How to Run a Meeting That Actually Produces Decisions

The single most consistent complaint inside any growing company is about meetings. There are too many of them, they take too long, they don't produce decisions, the wrong people are in them, the right people aren't, and they keep happening anyway because nobody knows how to fix them.

The fix is mostly mechanical. Once you know the principles of a good meeting, running one isn't hard. The hard part is the willingness to actually apply the principles, especially when meetings have become a cultural habit that nobody questions.

Here's how to run meetings that produce decisions instead of consuming time.

1. Know what kind of meeting you're running.

The first failure of most meetings is that they're trying to do multiple things at once and doing all of them badly. There are really only four kinds of meetings, and each one needs different design.

• Decision meetings. The goal is to make a specific decision. These are the most valuable meetings and the ones most often run badly. They require pre-reads, clear options, and explicit decision-makers.
• Information meetings. The goal is to share information that requires discussion. These should be rare; most information sharing should happen asynchronously (written updates, video walkthroughs, etc.) and a meeting should only be used when there's genuine need for back-and-forth.
• Working meetings. The goal is to actually do work together — design something, edit something, plan something. These need clear inputs, dedicated time, and the right people in the room.
• Relationship meetings. The goal is to build or maintain a working relationship. One-on-ones, informal team time, occasional dinners. These are not productivity meetings and shouldn't be designed as if they were.

A meeting that's trying to be three of these at once is a meeting that won't accomplish any of them well. Pick one. Design for it.

2. Send pre-reads. Demand that they be read.

The single most effective change you can make to your meetings is to send a pre-read 24 hours in advance and require attendees to read it before showing up. The pre-read should contain the context, the question being decided, the options on the table, and the data relevant to the decision.

Two things happen when you do this. First, the meeting itself gets dramatically shorter, because you're not spending the first twenty minutes catching everyone up. Second, the quality of discussion improves, because everyone arrives with thought-out positions instead of reacting in real-time.

The Amazon meeting culture famously requires this — six-page memos read silently at the start of the meeting, before discussion. You don't have to copy the format. But the principle — that the meeting starts with everyone having absorbed the same context — is one of the few genuinely transformative meeting practices.

3. Have an agenda with timeboxes and a designated decision-maker.

The agenda should specify, for each item: what's being discussed, how long it gets, and who has the final call. The "who has the final call" part is what's missing from most meetings. Without it, meetings turn into discussions that drift, find rough consensus, and produce no clear decision.

A clear designated decision-maker doesn't mean others can't disagree. It means there's someone whose job is to listen to the discussion, weigh the inputs, and make the call. This is dramatically more productive than collective decision-making by exhaustion.

4. Keep the room small.

The single biggest cost driver in meetings is the number of people in them. A meeting with 4 people is a working session. A meeting with 12 people is a town hall. Most meetings drift toward 8-12 attendees because nobody wants to be excluded, and most meetings would be twice as productive at half the size.

Two questions to ask before adding someone to a meeting: do they need to participate (not just observe), and would the meeting be meaningfully worse without them? If either answer is no, they shouldn't be in the meeting. Send them the notes afterward.

5. End with explicit decisions and owners.

The last five minutes of any meeting should be: what did we decide, who's doing what, by when, and how will we know it got done. This sounds obvious. It almost never happens. Most meetings end with a vague "great discussion" and a calendar invite for a follow-up.

The discipline of ending with explicit decisions and owners is what separates meetings that produce results from meetings that produce more meetings. Make this a habit. Write the decisions down. Send them in a follow-up email within an hour. Hold people to them next week.

6. Cancel the meetings that shouldn't exist.

Audit your recurring meetings every quarter. For each one, ask: is this meeting still serving its original purpose? Are the attendees the right ones for the current state of the work? Could it be a written update instead?

Most companies accumulate recurring meetings that have outlived their usefulness. Cancelling them is the highest-leverage productivity move you can make. This is especially true for status meetings — most of which can and should be replaced by written updates that are read asynchronously, with meetings reserved for when discussion is actually needed.

7. Default to no.

The single most undervalued meeting skill is the willingness to decline meetings that don't need you specifically. Most people accept all meeting invitations because saying no feels rude. As a result, calendars fill with meetings that don't require the recipient's input, and the recipient's actual work suffers.

The healthier default is to ask, before accepting any meeting: do I need to be there? If the answer is "I could be there but I don't strictly need to be," decline politely. Send the relevant input asynchronously. Spend the time on the work that actually requires you.

The deeper principle. Meetings are a tool, and like any tool, they're useful when applied to the right problem and harmful when applied to the wrong one. The companies that run meetings well treat them as one of several tools (alongside written communication, asynchronous updates, working sessions, and direct individual work) and use the tool only when it's the right one for the job.

The companies that run meetings badly treat them as the default tool for everything and end up with a culture where the calendar consumes the work. The first kind of company can move five times faster than the second, on the same team and the same problem.

If meetings feel like the bottleneck in your company, they probably are — and the fix is design, not avoidance. Run fewer meetings, but run the ones you do run with discipline. The output difference is enormous.