The two-minute rule, and when it backfires

If a task will take less than about two minutes, do it now rather than writing it down. That is the whole rule. It appears in David Allen's Getting Things Done, where it sits inside a much larger system for deciding what to do with incoming things, and it has spread far beyond that system, usually with the surrounding structure stripped off.

It is worth being accurate about what kind of claim it is. Nobody has run a trial of the two-minute rule. The number is not an effect size and was not derived from measurement; it is a threshold Allen chose because it is roughly the point at which recording a task costs about as much as doing it. That is a reasonable piece of practical reasoning, and it is not evidence.

What it is actually for

Small tasks have an overhead. Writing down reply to Sam about the room booking takes time, then takes attention every time you read past it, then takes a decision later about whether now is the moment. For a task of thirty seconds, that overhead can exceed the task several times over.

Small items also accumulate. Fourteen of them on one list stop reading as fourteen small things and start reading as one undifferentiated obligation, and the obligation is what gets avoided. Nobody avoids sending a two-line email. People avoid the list with forty things on it, including that email.

The rule prevents the pile from forming, which is a real and worthwhile thing to prevent.

Where it pays and where it costs

It pays in a bounded set of arrivals with a clear edge — the emails from one morning, the papers on one desk, the forms in one folder. You go through them, the short ones vanish as you meet them, and what remains is the smaller number of real items you can actually see. The relief people report is not imaginary. Removing thirty trivial entries makes the four significant ones legible for the first time in a fortnight.

It costs where the stream of arrivals has no edge. In an inbox that refills continuously, or a shared channel, or a household with several people in it, two-minute jobs are produced faster than you can absorb them, and the rule quietly converts your whole day into responding. At six o'clock you have done perhaps forty things, every one of them genuinely necessary, and the piece of work you are actually assessed on has not been opened. The day felt busy and effective throughout, which is exactly why the pattern survives.

The three ways it goes wrong

As displacement. Two-minute tasks are the most attractive work available when something difficult is waiting. They complete. They produce a visible result. They are unambiguous, which the difficult thing is not. Applied without a boundary, the rule gives permission to spend a whole day doing them, and it does not feel like avoidance, because every individual item was legitimate.

As an underestimate. Almost nothing takes two minutes. The reply needs a figure you have to find; the form asks a question you have to check; the message opens a conversation. This is not carelessness. Buehler, Griffin and Ross found that people systematically underestimate how long their own tasks will take, and continue to do so even when reminded that they have made the same error before (Journal of Personality and Social Psychology, 1994). Their work is about projects rather than two-minute jobs, but the direction of the error is consistent, and the practical version is that your two-minute category is mostly full of five- and ten-minute jobs.

As a category error. The most damaging use is on something that was never small. A task gets classified as two minutes because its first visible action is two minutes — just send the email to ask — and then it turns out to be the front door of something large. Once you are inside it you are committed, and you did not decide to be.

What the setting can do

Much of the time, the two-minute rule is a private absorbent for an organisational failure. A steady stream of small requests arriving one at a time at one person is a coordination problem, and the person at the receiving end has been left to solve it with a personal heuristic.

The structural answers are cheap, and they are the sort of thing a team lead, a supervisor or a course administrator can decide.

• One route, not five. Where requests arrive by email, chat, a form and in the corridor, the same request often arrives twice and is answered twice. A single named route for a given kind of request removes that duplication for everyone at once.
• A stated response norm. Requests of this kind are answered within two working days converts a continuous obligation into a scheduled one. It is a sentence in a team charter or a module handbook, and it costs nothing.
• Batched arrivals. An office hour, a weekly slot, one digest instead of individual messages. Each turns an unbounded stream into a bounded one, which is exactly the condition under which the rule works properly.

If small requests land on you continuously and you are the only person doing anything about it, the honest description is that a design gap is being covered by your goodwill. That is a reasonable thing to say out loud where the work is allocated.

Small things to try this week

1. Keep the rule, but put a boundary on it: run it for a fixed period once or twice a day, and outside that period write things down as normal.
2. For three days, note the actual duration of anything you classify as two minutes. Do not adjust your behaviour, just record. The gap is usually the persuasive part.
3. Before starting a supposedly two-minute item, ask whether it has a second step. If it does, it belongs on the list rather than in your hands right now.

If the pile still forms after a few weeks of this, that is useful information. It generally means the volume arriving is the problem rather than the way you are handling it, and volume is not something a personal rule was ever going to solve.