David Allen's two-minute rule is the most quoted line in Getting Things Done, and deservedly: if a task will take less than two minutes, do it now. The logic is clean. Filing, tracking, and later-remembering a tiny task costs more than the task itself. Reply to the email. Confirm the meeting. Done, gone, no overhead.
I've used it for years and it earns its reputation. But it has two failure modes that nobody quotes alongside it, and both of them live at minute three.
The first minute three: the task that lied
The rule depends entirely on your estimate, and human estimates are optimistic (the planning fallacy runs this whole series). "Quick reply" turns out to need checking a document, which needs finding the document, which surfaces a question, and suddenly your two-minute task is at minute nine with no end in sight. You didn't do a quick task. You got pulled into an unquick one by a friendly-sounding estimate, mid-whatever-you-were-actually-doing.
Do that four times a morning and the rule has quietly become a licence for interruption. Each invocation felt efficient. The morning, somehow, is gone.
The second minute three: the residue
The subtler cost hits even when the estimate was honest. A genuinely two-minute task, done mid-focus, doesn't cost two minutes. It costs two minutes plus the switch: attention residue on the way out, and by some well-worn research, the better part of twenty minutes fully finding your way back in. The rule prices the task. It doesn't price the interruption, and the interruption is usually the bigger number.
So the fine print reads: the two-minute rule is excellent between things and expensive during them. Allen's original context was processing an inbox, a between-things activity, where the rule is exactly right. Quoted without that context, it becomes advice to interrupt yourself efficiently.
My twist: the rule needs an escape hatch, and the hatch is two seconds
Here's the version I actually run. The two-minute rule holds, with one condition attached: it only fires when doing-it-now is cheaper than capturing it, and that comparison only stays honest if capture costs almost nothing. Two seconds, no decisions, any form.
With a two-second capture available, the choice at the moment a small task appears becomes clean. Between things, and it's genuinely tiny? Do it now; Allen was right. Mid-focus, or any doubt about the estimate? Two seconds of capture, and it's safely out of your head without costing the twenty-minute return trip. The rule stops being a licence for interruption and goes back to being what it was: a way to keep tiny things from generating tracking overhead. The escape hatch is what makes the rule safe to hold loaded.
Without cheap capture, notice, there is no real choice. "I'll just do it now" wins by default, because the alternative (hold it in your head and hope) is worse than any interruption. Most people don't over-apply the two-minute rule out of enthusiasm. They over-apply it because they don't trust anywhere to put the task instead.
How ka-do handles it
Ka-do's job here is the hatch. Capture is two seconds in any shape (why that's the whole design), so "do it now or dump it" is always a live choice, even mid-flow, even in a meeting. The dumped task lands with today or tomorrow attached and resurfaces on its own, which is the part that makes your brain accept the trade: the capture has to be as final as the doing, or the loop stays open anyway.
What ka-do won't do is nag you toward either side. No "you have 4 quick tasks!" prompts engineered to interrupt you efficiently. The choice stays yours; the tool just makes both options cost what they claim to.
Honesty clause: I still get caught by the lying two-minute task about once a week, usually an email. The difference is the second one gets dumped, because the hatch is there and the morning's already paid once.
The one-line version
The two-minute rule is right about the task and silent about the interruption. Keep it, and give it the condition Allen assumed: do it now only when now is between things; otherwise, two seconds of capture beats twenty minutes of getting back. That same honesty about what you're actually avoiding is the whole case for structured procrastination, done properly instead of guiltily.
Ka-do gives every small task a two-second exit. Try it free →