How to Take Better Meeting Notes Without Transcribing

A client once forwarded me her notes from a three-hour steering committee. Eleven pages, typed live, nearly word for word. She asked me to find the decisions in them. I read the pages twice and found two, both buried under a long exchange about a venue contract. She had captured everything and kept nothing. If you want to know how to take better meeting notes, start from that failure: the goal is not to record the meeting, it is to keep what you will need afterwards.

Notes fail for a simple reason. People write down what is being said when they should write down what changes. A meeting produces a small number of events that matter: a decision is made, a question is left open, someone commits to something, or a fact surfaces that contradicts what you believed walking in. Everything else is conversation, and conversation is already gone the moment it is spoken.

The second failure is timing. Notes taken in the room are raw material. Nobody can read raw material three weeks later and reconstruct the meeting, including the person who wrote it. The notes only become useful once someone sorts them, and that sorting rarely happens unless it is scheduled.

The third failure is mixing audiences. Your own working notes and the minutes you circulate to others are different documents. Minutes record agreed outcomes so a group can act on them, and we have written separately about how to write meeting minutes. Your notes answer a narrower question: what do I, personally, need to remember, chase or challenge? Keep the two apart and both get better.

Capture four things, and let the rest go

The working rule I give clients is four buckets. Decisions, open questions, commitments, surprises. That is the whole system, and it fits on one page for a meeting of any length.

A decision is anything the group treated as settled, even if nobody said the word. Write it as a sentence a stranger could understand: "Budget for the pilot capped at 12,000 euros, approved by finance lead." A question is anything raised and not resolved. Write who is expected to answer and by when, if that was said. A commitment is a promise with a name attached, yours included. A surprise is the bucket people skip, and it is often the most valuable: a number you did not expect, an objection you had not anticipated, an interlocutor who changed position.

Three or four lines per bucket is normal. If a bucket is empty, that is itself information. A one-hour meeting with no decisions and four open questions has told you something about the state of the project.

Resist the urge to record reasoning in full. When a group debates two options for twenty minutes, you need the choice, the one reason that carried the argument, and any objection that was overruled but not withdrawn. That is three lines, not three paragraphs.

Prepare ten minutes before, and the notes mostly write themselves

Most advice on note-taking starts when the meeting starts. The gain is earlier. Ten minutes before the meeting, write three things at the top of the page: why you are attending, what you need to leave with, and what you suspect will go wrong.

The first line tells you what counts as relevant. A consultant attending to confirm scope listens differently from one attending to defend a deadline. The second line gives you a test at the end: did I leave with it, yes or no? The third line primes the surprise bucket, because you are more likely to notice a contradiction if you wrote down the expectation it contradicts.

If an agenda exists, copy its items down the left margin and leave space. If none exists, that is worth noting too, and worth asking for next time. A meeting without an agenda produces notes without a spine.

Paper, laptop or microphone: what the evidence actually says

You will have heard that handwriting beats typing. The claim comes from a 2014 study by Mueller and Oppenheimer, in which students who typed notes during a lecture recorded more words but answered conceptual questions less well, apparently because typing invited near-verbatim transcription. It is a neat result and it travelled far.

It has not held up well. A later replication reported in an Association for Psychological Science summary found no reliable advantage for handwriting, and a replication and extension published in Educational Psychology Review reached a similar conclusion, with trends favouring longhand that did not reach significance. What survives is a narrower lesson: whichever tool you use, transcribing word for word is the habit to avoid. Paraphrasing forces a choice about what matters, and that choice is the work.

So the practical answer is plain. Use what lets you paraphrase. If you type fast enough to transcribe, slow yourself down by using the four buckets, which forces selection. If paper keeps you from checking messages during a call, use paper. The tool matters less than the discipline.

The microphone is a different question, and we will come back to it.

The ten-minute review is where notes become useful

Here is the habit with the best return. Within a day of the meeting, give yourself ten minutes and a pen in a different colour. Read the page once. Do four things.

