Getting out of one standing meeting is a personal manoeuvre and the exit message covers it. This is the other thing: reviewing the whole recurring set on a schedule, as a team, so that meetings stop accumulating by default.
They accumulate because creating a standing meeting takes four minutes and ending one requires somebody to volunteer an opinion about somebody else’s work. The asymmetry is the entire mechanism. No individual meeting is the problem; the absence of any moment at which the set gets questioned is.
A quarterly review fixes that by making the questioning scheduled rather than personal. Nobody is criticising your meeting. It is October, and in October we look at all of them.
The review in forty-five minutes
Once a quarter, one slot, the whole recurring set on one page. It is itself a meeting, which is worth acknowledging directly — this is the one recurring meeting whose job is to reduce the others, and if it stops earning its place you kill it too.
Four steps.
1. Build the list (before the meeting, 20 minutes, one person). Every recurring meeting the team holds or attends. For each: name, frequency, duration, headcount, organiser, and the date it was created. Calendar exports give you most of this; the creation date is the one you may have to dig for, and it is the most useful column on the page.
2. Compute the hours. Frequency × duration × headcount, per quarter. One column, one number per meeting, one total at the bottom. Do not convert to money — the cost arithmetic explains why the dollar figure argues badly even when it is correct. Hours per quarter is a number people accept; a salary-weighted figure reads as an accusation and the conversation becomes about the multiplier.
3. Sort each one into four outcomes. Keep, shrink, convert, sunset. That is the entire vocabulary and the constraint is deliberate — “revisit next quarter” is not on the list, because it is how everything survives forever.
4. Write the decisions where the meetings live. Each outcome gets an owner and a date. A decision recorded only in the review’s own notes has not happened.
The four outcomes
| Outcome | Means | Who does what |
|---|---|---|
| Keep | Unchanged, with a stated reason | Owner confirms they still own it |
| Shrink | Shorter, fewer people, or less often | Owner edits the invite this week |
| Convert | Becomes a written artifact | Someone commits to writing it |
| Sunset | Ends on a named date | Owner cancels; nobody needs to defend it |
Which one a meeting gets is a judgement, but it is not an unstructured one — the four diagnostic questions in the redesign guide sort a meeting into exactly these outcomes, and running them in the review is faster than arguing from impressions.
Two notes on the outcomes people avoid.
Shrink is underused and it is the best value on the table. Halving a weekly hour with nine people returns more hours than cancelling a fortnightly half-hour with four, and nobody has to concede anything. Most reviews should produce more shrinks than sunsets.
Convert is not free and the review should price it. Moving a status meeting into a written update means someone writes it every week. Name that person in the review or the conversion does not happen — the meeting quietly continues, which is the most common way a review produces nothing.
The sunset clause, which is the part that actually works
Everything above is a review of existing meetings. The sunset clause stops you needing it so much.
Every new recurring meeting gets an end date when it is created. Not a review date — an end date. Three months, six months, whatever suits the thing it is for. On that date it stops, and if it should continue, someone renews it.
The mechanics take about ten seconds: set a real end date on the recurrence in the calendar, and put the reason the meeting exists in the invite body.
That is the whole intervention, and it inverts the asymmetry that created the problem. Today, continuing requires nothing and ending requires an argument. With a sunset, ending requires nothing and continuing requires someone to say the meeting is still worth it. The same people, the same meetings, the opposite default.
What it does in practice is kill the project meetings. A weekly set up around a launch has a natural end that nobody ever gets round to declaring; with a date on it, it ends on that date and everyone is mildly relieved. The meetings that survive renewal are the ones with an owner who will speak up for them, which is exactly the filter you wanted.
The reason line in the invite body matters as much as the date. Twelve months on, nobody remembers why the Tuesday call started, and the review has to reconstruct it from folklore. One sentence written at creation — “set up to coordinate the billing migration; ends when we’re off the old system” — answers the hardest of the four diagnostic questions for free.
Retrofitting it. You do not need permission to put an end date on meetings you organise. Do that this week for yours, and propose it as a rule for new meetings at the next review. A rule about future meetings is a much easier sell than a judgement about present ones, and within two quarters it is doing more work than the review is.
Running it when you do not run the team
The review is a manager’s tool as described. The version available to someone without that authority is narrower and still worth doing.
Run it on yourself. Your own recurring commitments, the same four columns, the same four outcomes, once a quarter. You control your attendance even where you do not control the meeting, and “shrink” and “convert” both have personal versions — attending for one item, sending written input. The ladder in the alternatives to declining is the full set of moves available from that seat.
Then bring the page, not the opinion. Your manager will not act on “we have too many meetings”. A one-page list with hours per quarter and a proposed outcome next to three of them is a different object entirely — it is work you have already done, and the ask is a yes rather than an investigation. Lead with the shrinks. They are cheap to approve and they establish that you are tidying rather than escaping.
Offer to build the list for the team. This is the cheapest route to the whole thing existing. Assembling the calendar export is unglamorous, nobody else wants to do it, and the person who does it ends up framing the conversation.
When a standing meeting should simply be kept
Three kinds, and a review that sunsets them is a review doing harm.
The one holding a relationship together. Some standing slots exist so two teams stay in contact. They fail every efficiency test because efficiency is not what they are for. Keep, with the reason written down honestly, so the next review does not re-litigate it.
The one that is somebody’s only access to a senior person. A skip-level or an open office hour looks like low-value airtime on an hours-per-quarter table and is not replaceable by a document.
The new one. A meeting set up six weeks ago has not had time to prove anything. Give it its full sunset period before judging it; reviewing a meeting to death before it has run its course teaches people not to propose meetings, and some meetings are the right answer.
Everything else gets one of the four outcomes and a date. The review’s only real enemy is the fifth outcome nobody admits to using — leaving it exactly as it is and agreeing to look again in three months, which is how the set got this big.