If only 2 or 3 people are meeting, it is not a meeting. You are just scheduling time for a chat. No need to go overboard with prep, but it might help if the person knows why you want to talk to them first - unless you just want to build a relationship. 1-on-1 should be used for building relationships, for feedback, for any difficult conversations or actions that you parked from a bigger meeting. If it were up to me, all meetings would be this kind. This is where the real work happens. This or a workshop, but the most successful workshops start with a few 2-3 people meetings anyhow.
If 4 or more people are meeting, you have yourself a proper meeting and you should sign-post it as such with an agenda. How strict you need to be on the agenda depends on which of the four types of meetings you are running:
“Doing the doing” when this happens to need a diverse group of people to make it happen. Make sure everyone there is needed and knows what they are talking about. “Hangers-on” can severely disrupt the flow and creativity of the rest. Cut people ruthlessly and manage with a smaller number than required. You can always take your findings to people outside the core group afterwards if needed. This kind of work usually benefits from large amounts of unstructured time, like a half or full day. You may want to break it down into a couple of chunks to make sure you are making progress, but don’t go overboard on the timings. Go with the flow.
Cancel this meeting if the person needing a decision has not shared why the decision is needed, the options and estimated consequences of each option, and their recommendation. If you are the person needing the decision, make sure you have spoken to the key people beforehand. Like House of Cards, never call a vote unless you are sure you are going to win. You need to know who is on-side or not so you can prep everyone else. This kind of meeting, when done well is just acting out what everyone already knows. Time it ruthlessly. Allow some time to go through each option (as anyone you have not spoken to will never have read a pre-read), some time for discussion, and then push for a decision when the time is up. If you can’t get the meeting to a decision, then you either have a bunch of people that use “more info” as an excuse or you have not done enough prep to show that a decision is really required and that it is needed now.
Relationship building / development on scale:
For example, team meetings or anything where the main aim is to get people working together. This is fine. Introduce some structure and timing to keep things on track, but be upfront that the point of this is to work together better or give people the safe space for development e.g. running a team meeting or presenting.
People not being aware of which type of meeting they are in. It can be okay to break a bigger meeting into parts that may cross the four types but avoid it where possible. The biggest source of frustration for attendees is confusion over what sort of meeting they were in. For example, you don’t want lots of discussion at a deliverables update; You don’t want people pushing to record actions in the middle of a workshop; You don’t want people trying to solve the world in a cascade; And you don’t want people telling their life story in a decision meeting.
Similarly, no one likes a time-Nazi if the group is just about to get the result they are after and someone cuts them off for the sake of a couple of minutes. Let it run over and then readjust. The key is to be able to do this quickly and have structured your meeting so you can run over. E.g. don’t ask for a decision right at the end when everyone needs to leave for another meeting. The decision should be made 10mins before the end. Buffer it with AOB if needed.
You can do all this with a clock, but seeing the time tick down and communicating this will increase the effectiveness of your meetings massively. Having the ability to proportion delays across the rest of your agenda instantly, so you always have a plan to finish on time is something your co-workers will be eternally grateful for. Okay maybe not, but they might buy you a coffee. MeeTime does all these things and for less than the price of a clock. Google Ventures uses a time timer at $25-£50 dollars. For a tenth of the price, you can have something more sophisticated always available in your pocket. Shameless plug over. Get MeeTime.