Attention residue
Part of attention stays on the previous task after switching, and degrades the next one.
When a person switches tasks, some attention remains on the prior task, particularly when it was left unfinished or under time pressure. Performance on the new task suffers until the residue clears. Interruption research shows the same pattern: interrupted work gets finished, but at a cost in speed and stress.
How it shows up in software
Notifications, badges, modal upsells and forced context switches all impose this cost on the user's real work. It also shapes what a product should do when someone returns: an app that reopens where the person was clears residue, while an app that reopens on a promo screen adds a second switch on top of the first.
Using it well
- Batch non-urgent notifications and deliver them at natural boundaries instead of on arrival.
- Restore exact state on return: scroll position, draft contents, expanded panels, selected filters.
- Let a person close a loop in one gesture from where the notification appears, so they do not have to enter the other context at all.
- Rank interruptions honestly. If everything can interrupt, nothing gets the focus that matters.
Where it turns manipulative
- Deliberately fragmenting attention with engagement notifications, on the logic that a distracted user opens the app more often.
- Interrupting a task in progress with an upsell precisely because the person is engaged and therefore reachable.
- Dropping a returning user onto a marketing screen instead of the work they left, which adds a switch to a switch.
Where you have seen it
iOS
Focus modes and scheduled notification summaries hold non-urgent alerts for delivery at set times.
Slack
Threads keep a side conversation out of the main channel, and the app returns to the last read position per channel.
Gmail
Replying from a notification on mobile completes the task without entering the inbox.
What the research says
- Leroy, 2009Mixed evidence
Participants switching from an unfinished task performed worse on the next task than those who had completed the first, and the deficit persisted through the following task.
Laboratory switching tasks with modest samples, from a literature that has faced replication scrutiny.
- Mark, Gudith and Klocke, 2008Mixed evidence
Interrupted work was completed in the same or less total time, but with higher reported stress, frustration and effort.
- Leroy and Glomb, 2018Mixed evidence
Field and laboratory studies found residue from time-pressured unfinished work predicted worse performance on subsequent tasks.
Grades are a judgement about the evidence, not about how useful the idea is. Plenty of contested effects are still worth knowing, as long as you do not cite them as settled.