Choose a meeting when the topic is ambiguous, emotionally loaded, hard to reverse, or needs several people to hear the same thing at once. Choose writing when the information is settled, the decision is cheap to undo, or you mainly need people informed rather than aligned. Most recurring meetings fail all four tests — which is exactly why they feel like a waste.
That sounds obvious until you look at your own calendar. Half the meetings on it are there because you scheduled them, and you scheduled them because a meeting is the easiest thing to reach for when something feels unresolved. It requires no writing, no structure, and no decision about what you actually want — just a block of other people's time and the vague hope that talking will sort it out. Sometimes it does. Often it just moves the confusion into a room.
The point of a framework isn't to have fewer meetings for its own sake. It's to stop paying live-time prices for things writing does better, so the meetings you do hold have room to be good.
What does a meeting actually buy you?
A meeting has one genuine superpower: real-time back-and-forth between people who can read each other. Someone floats a half-formed idea, three people react, the idea changes shape, and you land somewhere none of you would have reached alone. Tone comes through. Hesitation is visible. You can tell the difference between "yes" and "yes, but I have reservations I'm not going to type out."
The price is steep and mostly invisible. A one-hour meeting with six people is six hours of salaried attention, but the bigger cost is what it does to the hours around it. A meeting at 11am doesn't cost an hour — it costs the morning, because the 45 minutes before it aren't long enough to start anything real. That fragmentation is the same mechanism behind context switching: the calendar chops the day into pieces too small for concentrated work, and the deep, uninterrupted stretches your team needs for their hardest problems never appear.
So the honest question isn't "would a meeting help?" — a meeting almost always helps a little. It's "does this need live time enough to justify what live time costs?"
What are the four criteria that decide it?
Four things reliably separate conversations that need a meeting from conversations that don't. Score your topic against each one.
Ambiguity. How well-formed is the thing? If you can write down the question clearly, you can probably get the answer in writing. If you can't even name the problem yet — "something's off with how we're handing off work" — you need people talking, because the exploration is the work.
Emotional weight. Anything involving disappointment, criticism, uncertainty about someone's role, or bad news needs a face and a voice. Written words harden. A message reads colder than you meant it, gets reread at midnight, and no one can ask the follow-up question that would have defused it.
Reversibility. Cheap-to-undo decisions don't need consensus rituals. Pick one, try it, change it next week. Expensive or hard-to-reverse decisions — a commitment to a client, a change to how someone's work is evaluated — deserve the friction of everyone in the same conversation, so objections surface before the door closes rather than after.
Number of people who need to be aligned. Not informed — aligned. Informing scales beautifully in writing. Alignment doesn't: if five people each need to adjust their plan based on what the others say, threading that asynchronously takes days and usually ends in a meeting anyway.
Which format fits which situation?
| Situation | Best format | Why | Trade-off to accept |
|---|---|---|---|
| Sharing a decision that's already made | Written message | No one needs to react live; people read on their own time | Feels impersonal — add the reasoning, not just the outcome |
| Status updates across a team | Written, in one place | Reading is faster than listening; skimmable and searchable | Someone has to actually write it, and read it |
| Exploring a vague, unformed problem | Meeting | Ambiguity needs back-and-forth to resolve | Costs live time and needs a facilitator or it drifts |
| Feedback, praise, or a difficult conversation | Live, one-to-one | Tone and follow-up questions matter more than the words | Doesn't scale; you carry the emotional load |
| A reversible choice inside one person's remit | No conversation — let them decide | Speed beats consensus; that's what delegation is for | You may not love every call they make |
| A hard-to-reverse decision touching several teams | Meeting, after a written pre-read | Objections surface live; the pre-read means you skip the briefing | Slow to schedule; needs prep discipline |
| A long, detailed proposal | Document, then a short meeting | Reading is far faster than being read to | Someone must write it properly |
| Recurring team sync with no agenda | Cancel it, or shorten it hard | An empty agenda means there was nothing that needed live time | You lose incidental connection — replace it deliberately |
The row people resist most is the fifth one. "A reversible choice inside one person's remit" gets escalated into a meeting constantly, not because it needs one, but because the person doesn't know how much authority they have. That's not a meetings problem — it's a delegation problem, and naming the level of authority up front removes more meetings from your calendar than any scheduling policy will.
How do you make the meetings you keep worth attending?
Once you've cut the ones that shouldn't exist, protect the ones that should.
Write the decision you need before you send the invite. One sentence: "By the end of this, we will have decided X." If you can't write it, the meeting isn't ready. If writing it produces an answer, you've just saved everyone an hour.
Invite people who will speak. Everyone else gets the notes. Adding an observer "to keep them in the loop" is using a meeting to solve an information problem, which is the exact thing writing is better at.
Send the context in advance and expect it read. Reading a page takes three minutes; being talked through it takes twenty. A pre-read converts most of a meeting into the part that actually needed people present.
Batch them. Three meetings back-to-back cost less total focus than three meetings scattered across a day, because they carve one hole instead of three. That's the same logic as protecting a block for deep work — clustered interruptions leave usable stretches behind.
Keep the one-to-one. When managers go on a meeting diet, the recurring one-on-one is often the first casualty, and it's the one meeting that scores high on every criterion: ambiguous, emotionally weighted, and impossible to do in writing. Cut the status sync instead.
FAQ
How do I cancel a recurring meeting without upsetting people?
Don't frame it as a cancellation — frame it as a trial. "Let's skip the next three and see what breaks; if we miss it, we'll bring it back." That costs no one anything to agree to, and in practice most teams don't ask for it back. Say clearly where the information will live instead, so nobody feels cut off.
Isn't async communication slower?
For a single decision, yes — a message thread can take two days to do what ten minutes of talking would. For the team's total output, usually no, because async doesn't require everyone to stop at the same moment. The right split is: use live time for the things that genuinely need it, and accept slightly slower resolution on everything else in exchange for uninterrupted work.
What if my team prefers meetings?
Often that preference is really about clarity or connection, not format. People ask for meetings when they're unsure whether they're allowed to decide, or when writing in your team is so poor that a meeting is the only reliable way to find out what's happening. Fix the underlying cause and the demand drops on its own.
Should a decision meeting have a pre-read every time?
For anything with real detail, yes. The pre-read is what stops the first twenty minutes being a briefing. If people habitually arrive without reading it, that's a signal the document is too long or the meeting was never necessary.
How short can a meeting be?
Shorter than the default. Meetings expand to fill whatever length you book, so booking 25 or 45 minutes instead of 30 or 60 usually loses nothing and hands the day back a usable gap. A focused single-decision meeting rarely needs more than a quarter of an hour.
Start with the next invite
You don't need a meetings policy. Take the next meeting you're about to schedule and run it through the four criteria: how ambiguous is this, how emotionally loaded, how reversible, how many people need to genuinely align? If it scores low on all four, write the message instead and watch what happens — which is usually nothing, which is the point. For more practical systems that keep a manager's week, hand-offs, and calendar under control, explore You Manage It.
