Meetings That Should Be a Document

The asynchronous meeting is not a meeting. It is a document with a deadline attached, and the useful question is not whether your team is async-first — it is which of the eleven recurring meetings on your calendar survive being written down, and what each conversion costs somebody.

The short answer: status converts, review converts with a deadline, brainstorming half-converts, and anything carrying disagreement or authority does not convert at all. The reason is not about tooling or seniority. It is that a meeting exists to do one of two things — move information, or produce a decision that several people have to be able to live with — and only the first one is a writing problem.

Here is the conversion table by meeting type, the cost of each row, and the three that stay meetings.

The test: what does this meeting produce?

Before the table, the one question that sorts every meeting you have.

Does the meeting exist to move information, or to get agreement between people who might not give it?

Information transfer converts to writing, always, and usually improves: it becomes searchable, skimmable, and available to people who were not in the room. Agreement does not convert, because agreement is produced by the awkward bit — the pause, the objection somebody nearly did not raise, the visible reaction that tells you the thing you proposed is not going to fly. Writing removes exactly that, and removes it silently.

Most meetings are a mix, which is why the honest answer is usually “convert most of it and keep twenty minutes” rather than a clean kill. the fix-or-kill procedure is the general version; this post is the conversion half of it, in detail.

The conversion table: which meetings go asynchronous

Read the right-hand column as the actual decision. A conversion nobody is paid to do does not happen; the meeting quietly continues and the team concludes that async does not work for them.

Status is the row worth acting on first. It is the most common recurring meeting, it is pure information transfer, and it is the one where writing is strictly better — a reader who does not need Tuesday’s detail can skip it, which no standup allows. What the written version has to contain to actually replace the meeting is a separate skill from running the meeting was, and it is covered properly in what a written status update has to do to replace the standup. The common failure is writing the spoken standup down verbatim, which produces three lines nobody reads and a team that believes it has tried this.

Project review converts to a pre-read, not to a document that replaces the meeting. Circulate the material with a named deadline for comments, then hold a short session on the comments only. The full-hour walkthrough is the part that dies; the conversation about the two contested points is worth keeping and takes twenty minutes.

Brainstorming half-converts, and the half matters. Fifteen minutes of everyone writing independently before anyone speaks produces more distinct ideas than an hour of talking, because it removes the anchoring effect of whoever speaks first and the reluctance of whoever is most junior. Then converge live. A brainstorm that is fully asynchronous tends to produce a list nobody builds on.

What each conversion actually costs

Three costs that decide whether a conversion sticks, none of which is the tool.

Somebody has to write it, every cycle, forever. The meeting was an hour of six people; the update is fifteen minutes of one person, which is a real saving and a real burden on a specific individual. Name that person out loud when you make the conversion. A conversion with no named writer reverts inside a month — this is the single most common way it fails, and it is entirely predictable.

Reading is not free and it is not enforced. A document that goes unread has not replaced anything; it has moved the failure somewhere quieter, where nobody notices for six weeks. The meeting at least forced attention into the same hour. If you convert, you need a way of knowing whether it is being read — a decision that has to be acknowledged, a question in the doc, something.

The quiet people get quieter or louder, and which one is not predictable. Writing helps the person who thinks slowly and hates interrupting. It disadvantages the person whose English is second-language, or who is dyslexic, or who is simply a better talker than writer. Both effects are real. Watch for who stops contributing after a conversion and ask them directly.

There is also a cost that lands on nobody in particular and so gets missed: converted meetings stop producing the sideways information that made people feel oriented. You will hear this as “I don’t know what anyone’s working on any more” about two months in. The fix is not to reinstate the meeting but to keep one live session that is explicitly for that, with no agenda filter on it, and the quarterly review is where you notice you need it.

The three that stay meetings

Anything where the answer is contested. If two people want different things and one is going to lose, that is a meeting. In writing, disagreement escalates — each round is drafted, re-read, and interpreted for tone by an audience, and positions harden with every message. Live, it takes eleven minutes and somebody concedes. A written thread on a contested point is the most expensive form this conversation takes.

Anything carrying authority or consequence for a person. Performance conversations, reorganisations, someone’s role changing, an escalation. Written, these are documents that can be forwarded, which changes what people are willing to say and what they hear. This is not a productivity judgement; it is what the thing is.

The 1-1. It does not convert and it does not shrink. Its value is almost entirely in the parts nobody would write down — the half-raised concern, the thing that is going wrong that is not yet a problem. A 1-1 replaced by a written check-in becomes a status report with a friendlier tone, which is a worse version of a thing you already have.

There is a fourth that is not a rule but is worth knowing: the meeting that is somebody’s only access to a person or a project. It fails every efficiency test and converting it removes something that will not be visible in any measurement you take afterwards.

Converting a meeting you do not run

Most of the table above assumes you own the meeting. The version available to someone who does not is narrower and still works.

Write the artifact first, then propose the conversion. Not “should this be a document?” — send the document. One week, unprompted, write the update the meeting would have produced and post it before the meeting. The argument is then about a thing that exists rather than about a principle, and the thing that exists has already saved everyone the hour once.

Ask for one item, not the meeting. “Could the status round go in the channel beforehand? Happy to write mine there on Wednesdays and see if it works.” Converting one item is a request the organiser can approve without deciding anything about meetings in general, and it is how most of these actually start.

Take the writing. The person who offers to write it gets the conversion approved, because the objection in the organiser’s head is never “writing is bad”, it is “who is going to do that”. the meeting types worth targeting first is where to spend that offer, and the downgrade ladder is the smaller move when even that is too much to ask.

How a converted meeting comes back

Watch for the slot being rebooked ad hoc “just this week to go through it properly”. Once is fine. Twice in a month means the document is not carrying what the meeting carried, and the useful response is to find out which part rather than to defend the conversion — usually it is one specific thing, and putting a fifteen-minute call back for that one thing is the right answer rather than a failure of the idea.

The conversion that lasts is almost never the full replacement. It is the hour becoming a document plus twenty minutes, which is a 65% saving that nobody argues with, and it holds for years.