Attention residue

You leave a meeting and sit down to work. Nothing about the room follows you — the meeting is over, the file is open, you have an hour. And for fifteen minutes of that hour you are not really there. Some part of you is still drafting the reply you did not get to make.

Sophie Leroy named this in 2009: attention residue. Her finding was not that switching is slow. It was that when you move from one task to another, a portion of your attention stays on the first one, and it stays in proportion to how unfinished the first one felt.

Which reframes the whole complaint. The problem with a fragmented day is not the fragments. It is that each fragment leaves something behind, and by the afternoon you are trying to think with an attention that is distributed across six places you are no longer standing.

Leroy's design was simple and the result is the reason the term stuck. Participants worked on a task, were interrupted, and moved to a second unrelated task. What varied was whether the first task felt complete.

Time pressure on the first task made the residue worse, not better. People who were interrupted while under deadline pressure — the condition where you would most want them to move on cleanly — carried the most residue into the next task and performed worst on it.

Finishing was not the operative variable. Closure was. In her follow-up work with Theresa Glomb, the thing that predicted residue was not whether the task was actually done but whether the person had a settled sense of where they were with it.

That distinction produced the one intervention in this literature with good evidence behind it. Before switching, spend a minute writing down where you got to and what the next move is — Leroy and Glomb call it a ready-to-resume plan. In their studies it measurably cut residue and improved performance on the interrupting task. Not a productivity slogan; an experimental result with a mechanism. The unfinished thing stops asking for attention once it has been written down somewhere it can be found.

Residue sits on top of a much older finding from cognitive psychology, and the two are often confused.

Switch costs are the reliable slowdown when you alternate between two tasks rather than repeating one. Rubinstein, Meyer and Evans laid out in 2001 why: switching requires reconfiguring the rules you are operating under, and reconfiguration takes time even when you know the switch is coming. Monsell's 2003 review of the field is the standard summary, and its most useful point is that the cost does not disappear with practice or preparation. It shrinks. It never goes to zero.

Switch costs are measured in hundreds of milliseconds. Attention residue is measured in minutes of degraded work. They compound, but they are different sizes of problem, and only the second one is worth reorganising your week around.

You will have seen it: it takes twenty-three minutes and fifteen seconds to refocus after an interruption. It is quoted everywhere, usually as though a stopwatch were run on someone's concentration.

It comes from Gloria Mark's 2008 observational study of office workers, and what it actually measured is different. Her team shadowed people and timed how long it took them to return to an interrupted task — during which interval the person was not staring into space, they were doing other work, often several other things. The figure is a description of how work queues in a busy office. It is not a recovery time for attention, and it does not mean twenty-three minutes are lost.

The genuinely interesting result from that same study is the one nobody quotes. Interrupted people did not work slower. They worked faster — and reported higher stress, more frustration and more time pressure. The cost of a fragmented day is not primarily throughput. It is that you pay for the throughput with something else.

We flag this because the inflated number is the kind of thing that makes a true claim easier to dismiss. Attention residue is real without it.

Batch by kind, not by size. Two tasks that use the same rules cost little to alternate between. Two that use different rules cost a great deal, no matter how short they are. The expensive day is not the busy one, it is the varied one.

Write the handover before you switch. One sentence, where you are, what is next. This is the tested one and it takes less than a minute.

Protect the transition, not just the block. People defend two hours of focus time and then walk into it straight out of a call. The first twenty minutes of the block are already spent. A two-minute gap in between buys back more than a longer block does.

Notice which residue is not about tasks. Sometimes the thing occupying the back of your attention is not an unfinished document, it is an unfinished conversation — and that runs on a different mechanism and does not respond to a to-do list. The entry on rumination is that one, and tired but wired is what it looks like when it follows you past midnight.

And if the residue is total rather than partial — if focus does not arrive even on a clear, uninterrupted day, and always has not — that is a different question from this one. The ADHD Traits test reads that pattern as a screening-style profile rather than a diagnosis, and a persistent lifelong version of it belongs with a clinician rather than a calendar.