Task switching is the bill your todo app doesn’t itemize

Tuesday, 2:14pm. I finished a task. A real one, with a satisfying close. I opened my todo app to start the next thing.

At 2:25pm, I was still in the app. I had not started anything. I had read seven different tasks. I had moved one to tomorrow. I had renamed another. I had opened a project I hadn’t touched in a month, felt a small wave of dread, and closed it. Eleven minutes had gone, and the only thing I’d actually done was switch tabs.

This is the part of ADHD productivity that nobody itemizes on the bill. Not the doing. Not even the starting. The switching. The walk between two tasks, which in any well-designed app should be trivial, and which in every list-based app I’ve used is the single most expensive thing I do all day.

The general cost of task switching, before we even get to ADHD

There’s a number that gets quoted a lot: 23 minutes and 15 seconds. It comes from Gloria Mark’s 2008 study at UC Irvine, which followed knowledge workers in the field and measured how long it took them to fully resume a task after an interruption. The same line of research found that people switched between “working spheres” on average every 10 minutes and 29 seconds, and in a controlled email-task experiment, every 3 minutes and 5 seconds. The 23-minute figure is the recovery, not the switch itself; the switch is roughly free in attention, and the resumption is where the bill lands.

The mechanism behind that bill has a name. Sophie Leroy’s 2009 paper called it attention residue: when you transition from Task A to Task B, part of your attention stays behind with A, particularly if A was unfinished or unresolved. You start B with a fraction of yourself. The fraction is small enough to ignore on any single switch and large enough, across a day of switches, to be the dominant cost of knowledge work.

Both numbers are from neurotypical participants in normal office conditions. They’re the baseline bill.

What ADHD does to the bill

The ADHD literature on task switching is more specific than the popular “easily distracted” framing suggests. A 2011 study on adults with ADHD found a selective impairment in attentional set shifting, the specific executive-function operation of suppressing one task rule and amplifying another. Earlier work on children with ADHD found substantially larger switch costs than non-ADHD controls, and that those costs collapsed back to baseline on stimulant medication. The deficit isn’t generic distractibility. It’s a specific tax paid every time the brain has to flip its rule set from “do A” to “do B.”

The popular phrase for the felt sense of this is “context-switching cost.” Inside the head, it isn’t the friction of changing. It’s the friction of deciding what to change to, while the previous task is still running in the background, and three or four other tasks are visually shouting from the screen.

The app you use is the thing standing between you and the next task. If it makes the deciding cheap, the bill stays small. If it makes you re-triage the entire list at every switch, the bill is enormous, and you pay it dozens of times a day.

What the list view actually does at the moment of a switch

Open a typical list app right after finishing a task. Watch what your brain does in the first three seconds.

1. Attention residue from the task you just finished. Leroy’s mechanism. Some of you is still on the prior task, especially if you didn’t quite finish or the close wasn’t satisfying.
2. Visual scan of the list. Twenty-three items, each one a small claim on attention. Some have tags, some have due dates, some are red.
3. Re-triage. You begin re-evaluating the order you established at the start of the day. Should I do that one first? It’s older. But this one is more urgent. But that one is what I was supposed to do today.
4. Stuck-cost on items unrelated to the next task. You notice item 17, which has been there for two weeks. Why haven’t I done item 17? Should I do item 17 now? Brief moral assessment. Snooze.
5. Maybe a pick. If you’re lucky.

That whole process is the bill. It’s paid in working memory. The list view is the cost. It is not “where you find the next task.” It is the structure that makes finding the next task expensive.

This is the part that’s specific to ADHD. A neurotypical user with ten items in the list pays steps 1 to 5 and pays them quickly, because their attention residue dissipates faster, their decision budget is larger, and their guilt response to item 17 is milder. An ADHD user pays the same five steps and pays each one harder, against a smaller budget, with more attention residue, and with item 17 generating real distress instead of a passing thought.

The cost of switching is not a multiplier on the cost of doing. It’s an additive tax that scales with list size and visual complexity. Most general-purpose apps optimize their list view for capacity (showing more items) and for filtering (showing relevant items). Both of those optimizations make the bill larger for ADHD users, not smaller.

