Why Switching Tasks Costs You: Attention Residue Explained

Published Sep 22, 2026 · ~6 min read

Answering a message takes ten seconds. Then you return to what you were doing and stare at the screen for a while, unsure where you were. The ten seconds are the visible bill. The blank stretch is the real one.

The cost is not the interruption

Most people think of an interruption as a brief pause. You time it, it comes to ten seconds, so it seems free. What actually gets spent is the period after the interruption ends, while you try to get back into the original task.

That period is rough because your brain never really left. It suspended the task instead of closing it. A suspended task keeps consuming processing capacity, so you are sitting in the new task with part of your attention still parked on the old one. That leftover is what researchers call attention residue.

How attention residue was demonstrated

The term comes from an organisational behaviour study (Leroy, published in 2009). Participants moved between tasks, and cognitive performance dropped noticeably when the task they had just left was unfinished and had no plan attached to it. The open loop kept drawing on resources.

The more useful half of the finding is the comparison group. When participants were asked to write down what they would do next on the interrupted task before switching, the residue shrank substantially. What helped was a plan, not completion. The brain does not need the task to be finished. It needs to know the task is handled. That detail carries most of the practical value here.

Why glancing at your phone is so expensive

Messaging apps hit both conditions at once. First, they create fresh unfinished goals: you read the message, and most messages either need a reply or linger in your mind. Second, they usually leave you no way back in. You read it and close the app, with no note on what happens next. Both conditions together is the worst combination for residue.

So the problem with a quick check is not screen time. It is that you have queued a task into your cognitive backlog without handling it or scheduling it. Queued tasks do not disappear on their own.

As for the widely repeated claim that a single distraction costs 23 minutes of refocusing, it traces back to early observational fieldwork, and researchers have since questioned how stable that figure is. Treat it as a direction rather than a law. The recovery cost far exceeds the length of the interruption. How many minutes depends on the difficulty of the task and how often you switch.

What a switch actually consumes

Break it down and a switch charges you three times:

• Goal switching: unloading one set of rules and loading another. The more similar the tasks, the harder this is, because the interference is stronger. Writing a report while answering email is messier than writing a report while picking up a child.
• Rule reloading: returning requires rebuilding context. Where was I, why did I write it that way, what comes next. This does not restore instantly from memory. It has to be reconstructed.
• Mood and momentum: coming back from something easy to something hard costs extra activation energy. It is why the hardest move after scrolling is simply going back.

Frequent switching also creates an illusion of productivity. You feel busy, but the completed list stays short. Every switch leaves a small loose end, and by evening the backlog holds a row of half-finished items.

Why unfinished work keeps buzzing

This connects to the classic Zeigarnik effect: unfinished tasks are remembered better than finished ones. One explanation is that an open goal stays active to remind you to deal with it. Evolutionarily useful. In today's information environment, it is a noise source.

The key is that an open goal wants a decision, not a completion. A line on a to-do list or a simple note saying handle this at three o'clock can retire a goal from active status. This is why writing something down feels like relief even when nothing has been done yet.

Four things that work

First, batch your switches. Messages, email, and group chats rarely need instant replies. Move them to fixed windows, say once every 60 to 90 minutes. Cutting twenty switches to five saves more than the seconds themselves. It saves the rebuilding after each one.

Second, write one line before you leave. When an interruption arrives, leave a sentence in the document about what comes next, even five words. This is the lowest-cost, best-supported move in the whole research area.

Third, give the return trip a ritual. The more concrete the re-entry, the better: reread the paragraph you just wrote, put the cursor on the break point, or spend two minutes on something mechanical to let attention settle back. Small micro-break style actions help here, not as a reward but as a boundary marker that the previous block is over.

Fourth, accept that switching happens, and aim only to stop it spreading. What wrecks a day is not the interruption itself but the ten minutes of scrolling that follows it. Getting good at returning is far more realistic than chasing a day with no interruptions at all.

A comparison you can run today

Pick a half day with a clear deliverable and one rule: check messages only at three preset times. At the end of the afternoon, ask two questions. What did I actually finish, and when was the hardest ten minutes?

Most people find the difference in output larger than they expected on the first attempt. If the mechanics of attention interest you further, continue with the section on anticipation in the reaction time test, or browse the article index for the rest of the series.