Advice about fast note-taking usually offers shorthand, abbreviations, and templates. None of those address the constraint that decides the outcome, which is how many actions stand between a thought arriving and it being saved somewhere you will look. Below that threshold you capture things during the moments that produce them. Above it you capture the calm moments only, which are not the interesting ones.
What speed costs, measured
The input method dominates everything else. In a MobileHCI 2019 study by researchers from Aalto University, the University of Cambridge, and ETH Zürich, 37,370 volunteers averaged about 36 words a minute overall; people typing with two thumbs averaged about 38, roughly a quarter slower than participants in a comparable physical-keyboard study.
Conversational speech runs at roughly 130 to 150 words a minute. So the choice between speaking and typing is worth more than every abbreviation system ever invented, by a factor of three or four. No shorthand recovers that gap.
‘Quickly’ and ‘effectively’ genuinely conflict
The best-known experiment is Mueller and Oppenheimer's 2014 paper in Psychological Science, ‘The Pen Is Mightier Than the Keyboard.’ Students typing lecture notes captured more words verbatim but performed worse on conceptual questions than students writing longhand, who had to select and rephrase because they could not keep up.
It is worth knowing that the picture is less settled than the paper's fame suggests. Morehead, Dunlosky, and Rawson published a direct replication attempt in 2019 and did not reproduce the main effects. What survives both is narrower and still useful: transcribing verbatim is a poor use of attention when you are trying to understand something in the moment.
Which resolves the apparent conflict. Speed is the right goal when you are capturing something you already understand and want to preserve—a decision, a name, an idea, a commitment. Slowness is the right goal when the writing is how you are doing the thinking. Almost all phone note-taking is the first kind.
The three-action rule
From a locked phone to a saved thought should take three actions. Unlock, open, speak or type. If it takes five, capture quietly stops happening while you are walking, cooking, or between two meetings.
Count yours honestly, including the ones that do not feel like actions: choosing which app, choosing a folder, tapping a title field you then leave empty. Every decision at capture time is a place where a thought can be lost, and the thought does not wait.
Write the trigger, not the summary
A complete note takes a minute you do not have. A trigger takes eight seconds and is enough, provided it contains the two things that decay fastest: who or what it is about, and why it mattered.
- Weak: ‘call back’. Strong: ‘call the clinic back about the referral date, they close at 4’.
- Weak: ‘good idea from the podcast’. Strong: ‘podcast idea—charge for the migration, not the seat count’.
- Weak: ‘Marta’s thing’. Strong: ‘Marta wants the training days itemized separately on the renewal’.
- The pattern is the same each time: name the subject, then say the one fact that makes it worth keeping. Everything else can be reconstructed; those two cannot.
Repair in the first hour, not the first month
Fragments lose meaning fast. ‘Ask him about the thing before it closes’ is recoverable within an hour and largely unrecoverable in six weeks. So the second half of fast capture is one clarifying sentence added soon after, while the context is still free—not a reorganization, not a review, one sentence.
This is also the only maintenance habit worth defending, because it is what stops a fast archive from becoming an unreadable one. A system that asks a short follow-up question about an ambiguous capture does the same work at the same cheap moment, which is better than relying on remembering to do it yourself.
Techniques that do not survive contact with a phone
Shorthand systems require practice and can produce notes that are hard to search. Cornell-style layouts assume a page and a lecture. Tagging at capture time is a filing decision wearing a different hat, and it creates the same friction. Color coding works only while the palette remains small enough to remember without stopping to choose.
The two that do survive are unglamorous: one destination for everything, so there is no choosing; and speaking instead of typing whenever your hands are busy or the thought is longer than a line.
What fast capture will not do
It does not produce understanding. A perfectly captured lecture is not a learned lecture, and that is what the 2014 experiment was actually about. If the goal is to understand rather than to preserve, write less and slower, on purpose.
It also cannot recover what was never captured, and it will not feel useful in the first few weeks, when the archive is too small to answer anything. And speed has a real cost in precision: spoken notes ramble, leave the subject implicit, and mistranscribe names. The trade is worth making, but it is a trade rather than a free win.
Get capture under three actions and speak instead of typing when your hands are busy—that is worth more than any shorthand. Then write triggers, not summaries: name the subject, say the one fact that made it worth keeping, and add a clarifying sentence within the hour while the context is still free.