The metric nobody optimizes

If you read app-review posts about productivity tools, the metrics that get compared are roughly: capture speed, sync speed, integrations, customizability, calendar integration, AI features, theme options, pricing.

The metric that would actually predict ADHD outcomes is: how much working memory does this app cost me to switch from one task to the next?

I have never seen this metric in a review. I have never seen an app marketing itself on it. The reason is that, for the median productivity-app user, the cost is small enough to not show up as a complaint. You can ignore it. The app gets reviewed on capture speed because capture speed is what the median user notices.

For ADHD users, transition cost is the dominant cost. Every other feature is downstream. An app with a beautiful capture flow and a 23-item default list view is, for an ADHD user, an app that subsidizes the cheap part and taxes the expensive one.

The design move that actually reduces the bill

There is one structural intervention that matters more than every other ADHD-friendly tweak combined: separate the planning view from the execution view, and put the execution view on top by default.

Planning is the activity where you look at the whole list, prioritize, sequence, defer, drop, and pick what to do today. It’s a deliberate, intermittent act. You do it once a day, or once a week, or whenever the day’s plan stops describing reality. It needs the full list. It needs filters and tags and dates.

Execution is the activity where you do the next task. It needs exactly three things on screen: the task, the first action, and elapsed time. It does not need the other 22 tasks. It actively suffers from their presence.

In a well-designed app, switching from one task to the next is: finish task. Tap once. See the next task already picked. The list view is not the surface you return to between tasks. The execution view is. The list comes out only when you’re deliberately re-planning, which is a small fraction of the day.

This is what Focus mode is in Ikoi, and it’s what the time blindness post was getting at from a different angle: the same architectural choice that helps with seeing time also helps with switching between tasks. Both problems are downstream of “the screen is too loud during execution.” Hiding the list during the work isn’t a feature for time perception or for task switching specifically. It’s the substrate that makes both work.

The picking still happens. It just happens once, when you plan the day, instead of n times, once per transition. You’re not trading control for simplicity. You’re paying the planning cost once and refusing to pay it again every time you finish a task.

The objection

But what if I need to change what’s next, mid-day? Plans change. I need to see the list.

Yes. The list is one swipe away. The argument is not that the list shouldn’t exist. The argument is that the list should not be the default surface between tasks. There is a difference between “the list is available” and “the list is what you see every time you complete something.”

In practice, the times you genuinely need to repick are far rarer than the times you stumble into the list because the app made it the default. The cost of the rare deliberate re-pick is small. The cost of the constant accidental re-triage is huge. The asymmetry is what makes the design choice clean.

If you do need to re-plan mid-day, do it once, on purpose, with intention. Then close the list and go back to executing. Don’t live in the planning view between tasks. The planning view is a tool, not a home.

The bill, re-itemized

If your todo app charges you for switching tasks, and the charge is paid in working memory, then the design choices that reduce that charge are the most load-bearing ones in the entire product. They aren’t cosmetic, and they aren’t ADHD-specific in the sense of being weird or niche. They’re the response to a real cognitive cost that has measurable research backing for the general population and a much larger version of the same cost for ADHD brains.

The features that matter, in roughly descending order of impact:

• A single-task execution view that’s the default, not a setting.
• A clear, deliberate way to plan once and stop re-planning.
• A way to pick the next task that doesn’t require scanning the full list.
• An explicit acknowledgement that the list view is for planning, not for execution.
• A timer and a first-step field inside the execution view, so the cost of resuming attention on the new task is also reduced. (The first-step field is the lever for that resumption cost.)

What does not matter as much as people think: capture speed, syntax for natural-language input, AI summarization, calendar integration, prettier checkboxes. These are the things that get compared in app roundups. None of them touch the actual bill.

If you’re building for ADHD users: the metric is transition cost. Optimize that, ship the rest second.

If you’re an ADHD user choosing a tool: open the app right after finishing a task and notice what happens to your attention in the next ten seconds. If it goes to the next task, the app is on your side. If it goes to a re-triage of seventeen unrelated items, the app is the bill.

Related: Time blindness is a visual problem covers why a timer in a cluttered list view doesn’t help, and the 2-minute door opener covers what to put in the execution view so the resumption cost goes down too.