How Much Time Do You Actually Lose Switching Between Projects on a Mac?
TL;DR
Gloria Mark's UC Irvine research measured an average of 23 minutes and 15 seconds to fully resume a task after an interruption, with roughly two other tasks touched in between. On a Mac that keeps every active project's apps, tabs, and windows open at once, you're not avoiding that cost by "getting better at multitasking" - you're re-triggering it every time you glance at the wrong window. The fix that actually removes the trigger is a separate workspace per project, so switching is a single action rather than a manual cleanup, which is what Ikuna does.
What the 23-minute number actually measured
The number comes from a specific 2008 study: Gloria Mark, Daniela Gudith, and Ulrich Klocke tracked office workers' interruptions across a full week and timed how long it took them to return to an interrupted task (Mark, Gudith & Klocke, 2008, CHI Proceedings). The average recovery time was 23 minutes and 15 seconds, and workers touched roughly two other tasks before getting back to the original one. That's a real, directly measured finding about task resumption in an office setting - not a general claim about multitasking ability, and not specific to any particular kind of interruption.
A separate line of research looked at who's actually worse at handling interruption and switching in the first place. Ophir, Nass, and Wagner's Stanford study compared self-identified heavy and light "media multitaskers" and found heavy multitaskers were more susceptible to interference from irrelevant environmental stimuli and irrelevant information already in memory (Ophir, Nass & Wagner, 2009, PNAS). Worth noting: that finding has had a mixed replication record since - later studies didn't consistently reproduce the original switch-cost effect, and some found the opposite pattern in certain tasks (Cyberpsychology meta-analysis, 2020). Use it as evidence that heavy multitasking habits can correlate with worse filtering, not as a settled law.
Why a shared Mac desktop keeps re-triggering the cost
Both studies are about interruptions and switches that already happened. The practical question for someone managing multiple projects on one Mac is upstream of that: how many potential triggers are sitting on your screen at any given moment? If Client A's Slack, Client B's Figma tab, and your side project's terminal are all open while you're working in one of them, every one of them is a candidate for the next unplanned switch - and each switch that happens re-runs the 23-minute-plus recovery cost measured by Mark's study.
What a shared, unmanaged desktop actually costs you
- Every open unrelated app is a standing decision point. You don't have to click it to pay part of the cost - noticing it and choosing not to click is itself a small tax.
- Manual cleanup between projects is its own interruption. Closing twelve tabs and quitting three apps before opening the next project's set is a task-switch in itself, with its own resumption cost for whatever you were doing before you started cleaning up.
- Because cleanup is annoying, most people skip it. The result is a permanently cluttered desktop that recreates the original problem instead of solving it.
Removing the trigger instead of managing around it
Batching similar tasks and asking "can this wait 25 minutes" are reasonable coping strategies for interruptions you can't prevent - a colleague's question, an unplanned call. They don't address self-inflicted switching caused by simply having every project's apps and tabs visible at once, because the trigger there isn't a person asking something, it's your own screen.
Ikuna addresses that specific trigger. A context defines the apps, browser tabs, and window layout for one project; switching into it quits or hides everything that isn't part of it, and switching back restores your exact prior state. Instead of a manual, error-prone cleanup between projects, the switch is a single action - so the environment stops being a source of unplanned interruptions in the first place, rather than something you have to resist every few minutes.
Comparison: coping with interruptions vs. removing the trigger
| Approach | What it addresses | What it leaves untouched |
|---|---|---|
| Batching tasks / the "25-minute rule" | External interruptions you can defer | Self-inflicted switching from open apps/tabs you can see right now |
| Manual desktop cleanup between projects | Removes clutter, eventually | Costs time and focus itself; usually skipped under pressure |
| macOS Spaces / Stage Manager | Groups or separates windows already open | Doesn't quit apps or restore a saved multi-app state |
| Ikuna | Removes/hides unrelated apps and tabs per project, restores prior state on return | Doesn't defend against externally-initiated interruptions (calls, questions) |
FAQ
Does the 23-minute figure apply to every kind of switch equally?
The original study measured average resumption time across a range of office interruptions during a normal workweek; it wasn't broken out by switch type in a way that lets you assume a quick tab-check costs exactly the same as a 45-minute meeting interruption. Treat it as a representative average, not a per-switch guarantee.
Is multitasking really impossible, or just less efficient?
For genuinely complex cognitive tasks, what looks like multitasking is almost always rapid serial switching rather than true parallel processing - the Stanford research on heavy multitaskers is about how well people filter and manage that switching, not about whether simultaneous deep cognition is possible.
Will removing open apps and tabs eliminate context-switching costs entirely?
No - it removes the self-inflicted, environment-driven switches. Externally initiated interruptions (a real emergency, a colleague at your desk) still carry a resumption cost; environment design reduces one major source, not all of them.
How is this different from just closing tabs manually every time I switch projects?
The mechanism is the same - remove what doesn't belong - but manual cleanup has to be redone from memory every time and is easy to skip when you're busy. A defined context makes the removal (and the restoration) a single repeatable action instead of an ad hoc chore.
Sources
- Mark, G., Gudith, D., & Klocke, U. (2008). The Cost of Interrupted Work: More Speed and Stress. CHI 2008 Proceedings, UC Irvine. (Directly tested: office-worker interruptions and task-resumption time.)
- Ophir, E., Nass, C., & Wagner, A. D. (2009). Cognitive Control in Media Multitaskers. Proceedings of the National Academy of Sciences, 106(37), 15583-15587. (Directly tested: self-reported heavy vs. light media multitaskers on filtering and task-switching tasks; later replication record is mixed.)
- Cyberpsychology: Journal of Psychosocial Research on Cyberspace (2020). "Cognitive Control in Media Multitaskers" Ten Years On: A Meta-Analysis. (Directly tested: aggregate replication attempts of the 2009 finding.)
