The two-minute rule, and where it stops working

Everything below about the two-minute rule comes from what actually happens rather than from what is supposed to.

Deferred small tasks are expensive

A task that takes ninety seconds costs far more than that once it is remembered, written down, prioritised and returned to. It also occupies a small amount of background attention for as long as it remains outstanding.

Doing it at the moment it appears removes all of that overhead at once. The cost compounds with the number outstanding as well, since a list of forty small things is harder to face than forty separate ninety-second jobs ever were.

The threshold is arbitrary and useful

Two minutes is not a magic number; it is short enough that the interruption costs less than the tracking would. Some people find five minutes works better at home and two at a desk. The value is in having a threshold at all rather than in which one you pick.

The useful part is this: it should also fall as the work gets more demanding, because two minutes taken out of deep concentration costs considerably more than two minutes taken out of email.

Where it turns into avoidance

A day spent clearing small tasks can feel productive while the significant piece of work remains untouched. Small tasks are attractive precisely because they are finishable, which is what makes them a good hiding place. Noticing that you have cleared forty small things and started nothing large is the diagnostic.

The tell is a sudden appetite for tidying or admin arriving exactly when the difficult thing became due, which is a consistent enough pattern to use as a warning.

Protect the large thing first

Doing the significant work first, before the small tasks are visible, is the standard correction. That usually means before email, before the list and before the messages. The small tasks will still be there and will take the same ninety seconds each.

Starting is the part that needs protecting rather than finishing, because a large task that has been opened tends to pull you back to it while an unstarted one does not.

Batching is the alternative

Where immediate action is disruptive — in the middle of concentrated work — collecting small tasks and clearing them in one block works. The requirement is that the block actually happens, or the collection becomes a backlog.

A fixed daily slot for admin is what makes batching viable rather than aspirational. Keeping the collected items in one place rather than four is what stops the batch becoming its own search, and paper does this as well as anything.

Adjust the size of it until it is something you would actually do tired.

Some of them should not be done at all

A two-minute task performed daily is close to twelve hours a year, which is usually enough to justify automating, delegating or eliminating it rather than doing it efficiently. The rule makes small jobs so cheap to complete that it quietly removes the pressure to ask whether they were worth doing. Tasks that exist because of how something is configured — an alert that always needs clearing, a form that always needs the same correction — are better fixed once at the source than done quickly forever.

In practice, estimation is also unreliable: plenty of jobs that look like two minutes turn out to be twenty once opened, which is why the rule works best on things you have already done before.

The takeaway

Do the small things immediately, and only after the large thing has started.

The version you keep doing is the version that works.