The cost side is arithmetic. Eight people at an average 75,000 salary, loaded at 1.3, cost about 375 for an hour — and a weekly repeat of that is roughly 18,000 a year. That is the price on the table.
The value side is a set of questions, and the first one decides most cases.
| Question | If no |
|---|---|
| Can you write the outcome in one sentence — what will be different afterwards? | Cancel. Not ready. |
| Is there a decision to make, or ambiguity to resolve? | Send a document instead. |
| Will the person who can actually decide be in the room? | Postpone. You will meet twice. |
| Does everyone have what they need to contribute? | Send the pre-read, move the meeting. |
| Does every invitee affect the outcome? | Cut the list. Cost scales linearly. |
| Would async genuinely be slower? | Do it async. |
The first row alone eliminates a large share of proposed meetings, because a meeting whose purpose cannot be stated as a change in the world is a status update wearing a costume.
Anti-meeting arguments almost always ignore this, and it is the strongest case for meeting. Async work has a latency cost, and for anything requiring back-and-forth it compounds:
| Route | Round trips | Elapsed time | Direct cost |
|---|---|---|---|
| 20-minute meeting, 4 people | 1 | 20 minutes | ~62 |
| Thread, 4-hour reply latency | 6 | ~3 working days | ~40 in reading and writing |
The thread is cheaper in salary and vastly more expensive in elapsed time. If four people are blocked while it resolves, three days of delay dwarfs 62 of meeting cost. So the honest rule is not that meetings are wasteful — it is that meetings buy latency reduction, and you should only pay when latency is what is hurting you.
Text is genuinely bad at several things, and pretending otherwise produces its own failures:
This test produces a prompt, not a verdict, and its inputs are estimates — you rarely know colleagues salaries, and the loading factor varies widely by country. More importantly, cost is not waste. A one-hour meeting that prevents a month of misdirected work is exceptional value at 375, and no scoring rubric can see that in advance.
There is also an organisational reality the arithmetic ignores: declining a meeting has political cost, and in some cultures attendance is the signal rather than the content. A framework that tells you to cancel your manager weekly is technically correct and practically useless. The realistic application is to your own invitations, where you control the decision.
Finally, be wary of overcorrecting. Organisations that eliminate meetings aggressively often discover they have simply moved the coordination cost into an invisible form — slower decisions, more rework, and people who no longer know each other well enough to disagree productively.
Shorten the invite list and share notes with everyone else. Cost is linear in attendees, so halving the room halves the cost immediately, while shortening a meeting often just compresses the same conversation.
Offer the alternative rather than the refusal. Asking whether the decision could be made in the thread, or offering to send written input and read the notes, is almost always accepted.
Yes. Setting an end date after eight or twelve occurrences forces a deliberate renewal, which is the only reliable way recurring meetings ever get cancelled.
No. Attendee counts and salary estimates stay in your browser and nothing is transmitted or saved.
Every tool comes with a written guide, and every category is one click away.