First, rewrite each decision as a full sentence if it is not one already. Second, convert every open question and commitment into an entry in whatever system you actually use for tasks, with a date. Third, circle the one surprise that changes what you will do next. Fourth, write a single summary line at the top: "What this meeting changed."

That last line is the most useful thing on the page. When you search your notes in four months, you will search for it. Highlighting is choosing. Choosing is understanding. The summary line is that choice made visible.

If you skip the review, the notes decay quickly. Not because the paper fails, but because the context that made your abbreviations legible disappears. A shorthand mark that meant "ask Marta about the contract clause" on Tuesday reads as noise on Friday.

Where AI recording tools help, and where they quietly hurt

Recording and transcribing meetings has become cheap, and I use it on client calls where I have permission. The strength of these tools is completeness: they catch the sentence you missed while writing the previous one. The weakness is that completeness is exactly the failure described at the start. A full transcript is the eleven pages again, only faster.

Used well, the recording is a safety net behind your own notes, not a substitute for them. Take your four-bucket notes live. Afterwards, check them against the transcript for anything you missed, and use the transcript to confirm exact wording when it matters, for a figure or a commitment someone may later dispute.

Three cautions from practice. Ask before you record. In many jurisdictions consent is a legal requirement, and in every one it is a courtesy that changes how candid people are. Check the automated summary against the transcript before you trust it, because summaries smooth over disagreement and can attribute a decision to the wrong person. And do not let a tool's action-item list replace your own commitments bucket: a generated list tells you what was said, not what was agreed.

Granola is the one I would try first if you want a tool that works alongside your own typed notes rather than replacing them. It takes the rough notes you write live and uses the recording to fill gaps, which keeps you in the habit of selecting.

Fathom records and summarises calls with a free tier suited to occasional use. It suits people who attend many calls and mostly want a searchable record, with the understanding that you will still do the review yourself.

Otter.ai is the long-standing name in live transcription. It is strong on searchable transcripts, and weaker if you hope it will decide what mattered. If you are weighing it against others, our comparison of Otter alternatives sets out the trade-offs.

Different meetings need different notes

The four buckets hold up everywhere, but their weight shifts with the meeting type.

In a one-to-one, the commitments bucket dominates, and it should include what the other person promised you. Write the promises down in front of them. Nobody objects, and it quietly raises the standard of what gets promised.

In a large status meeting, surprises matter most, because most of the content repeats what you already knew. Your job is to spot the one item that does not fit.

In an interview or fieldwork call, as journalists and researchers know, exact quotes matter more than paraphrase. Mark them clearly, note the time on the recording, and paraphrase everything else. The mix of the two is the method.

In a workshop or brainstorm, do not try to record ideas in sequence. Capture clusters and the criteria the group used to rank them. The ranking is the outcome; the list of ideas is only the input.

Where to keep them so you can find them later

A good note is worthless if you cannot find it. The simplest scheme I have seen work is one note per meeting, titled with the date and the project, with the summary line at the top. Keep them in a single place. People who split notes between a notebook, a phone app and a shared document spend more time searching than reading.

Search is only half of retrieval. The other half is linking: when a decision touches a project page, paste the decision sentence there and link back to the meeting note. In a tool like Notion, that is two minutes and it means the project page carries its own history.

Notion works for this if you already live in it, and there is no reason to adopt it only for meeting notes. A plain folder of dated files does the job as well, provided the summary line is always the first thing in the file.

If you handle long documents as well as meetings, the same discipline applies to reading: select first, summarise second, store third. Our guide to the Cornell note-taking method covers one structure that adapts well to both.

What to do on Monday

Pick one recurring meeting this week. Before it, write your purpose, your required outcome and your suspected risk. During it, use only the four buckets. Within a day, spend ten minutes on the review and write the summary line.

Do that three times and compare the notes with whatever you produced before. The difference is rarely about handwriting speed or app choice. It is that you stopped recording a meeting and started keeping a decision trail. Whether you then add a recording tool behind it is a smaller question than it looks, and a better one to answer once the habit is in place.