The decisions are clear for about ten minutes. Use them.

A meeting ends and everyone in it agrees, for a little while, on what was decided. The deadline, who owns which piece, what got cut from scope. Ten minutes later, the next call has already started to crowd it out. By the time you'd normally sit down to type the follow-up — after lunch, at the end of the day, whenever the inbox lets you — half of what was sharp in the room has gone soft. Dictation closes that gap by matching the speed of your typing to the speed you still remember things: you hold a key and say the recap the moment the call ends, before anything else gets a chance to write over it.

Send the recap before you do anything else

The habit that makes this work is ordering, not speed. Do not open your task list, do not check Slack, do not glance at the next event on your calendar. The moment the call window closes, hold the key and speak the recap straight into a reply to the meeting's own email thread, or into a new message to the people who were on the call. It takes less time to say than it took to read this paragraph, and it is the single highest-leverage minute in most people's meeting day, because it is the only point at which the whole group's shared understanding of what just happened is still, briefly, identical.

Say it as you'd say it out loud to someone: what got decided, who's doing what, and by when. You don't need to draft a formal minute. A follow-up that reads like a person talking is more likely to get read than one that reads like a document, and dictation naturally produces the former — the words that land are the words you said, not a tidied-up version of them.

Keep a shorthand list during the call, not a transcript

None of this depends on Coii VoiceInput hearing the meeting. It doesn't, and it isn't built to — it's not listening while you're not holding the key, and the app has no way to know a call is even happening. What it needs is you, holding three or four words per decision in a notes app or on paper during the meeting itself, and then turning that shorthand into full sentences the moment you're free to dictate.

"Priya — migration doc — Fri" becomes, spoken out loud, "Priya owns the migration doc and will have a draft by Friday." The shorthand is there so you don't have to trust memory alone for forty-five minutes; the dictation is where it becomes something someone else can act on. Keeping the list short is deliberate — a shorthand list you can glance at while composing out loud is more useful than a full transcript you'd have to read back to yourself first.

Turn the recap into task-tool entries, one at a time

The group follow-up email is usually one dictation. Individual action items going into a task tool — Things, Linear, a shared board — work better as separate ones. Click into the task title field, hold the key, say the task as an instruction: "Send the contract redlines to legal by Wednesday." Let go. Click the next field, repeat.

This is slower than dictating one long list and splitting it up by hand afterward, but it's more reliable: each task is its own row in your history, so if a paste into the task tool fails or lands in the wrong field, you're recovering one line, not re-parsing a paragraph to find which sentence was meant to be which ticket.

Names, tools and jargon that keep coming out wrong

Meeting follow-ups are dense with the kind of words a general-purpose engine gets wrong by default: colleagues' names, internal tool names, acronyms specific to your team. Fixing how a term comes out is mostly a one-time cost. Add the attendee's name, the project's internal codename, and the two or three acronyms your team uses every week to your dictionary once, and every follow-up after that gets it right without you having to slow down and enunciate.

This matters more for meeting follow-ups than almost any other use of dictation, because a follow-up email with a wrong name in the first line reads as careless in a way a wrong word in a private note never does. Spend the five minutes setting up the dictionary before your next standup, not during your first attempt at a recap.

Why this is safe to do with sensitive material

A meeting follow-up is often the most sensitive text anyone dictates all week — a compensation discussion, a client's contract terms, a note about a personnel issue. Nothing about it leaves the Mac: no audio, no transcript, uploaded anywhere, and the app works exactly the same with the network off as with it on. That's not a setting to enable — there's no cloud mode to turn off in the first place. For a follow-up you wouldn't want sitting on a third party's server, that's the whole argument.

Recovering a follow-up that didn't land where you expected

Every dictation is written to your local history the instant you let go of the key, before the app has tried to put it anywhere. If a follow-up lands in the wrong field, or a task tool's paste doesn't take, the sentence isn't gone — it's sitting in history exactly as you said it, ready to copy into the right place. The mechanics of recovering it are the same as for any other dictation; the only difference with a meeting follow-up is that you're more likely to notice quickly, because you're usually looking right at the inbox or the board you just tried to fill in.

