Switch from one task to another and you don’t arrive whole. A piece of your attention stays behind on the thing you just left, replaying it, second-guessing it, wondering if you missed something. Researchers call this attention residue, and its most uncomfortable finding is that the residue doesn’t come from the task you’re moving to. It comes from the one you didn’t finish.
That last word matters. A task you’ve closed cleanly lets go of you. A task you abandoned mid-thought keeps a hand on your sleeve. The more unfinished and unresolved it was, the more of your head it keeps, so you sit down to the new thing at a fraction of your real capacity, and you feel vaguely worse at it than you should, without quite knowing why.
Products create this state constantly, usually by accident. Six half-completed setup steps. A dashboard with four things blinking for attention at once. A notification that yanks someone out of a flow they were three steps into and drops them somewhere else. Every one of those is a task the user couldn’t close, so every one of them is a shard of attention that doesn’t make it to the next screen.
The instinct, especially for a builder proud of everything the product can do, is to let people juggle. Keep it all open, all live, all one click away. It looks like power. It reads to the brain as a room full of unfinished tasks, each quietly taxing the one the user is actually trying to do.
So the design decision is to protect a single thread at a time, and to give tasks real endings. Concretely:
Let the current task finish before you offer the next one. A flow that says “done, here’s what you did” and then surfaces the next step beats a flow that shows step two while step one is still ambiguous. The little moment of closure isn’t a nicety; it’s what empties the residue so the next screen gets a full brain.
When you must interrupt someone, a required verification, an outage, an expiring session, give them a way to park, not just abandon. The same research line that found the residue also found what shrinks it: a quick “here’s where I was, here’s what’s next” plan before the switch. A one-line “you were setting up X; pick up at step 3” does for your user what that plan did in the lab. It lets the brain set the task down instead of carrying it.
And be suspicious of surfaces that show many open loops at once, the six-badge dashboard, the endless “finish setting up your account” checklist that never fully clears. Each unfinished item is not a helpful nudge. It’s a small tax on everything else the person tries to do while it sits there.
Depth is worth building. But depth offered as breadth, everything open, everything half-done, nothing closable, doesn’t feel like capability from the inside. It feels like a desk covered in sticky notes, each one saying don’t forget me, while you try to do the one thing you sat down to do. Give people one thread and a clean way to end it, and you hand back the attention your interface was quietly keeping.
The science, to look up: attention residue (Sophie Leroy, 2009, “Why is it so hard to do my work?”); the “ready-to-resume” intervention (Leroy & Glomb, 2018); related: the Zeigarnik effect and open-loop memory. Well established.
Sources
- The science, to look up: attention residue (Sophie Leroy, 2009, 'Why is it so hard to do my work?')
- the 'ready-to-resume' intervention (Leroy & Glomb, 2018)
- related: the Zeigarnik effect and open-loop memory.
Liked this? Get the next one in Working Theory.
Going weekly in August (it's in beta now). One genuinely interesting read on building, the brain, and the science most people missed.
Subscribe →