Minutes, recap, or decision log: what to write when

Minutes record what happened in the room, a recap tells the people who were there what happens next, and a decision log stores what was settled so someone can find it later without asking you. Three jobs, three formats. The usual mistake is writing one document that half does all of them.

You know the shape of that document. Nine paragraphs of who said what, two real decisions buried in the middle, sent a day and a half late, filed on a page nobody searches.

How to choose in one question

Ask who reads it and how long it has to survive. The format falls out of the answer.

• The readers were in the room and the useful life is about a week: write a recap.
• The reader was not there and may turn up months later: write a decision log entry.
• The reader could be an auditor, a board, or a lawyer: write minutes.

Nearly every meeting you sit in is the first case. Some are the second, and those deserve the extra effort, because a decision that survives is worth more than a decision that gets made twice. Almost none are the third, whatever your note-taking habit suggests.

The question is worth asking out loud because the default is inertia. Whatever your team did last time is what gets done again, long after the reason expired.

What minutes are actually for

Minutes exist to prove that a meeting happened and to record what was formally resolved, with the attendance attached. That is why they read like something no human opens by choice.

Proper minutes carry the attendees, the apologies, each motion, the wording of what was passed, and the vote. The voice is deliberately flat. The point is evidence, and evidence should be boring.

If nobody will ever have to prove any of it, minutes are theatre. I have sat in meetings where someone typed nearly every sentence spoken, and the file was never opened again by anyone, including the person who typed it.

Keep minutes when a board, a works council, a regulator, or a contract asks for them. Otherwise you are producing a transcript to make an ordinary meeting feel important, and paying for it with the one hour after the meeting when a recap would still have been worth something.

A recap is a message, not a document

A recap is a short message sent to the people who were in the room, quickly, so everyone leaves holding the same version of what was agreed.

It belongs in the thread or the inbox where those people already are. The moment it grows a heading structure and moves to a page of its own, it has stopped being a recap and started being homework.

Speed is most of its value. Send it before the room's memory hardens into several slightly different versions.

How to write one well, what to capture and in what order, I have already put down elsewhere and will not repeat here. One line from it belongs in this argument though: put a name against every open question, because a question with no owner is a question nobody answers.

The decision log is written for a stranger

A decision log is a durable list of settled questions, written so a person who was not in the room can act on the answer without tracking you down first.

One entry per question. Each entry holds:

• The question, phrased the way someone would search for it later.
• The answer, in one sentence, saying what happens now.
• The date it was settled and the person who owns it.
• The options you turned down, each with the reason you turned it down.

Most logs stop after the third line. The fourth is the one that compounds. A decision recorded without its rejected options gives you the answer and hides the thinking, so the next person to arrive restarts the argument from zero, in good faith, with no idea they are running a rerun.

The entries that have fooled me while building Unvent are the tidy ones. Outcome written down, reasoning left out, and months later I take my own word for it. A note that records only what was decided reads as settled. It hands future you nothing to pull on, which is worse than leaving the question visibly open.

The entries I resented writing are the ones I have reread most.

What breaks when you pick the wrong one

The wrong format fails quietly. A document exists, so nobody notices it is doing the wrong job, and the bill arrives weeks later as a repeated argument or a dropped handoff.

• Minutes where a recap was needed: everyone gets a transcript and nobody gets an instruction.
• A recap where a log entry was needed: the decision sits inside a thread that is unfindable a few months on.
• A log entry where a recap was needed: the page updates in silence and the people who were in the room hear nothing at all.

Watch for the third. It looks like diligence. You wrote it down, and you wrote it down somewhere nobody was looking.

Writing two of them from one meeting

One meeting can produce both a recap and a log entry, and the second one costs a fraction of the first.

Send the recap first, while it still has value. Then lift the decisions out of it, rewrite them in the phrasing someone would search for, and add the options you rejected with the reasons. The recap was written for people who already have the context. The log entry is written for people who have none, which is why copy and paste does not work between them.

Skip the log entirely for meetings that settled nothing. Status meetings, most standups, anything that ends with people better informed and no question closed: those earn a short written update at best, and a written update is a different job again.

One test tells you whether any of this is working. Count the questions your team decides twice. If a thing gets settled in a room, then reopened, then settled the same way by different people, either your record is not being read or it never held the reasoning in the first place. Fix that before you go shopping for a better wiki.