Making it the default, not a one-off

The first few times you dictate a follow-up instead of typing it, it'll feel like an extra step — reaching for the hotkey when your hands were already near the keyboard. That friction goes away with repetition, the same way any new default replaces an old one. Building the habit is worth doing deliberately rather than leaving it to chance, because meeting follow-ups are exactly the kind of recurring, time-pressured task where the habit pays for itself fastest — you do it several times a week, the stakes of forgetting something are real, and the old way of doing it (typing from memory, later, worse) has a cost you can feel.

Developers doing the same thing after a design review, or an account manager after a client call, run into the identical shape of problem: a workflow built around saying things once, right after they happen, rather than reconstructing them from memory an hour later.

Internal recaps versus external ones

Not every follow-up is written the same way, and dictation adapts to both without you having to think about which mode you're in — there is no mode, just what you say. A recap to your own team can be short and informal: names, owners, dates, done. A recap to a client or an external partner usually wants a sentence of framing before the list — "Thanks for the time today, here's where we landed" — because the reader wasn't in the room for the tone of the discussion the way your teammates were.

Both are still faster dictated than typed, but it's worth taking the extra few seconds to say the external one as you'd actually want it read, rather than as a bare list of bullet points. Dictation doesn't clean up register for you — what you say is what lands — so a follow-up that needs to sound considered has to be said that way out loud, not fixed afterward by a model guessing what you meant.

Back-to-back meetings and batching the follow-ups

Some days hand you three or four calls in a row with no real gap between them. The ten-minutes-after rule bends here: if the next call starts in ninety seconds, you don't have time for a full recap, and trying to rush one usually produces something worse than doing it properly a few minutes later. In that case, take five seconds for the shorthand list — the same three or four words per decision you'd use during any meeting — and dictate the full recaps for all of them together the moment there's an actual gap, working through the shorthand lists in order while each one is still recent enough to expand accurately.

This is slower than doing each one immediately, but it's a lot faster than the alternative most people fall back to, which is trying to reconstruct four meetings' worth of decisions from memory at the end of the day, once everything has blurred together. A shorthand list taken during the meeting is worth more than it looks like at the time — it's the thing that makes batching possible at all.

What a week of this actually looks like

Monday's standup produces two lines to the team channel. Wednesday's client call produces a recap email and three tasks. Friday's retro produces a longer note to yourself about what to change next sprint. None of these take more than a minute of speaking, and none of them get typed from a blank page an hour later, because the ten minutes right after each call is exactly the window this is built for. The habit is not "dictate more" — it's "dictate the recap before you do anything else," every time, until it stops being a decision.

If meeting follow-ups are the first place you try dictation, they're a reasonable place to start: the text is short, the stakes are real enough that you'll notice if it saves you time, and the moment to do it — right after the call — is the same moment every time. Get Coii VoiceInput and try it on your next follow-up.

A recurring meeting is a recurring dictation

If you run the same meeting every week — a standup, a client check-in, a status call — the follow-up ends up close to the same shape each time, which makes it one of the easiest places to notice whether dictation is actually saving time. Compare the third or fourth week's follow-up to the first: the words you say get more efficient as the format settles, even though nothing about the app changed. That's the habit forming around the specific task, not a general improvement in how fast you talk.

It's also where dictation quietly earns its keep against typing the most. A recurring follow-up typed from a template still requires filling in the parts that changed — who's covering what this week, what shifted, what's now blocked. Those are exactly the parts that are fastest to say once you already know them and slowest to type carefully, because careful typing means checking spelling on names and dates you already know cold. Saying them once, correctly, and letting the words land is where the time comes from.

One recap, not a meeting record

It's worth being precise about what this is and isn't. A follow-up dictated after a call is a short piece of writing produced from what you remember and noted — it is not a transcript of the meeting, and Coii VoiceInput has no way to produce one, because it was never recording. If what you actually want is a full record of who said what during the call itself, that's a different tool and a different kind of product than this page is about. What's covered here is narrower and, for most people, more useful day to day: turning "what we just decided" into a sent message before it has time to fade, using nothing more than the notes you took and the hotkey you already have set up.