Most note-taking advice assumes a stable relationship with future intentions: you file something now because you will want it later, and you will remember to look. When attention is inconsistent, that assumption breaks in both directions—filing does not happen at capture time, and the looking does not happen at retrieval time. What follows is a system built for that reality rather than around it. It is a workflow, not medical advice, and it is not a treatment for anything.
Why conventional systems fail here specifically
The standard advice—choose a folder, apply tags, review weekly—places its cost at exactly the two moments that are hardest: the moment of interruption and the moment of scheduled maintenance. A note taken while three things are half-finished cannot survive a filing decision, and a weekly review with no external deadline loses to everything that has one.
There is also a measurable cost to the interruption itself. Rubinstein, Meyer, and Evans, in work published by the American Psychological Association in 2001, found that switching between tasks introduced significant time costs, with people losing as much as forty percent of productive time in the conditions studied. That research was on task switching generally, not on ADHD specifically—but it explains why any capture method requiring a decision imposes a cost far larger than the seconds it appears to take.
Rule one: capture must cost nothing
The single highest-value change is reducing capture to something you can do without stopping what you are doing. Three actions from a locked phone to a saved thought is the threshold; above it, capture reliably fails during the moments that produce the thoughts worth keeping.
That means no titles, no folders, no choosing between apps at the moment of capture. Speaking one sentence should be a valid note. An unfinished fragment should be a valid note. The archive can absorb mess; your attention at that moment cannot.
Rule two: never rely on remembering to look
A note that requires you to remember to reread it has moved the problem rather than solved it. Anything with a consequence—a promise, a deadline, a form to return—needs to leave the note and become something that surfaces on its own.
- Say the deadline out loud inside the note: ‘tell Marta by Thursday’ is extractable; ‘tell Marta’ is not.
- Name the person. Commitments are recalled by who they were made to far more reliably than by topic.
- Let promises become tracked tasks that keep a link back to the note they came from, so the reason survives alongside the reminder.
- Accept that some items will resurface late. Late is a different failure from never, and it is the one worth optimizing for.
Rule three: search by description, not by filename
Trying to recall where you put something is a memory task with a high failure rate. Describing what it was about is far easier and much more robust: ‘the thing about the school form and the deadline’ works even when nothing in that phrase appears in the original note.
This is why retrieval by meaning matters more here than filing does. It replaces a task attention is poor at—remembering a location chosen months ago—with one it handles well: recognizing something once it is in front of you.
Rule four: repair context in the first hour, not the first month
Fragments captured at speed lose their meaning quickly. ‘Ask him about the thing before it closes’ is recoverable within an hour and largely unrecoverable in six weeks. The cheapest possible maintenance is a single clarifying line added soon after capture—one sentence, no reorganization.
A system that asks you a short follow-up question shortly after an ambiguous capture does this work at the only moment it is cheap. This is the one habit in the list worth defending, because it is what keeps a fast archive from becoming an unreadable one.
What this approach does not fix
It does not improve attention, and it should not be sold as doing so. Nothing here is a clinical intervention, and any productivity system marketed as an ADHD remedy deserves scepticism—the evidence for tooling as treatment does not exist in a form worth citing.
It also does not recover what was never captured, and it will not remove the friction of the first weeks, when the archive is too small to answer anything. Automatic topics will mislabel material, and a captured task with a vague deadline will surface at a vague time. The realistic promise is narrower than most productivity writing admits: fewer dropped commitments, and a much shorter path from ‘I know I wrote this down’ to actually holding it.
Optimize the two ends and ignore the middle. Make capture cost under three actions, make anything with a consequence leave the note and become a tracked task, and let search work from descriptions. Organizing is the part you can skip; capturing and resurfacing are not.