When you leave task A unfinished to switch to task B, part of your attention remains “stuck” on A. This cognitive residue isn’t metaphorical — it actively consumes processing capacity while you’re trying to work on B.
The mechanism
The brain maintains a mental buffer of the interrupted task. This buffer generates involuntary cognitive intrusions: thoughts about A surface while you’re working on B. These intrusions are often subconscious, but they consume working memory.
Result: B gets done with less cognitive capacity than it needs.
The hidden cost of interruptions
| Situation | Direct cost | Attention Residue cost |
|---|---|---|
| Resolved interruption | Time lost | Low — A is “closed” |
| Unresolved interruption | Time lost | High — A stays open in the buffer |
| Manager→Maker switch | Transition | Maximum — opposite cognitive contexts |
Leroy’s finding: subjects switching from an incomplete task to a new one scored significantly lower on the new task compared to those who finished their previous task first.
The fix: close before you leave
Writing “next action = …” before any interruption partially treats the task as “cleanly paused” in working memory — rather than “abandoned.” The brain interprets the note as a temporary closure.
This is the exact mechanism behind the Smart Resume feature in the Time Tracker: forcing the write of a resumption point before any interruption reduces attention residue in the next session.
Why “short” interruptions are deceptive
A 2-minute interruption doesn’t cost 2 minutes. It costs:
- 2 min for the interruption itself
-
- however long attention residue from the previous task degrades your performance
-
- the context switching ramp-up to return to the original task
This is why a day fragmented by 10 small interruptions can produce less than a continuous half-day — even if the declared “work time” is identical.
Sources
- Leroy, S. (2009). Why is it so hard to do my work? The challenge of attention residue when switching between work tasks. Organizational Behavior and Human Decision Processes, 109(2), 168-181.