-range material lives in Unscheduled as direction waiting for its fortnight, and the app never pretends that a task parked eight weeks out is "planned." It's stored. Planned means it's in the window.
What ka-do won't do is run your goal cadence for you. No OKR machinery, no quarterly cycle features, no scorecard rituals. Have those wherever you like. The app holds the one opinion the whole compression argument lands on: the plan is this week and next, and everything else is honestly labelled as not-yet.
Full disclosure, as ever: I aim in quarters and plan in fortnights, and the aiming is the wobblier of the two. But I've never once opened a two-week window and found something hibernating in it. There's nowhere for it to sleep.
The one-line version
Compressing the year to twelve weeks was the right idea stopped early. Keep compressing until the hiding space is gone. That's a fortnight. Compress it one step further and you land on three tasks, today — the same instinct, just carried all the way down.
Ka-do plans in the only window nothing can hide in. Try it free →