The task you just left is still running

The cost of switching is not paid at the moment you switch. It is paid afterwards, quietly, in the work you thought you had moved on to.

The UnaTask team, Building UnaTask. Aug 21, 2026. 5 min read.

You close the document, walk into the meeting, and you are there. Physically. But some part of you is still holding the paragraph you were in the middle of, and it stays there for the next forty minutes while you nod at things.

That feeling has a name and a measurement, and the measurement is worse than the feeling suggests.

What Leroy found

In 2009 Sophie Leroy published a set of experiments in Organizational Behavior and Human Decision Processes on what she named attention residue. People moved from one task to another under controlled conditions, and their performance on the second task was measured.

It was worse. Less accurate, slower, shallower. That much you might have guessed. The part worth knowing is what did not happen: it did not recover. There was no brief settling in period after which people were back to normal. The degradation persisted through the second task.

You do not arrive at the next task. You arrive with the last one still running in the background, and you do the new work with what is left.

The condition that makes it worse

Residue is not uniform. It is strongest when the task you left was unfinished, under time pressure, or emotionally charged.

Sit with that list for a second, because it is a fairly precise description of an ordinary working day. Observed office workers spend about eleven minutes on a thread of work before switching or being interrupted, and more than half of those threads get cut off before they are done. Unfinished, under pressure, and often the thing you care about. That is not the exception. That is the material.

Which changes what a block is for

The obvious lesson is protect the hour, and that is true but incomplete. The less obvious one, and the useful one, is that how the hour ends matters as much as whether you got it.

A block that closes at a natural stopping point leaves very little behind. A block that gets cut off mid sentence by a timer or a calendar alert leaves the maximum amount behind, and you carry it into the next thing and pay for it there. Two blocks of the same length can cost completely different amounts depending on how they finished.

Three things that follow

End on a landing, not a cliff. Give the last five minutes of any block to closing: finish the paragraph, commit the change, write the one line that tells tomorrow where you were. It costs five minutes and saves more than that from the next block. In practice this means setting a ninety minute block and treating the last five as the landing, rather than setting eighty five and pretending.
When you truly cannot finish, write the residue down. The open question, the next step, the thing you were about to try. One sentence, on the task itself rather than in a separate note, so it is waiting where you will land next time. Externalising it does more than help you remember, for reasons that turn out to be well studied.
Put the transition inside the block. Back to back blocks guarantee residue, because the switch happens with no room to close anything. Book blocks fifty or eighty minutes long on the hour, not sixty or ninety. The ten minute gap is not slack. It is the mechanism working.

None of this asks you to be interrupted less, which is mostly not in your gift. It asks you to change what an interruption costs, which is.

This is one chapter of Blocks No. 01, a 26 page guide to time blocking. Free PDF, no email, and every number in it traced to the paper it came from. Get the full guide