StoweThe brain dump has survived every productivity fashion because the first half of it genuinely works: getting everything out of your head onto somewhere external is measurably calming, and the research on unfinished tasks intruding on attention backs the felt experience. (We covered that evidence in what a brain dump actually does.)
The second half is where it dies. You are looking at forty lines of half-sentences, and every guide says the same thing: now organise it. Categorise, prioritise, assign projects, build the board. That hour of sorting is the tax that makes the next brain dump less likely — and a brain dump you stop doing is worth nothing.
Here is the loop I build around, which has exactly four steps and no system to maintain.
Step one: say it, all of it
Talk, out loud, the way the thoughts actually arrive — half sentences, tangents, the guilty thing you keep re-remembering. Do not phrase for the page. The only rule of a dump is that it be complete; tidiness at this stage is friction wearing a virtue's clothes.
In Stowe this is the mic: speech is transcribed on the phone, and on-device AI splits the ramble into discrete items — tasks, routines, things — each with a guessed type and date you can fix with one tap. But the step stands without any app: speak into a voice memo and split it yourself, or write fast and messy. The point is that splitting comes after emptying, never during.
Step two: put a word on the few that want one
Not a category for everything — a label for the items where a label is free. "Work." "Home." "Errands." One word, one item, and only where the word is obvious. Anything that makes you stop and think about where a thing belongs is postponed or skipped; an unlabelled item is a normal citizen, not a filing failure.
This is the whole argument of tags versus folders, compressed: capture first, label cheap, keep the label set small. In Stowe the tags are free, capped at twelve, one per item, each wearing a colour derived from the word itself — so labelling costs one word and zero decisions.
Step three: arrange the room, don't file the contents
When the backlog gets long enough to need finding-things-in, resist the folder urge. Group what you can already see: by label, by kind, by day. Sorting and grouping are views — they rearrange the room without moving anything into boxes, and they can be switched off without anything being lost. A pile you can re-pile at will is fundamentally different from a filing cabinet you must maintain.
(In Stowe, arranging the backlog — grouped by tag, kind or day, sorted inside each pile — is part of Pro; the capture and the tags stay free. Nothing is ever hidden by an arrangement, only ordered.)
Step four: the day is 3–5 things
The dump is not the plan. From everything captured, today gets three to five things — a number the planning research keeps circling back to, because days sized to best-case estimates fail by design. Everything else waits its turn where you can see it, and tomorrow starts from what actually happened rather than from the whole mountain.
That is the entire system: empty your head, label what is cheap, arrange instead of filing, and cut a small honest day from the pile. Four steps, no board to groom on Sunday, and — this is the part that matters — nothing that punishes you for skipping straight from step one to step four on a bad day. A system you can half-use is a system you will still be using in March.
Frequently asked questions
What do I do after a brain dump?
Resist the urge to sort everything. Split the dump into discrete items, put a cheap label on the few that want one, and pick the 3–5 things that belong to today. Everything else waits its turn, visibly, without being lost. The sorting hour after a brain dump is where most systems die; the fix is to make sorting optional rather than a gate.
Should I organise my brain dump into categories?
Only as far as it stays cheap. A label per item (one word — work, home, errands) is cheap. A hierarchy of projects and sub-lists is not: it must be decided at capture time and maintained forever after. If categorising takes longer than capturing did, the system is now the task.
Is it better to brain dump by voice or in writing?
Whichever you will actually do at the moment your head is full. Voice has one structural advantage: speaking tolerates half-sentences and tangents, so nothing is lost to the act of phrasing. The transcript can be split into items afterwards — by you, or by software — which writing forces you to do while capturing.