Guides · Guide
Make the Meetings You Keep Worth Having
Published
Cutting meetings is half the job. The other half is what happens in the ones you keep.
When decisions move out of meetings and into writing, the calendar gets lighter. What is left should be the meetings that genuinely need people in a room at the same time. Those deserve more care than they usually get, because they are now the expensive exception, not the background noise.
This guide covers which meetings survive the cut and how to run them so they are worth the hour.
Which meetings survive the cut
If you have applied the test from How to Have Fewer Meetings, most of your decision meetings are candidates to go async. What remains tends to fall into three groups.
Conflict that involves trust or emotion. When two people disagree in a way that touches how they work together, writing makes it worse. Tone disappears, positions harden, and every message gets read in the least generous way. A live conversation lets people hear each other, adjust, and repair. Keep these meetings.
News people need to hear together. A reorganization, a lost customer, a change in plans that affects people’s jobs. Some news should not arrive as a message read alone at 11pm. Delivering it live, with room for questions, is a form of respect.
Creative work that builds in real time. Some work genuinely benefits from people riffing on each other: early product shaping, naming, a hard design problem where the third idea only appears after the first two are said out loud. Not all creative work is like this, but some is.
Everything else deserves a second look. Status updates, approvals, and most decisions can move to writing. If a meeting does not fit one of the three groups above, ask what would be lost if it became a written question with a deadline.
Who belongs in the room
A meeting worth keeping still goes wrong if the wrong people are in it.
Invite lists grow by habit. Someone was added to a recurring meeting once, for a reason that no longer applies, and they are still there. Someone else is invited to stay in the loop, which is a job a written outcome does better and at a fraction of the cost.
Before each meeting you keep, go through the list and ask one question per person: what would this meeting lose if they were not here? There are only a few good answers. They hold information nobody else has. They own part of the outcome. They will be directly affected and need to be heard before it is decided. Or, for a hard conversation, they are one of the people it is about.
If the honest answer is “they would want to know what happened,” they need the outcome, not the meeting. Send it to them afterwards, in writing, as soon as it is decided.
A smaller room also changes the meeting itself. People speak more plainly when the audience is the people who need to hear it. Disagreement surfaces earlier. And the meeting ends sooner, because fewer people are waiting for a turn.
Run it like it matters
The person who called the meeting owns how it goes. That does not mean doing most of the talking. It means three things.
Start on time, even if someone is missing. Waiting teaches everyone that the start time is a suggestion, and the cost lands on the people who showed up.
Keep the room on the question. When a discussion drifts into a second topic, name it, write it down, and decide later whether it needs its own question and its own time. Most of the time it does not.
Leave room for the quieter people. In a live discussion, the first and loudest views tend to set the direction. Before you close an item, ask directly whether anyone sees it differently, and wait for an answer.
Agenda as questions, not topics
Most agendas are lists of topics. “Q3 roadmap.” “Hiring.” “Brand refresh.” A topic tells people what the meeting is about. It does not tell them what the meeting is for.
Write each agenda item as a question that can be answered instead.
“Brand refresh” is a topic. “Which concept should we take into the next design round?” is a question. The question tells everyone what they need to have thought about before they arrive, and it tells the room when the item is finished: when the question has an answer.
This small change does three things.
It forces whoever wrote the agenda to know what they want. If you cannot turn a topic into a question, the item is probably not ready for a meeting.
It exposes updates disguised as discussion. If the honest question is “Does anyone have questions about the hiring plan?”, that item is an update. Send it in writing and take questions there.
It gives you an order. Put first the questions that block the most people. If the meeting runs short, the important answers still happen.
Pre-reads
A meeting that starts with someone presenting background is spending everyone’s time on something each person could have read faster alone.
Send a short pre-read a day ahead. Short means something a person can read in about five minutes: the question, the context, the options, and what you need from them. If it needs more than a page, the meeting is probably trying to do too much.
Then open the meeting assuming everyone has read it. Do not present the pre-read aloud. This feels rude the first time and becomes normal fast. People learn that the pre-read is the meeting’s first half, and they read it.
Sometimes the pre-read answers the question on its own. People reply in writing, the answer is clear, and there is nothing left to discuss. When that happens, cancel the meeting. That is the pre-read doing its job, not a failure of planning.
Decision and owner before anyone leaves
The most common way a meeting fails is that it ends without anyone being sure what was decided.
Everyone leaves with a slightly different memory. A week later the same question comes back, because nothing was written down and nobody owned the next step.
Fix this with one habit: end each agenda item by saying the outcome out loud and naming one owner.
“We’re taking the line wave into the next round. Maya owns the next version.” One sentence, one name. If the room cannot produce that sentence, the item is not decided, and it is better to know that now than next week.
Write the outcome down in the meeting, not after it. Notes written later drift toward what the note-taker remembers. Notes written in the room can be corrected on the spot.
Then send the written outcome to everyone affected, including people who were not in the meeting. The outcome is what matters to them. The discussion mostly is not.
The 25-minute default
Calendar tools default to an hour. Meetings expand to fill it.
Change your default to 25 minutes. Extend only when there is a reason you can name: a hard conflict that needs room, a creative session that needs time to warm up.
Shorter slots do more than save time. They force the question to be framed before the meeting starts, because there is no time to figure out what the meeting is about once everyone is in the room. They make pre-reads necessary rather than nice to have. And they leave a few minutes between meetings, which protects people’s ability to think instead of running from one call into the next.
A 25-minute meeting with a clear question, a pre-read, and an outcome named before the end will usually accomplish more than an hour-long meeting without them.
Putting it together
The meetings you keep should be the ones that need to be live. Give each one a question instead of a topic, a short pre-read, a written outcome with an owner, and a shorter slot.
None of this requires new tools. It requires treating meeting time as scarce, which it is.
If you are still deciding which meetings to keep, start with How to Know If You’re Having Too Many Meetings for a self-audit, and What Causes Too Many Meetings for why the calendar filled up in the first place.