How teams run on far fewer meetings

The budget, not the ban

Teams that meet less are rarely the teams that banned meetings. Bans produce theatre: the meeting reappears as a “quick call”, a “sync”, or a thread of forty messages that would have been eight minutes of talking.

The teams that genuinely spend fewer hours in calls do something duller. They treat meeting time as a budget with a visible balance, and they make the default cost of a meeting slightly annoying to pay.

Defaults do most of the work

Every calendar tool ships a thirty-minute default, and almost every meeting therefore takes thirty minutes. Change the default to twenty and the same conversations finish in twenty. We have watched this happen across enough workspaces that it stopped being surprising.

Two related defaults matter as much:

The default attendee count should be small. A page that invites four people by habit spends four times the budget of one that invites one. Make the extra attendees an explicit choice.

The default recurrence should be none. A weekly meeting is a subscription, and like every subscription it outlives its usefulness by several months.

Write the pre-read, then decide

The most effective filter we have seen is a rule that a meeting needs a paragraph before it needs a slot. Not a document — a paragraph, in the booking request, answering what decision the meeting produces.

Roughly a third of requests do not survive the paragraph. The author writes it, realises the answer is already known, and sends the paragraph instead. That is not a meeting saved; that is a decision made a day earlier.

Measuring the budget honestly

Teams usually measure meetings by counting them, which is the least useful number available. Count hours, weight them by attendees, and split by whether the meeting produced a written outcome:

// Meeting load in person-hours, split by whether a decision was written down.
export function weeklyLoad(meetings: Meeting[]): Load {
  return meetings.reduce<Load>(
    (total, meeting) => {
      const hours = (meeting.minutes / 60) * meeting.attendees.length;
      const bucket = meeting.outcomeNote ? 'decided' : 'undecided';
      total[bucket] += hours;
      return total;
    },
    { decided: 0, undecided: 0 },
  );
}

The ratio between those two buckets is the number worth watching. A team with twelve person-hours of calls and eleven written outcomes is fine. A team with six person-hours and one outcome has a problem that removing meetings will not fix.

The standing no

Every team we spoke to that held its meeting budget for more than two quarters had one shared habit: someone was allowed to say no to a recurring meeting without explaining why, and that no was respected.

It is not a process. It is permission, and it is the part that no scheduling tool can install for you.

What tooling is actually for

Given all of the above, what is left for software? Three things: making the wanted meeting take ten seconds to arrange, making the unwanted one visible in a weekly summary, and making the written outcome the path of least resistance.

Everything else — the reminders, the time zones, the rota — is plumbing. Good plumbing, worth having, but not the reason a team gets its week back.