Focus

Focus is a state, not a trait

"I have terrible focus" describes a person. "I couldn't focus in an open-plan office on four hours of sleep with a task I didn't understand" describes conditions. The second one is more accurate almost every time, and it has the advantage of pointing at something changeable.

Three conditions do most of the work. You need enough baseline capacity to attend at all, which is mostly a question of sleep, food, and stress rather than technique. You need a clear enough target that attention has something to grip. And you need an environment that isn't continuously offering alternatives.

When focus fails, one of those three is usually missing, and it's worth finding out which before reaching for a method. Adding technique to a sleep problem produces frustration and a conclusion about your character that isn't warranted.

Attention residue

Sophie Leroy's 2009 research identified a phenomenon that explains a lot of unproductive afternoons.^[1] When you switch from one task to another, part of your attention stays behind on the first one. She called it attention residue, and found it's strongest when the first task was left unfinished.

The task you've switched to gets a degraded version of your attention. You generally can't tell. Subjectively you feel present on the new thing; measurably you're carrying the old one.

"I'll just quickly check" costs ten seconds and then taxes the next forty minutes.

This is the mechanism that makes quick interruptions so much more expensive than they appear. The interruption's duration is the visible cost, and it's the small one. The residue on everything after it is the real bill, and it never itemizes itself.

It also suggests something practical about how to leave a task. An unfinished task leaves more residue than a finished one, and a task left at a defined stopping point — with a note about what comes next — leaves less than one abandoned mid-thought. If you have to stop, spend fifteen seconds closing the loop before you go.

What context switching costs

Gloria Mark and colleagues studied how work actually gets fragmented in real offices.^[2] The picture is not one of long uninterrupted stretches punctuated by occasional interruptions. It's continuous fragmentation, with people frequently working several intervening tasks before returning to what they were pulled away from.

A related study produced a result that keeps surprising people: interrupted work was often completed faster than uninterrupted work.^[3] People compensated by working more intensely. What they reported was significantly higher stress, frustration, time pressure, and effort.

That's the finding worth sitting with. The cost of a fragmented day frequently doesn't appear in output at all. It appears in how depleted you are by six o'clock — which is precisely the kind of cost people attribute to their own inadequacy rather than to the conditions that produced it.

The multitasking illusion

For anything that genuinely requires attention, multitasking isn't a thing you're doing. Rapid switching is. Each switch carries a cost in time and accuracy, and the costs accumulate quietly.

Ophir, Nass, and Wagner tested whether people who multitask heavily with media get better at it.^[4] They compared heavy and light media multitaskers on filtering irrelevant information, managing working memory, and switching between tasks. Heavy multitaskers performed worse across the board.

The likely reading is that habitual switching trains a broader attentional style — one that's sensitive to everything in the environment, which is useful for noticing and costly for sustaining. Practice made people better at being interrupted and worse at not being.

Flow, and what it needs

Mihaly Csikszentmihalyi's research on flow described the absorbed state where action and awareness merge, self-consciousness drops away, and time stops feeling normal.^[5] It's the state people mean when they say they want to focus better.

You can't summon it. You can arrange the conditions it depends on, and the conditions are fairly specific.

A clear goal. Not "work on the report." Something specific enough that you'd know if you'd done it. Ambiguity is the most reliable flow-killer, because attention can't grip something with no edges.

Immediate feedback. Some signal that you're making progress. Writers get sentences; coders get output. Work that gives no signal for hours is genuinely harder to stay in, which is an argument for building intermediate checkpoints rather than for trying harder.

Challenge matched to skill. Too easy and attention wanders. Too hard and you get anxiety, which is its own attention thief. The band is narrower than people expect, and the fix is usually adjusting the task rather than adjusting yourself.

Note how much of that is about task definition rather than concentration. Most flow failures are specification failures wearing a focus costume.

Building a session that holds

A practical sequence, ordered by what fails first.

1. Decide the one thing before you sit down. Deciding while sitting there is how a session becomes twenty minutes of tab-opening. Ideally this was decided earlier — see attention management.
2. Make it concrete enough to finish. If you can't say what "done" looks like, you've got a project rather than a task. Cut a piece off it that has an edge.
3. Remove the triggers before you start. Notifications off, phone in another room. Doing this after you've started means starting twice.
4. Set an end, not just a start. A defined length makes the session safe to begin. Open-ended focus is intimidating in a way that thirty minutes isn't.
5. Park interruptions rather than following them. When something surfaces mid-session, write it down and keep going. The note is what lets you drop it; without one, working memory keeps holding it and you'll switch eventually.
6. Close the loop when you stop. Ten seconds on where you are and what's next reduces residue and makes returning cheaper.

When you can't focus at all

Sometimes none of this applies because you're not choosing between tasks — you're sitting in front of the right task, entirely unable to start.

That's a different problem with a different fix. It's usually activation rather than focus, and it comes in three varieties: you don't actually know what the task is, there's an unnamed emotion attached to it, or you can't see the first physical move. How to start a task when your brain won't works through all three.

If it happens most days rather than occasionally, the underlying machinery is worth understanding directly. See executive function.

Frequently asked questions

What is attention residue?

The part of your attention that stays stuck on a previous task after you switch away. It's strongest when the first task was left unfinished, and it means the task you switch to gets a degraded version of your attention that you usually can't feel.

How long does it take to refocus after an interruption?

Longer than the interruption, and it varies by task and person. The more useful framing is that the cost isn't the interruption's duration — it's the degraded quality of the work that follows it.

Is multitasking real?

Not for tasks that both require attention. What feels like multitasking is rapid switching, and each switch costs time and accuracy. Heavy media multitaskers performed worse at filtering and switching than lighter ones, not better.

What is flow and can you make it happen?

Flow is the absorbed state where action and awareness merge. You can't summon it, but you can build the conditions: a clear goal, immediate feedback, and a challenge roughly matched to your skill. Ambiguity about the task reliably prevents it.

How long should a focus session be?

Short enough that starting feels safe. Twenty-five minutes is a reasonable default, but if you're avoiding a task, that commitment can itself be the barrier. Five or ten minutes with permission to stop usually gets you moving, and momentum extends it.