The Real Cost of Meetings (and How to Reclaim the Week)

It is 10:02 on Monday. Thirty-four people sit muted in a video room while two directors read bullet points from slides they assembled the evening before. Most of the room is skimming Slack, waiting for the one slide that touches their team. Nobody decided to spend the hour this way. The meeting is on the calendar because it has always been on the calendar.

The question worth asking is not whether the meeting is annoying, which is a matter of taste. It is what the meeting costs, how much of that cost we can actually measure, and whether the organisation is getting the value back. Most of the numbers people quote about meetings mix those things together, which is why they rarely survive contact with a finance team.

Four things about a recurring meeting are measurable rather than estimated: how many people attend, how long it runs, how often it repeats, and the fully loaded hourly cost of the people in the room. Multiply them and you have a figure you can put on a slide without flinching.

As a worked example with round numbers: forty attendees, one hour, at an assumed fully loaded cost of £60 per person per hour, is £2,400 per occurrence, or roughly £115,000 across forty-eight working weeks. Only the £60 is an assumption there, and your finance team can replace it with a real average. Everything else comes off the calendar invite.

That total is not new spending. The salaries are already committed, so the hour becomes a real cost only to the extent that it displaces work of higher value. It is still worth calculating, because it converts a vague sense of too many meetings into a number a leadership team can argue about.

A one-hour meeting rarely occupies only sixty minutes. People stop what they are doing before it starts, and take some time to get back into concentrated work afterwards. Preparation happens somewhere too, usually on the calendar of whoever builds the material.

We do not know how large that surrounding cost is for your team, and neither does anyone else who has not measured it. So do not pick a figure that sounds credible. Model it instead: run the same calculation assuming 10, 20 and 30 minutes of disruption per attendee, and present all three. If the decision changes between the low and the high scenario, you have learned that the honest answer is you need to measure it. If the decision is the same in every scenario, the uncertainty did not matter and you can stop arguing about it.

Keep the two layers visually separate when you present them. A measured number and a modelled number added into one authoritative total is the fastest way to lose the room, because the first person who disputes your context-switching assumption gets to dismiss the whole thing.

In a 2024 survey of knowledge workers commissioned by Atlassian, most respondents said the majority of their meetings are ineffective. That is self-reported perception rather than a measurement of what the meetings produced. It tells you how the hour feels from the inside, which is worth knowing, and it does not tell you the hour produced nothing.

Microsoft's 2023 Work Trend Index, based on usage data from Microsoft 365 and its own survey work, reported that a large share of the working day went to communication — meetings, email and chat — rather than to creating. That was 2023, and it describes a population of Microsoft 365 users rather than your company. For a support lead or an account manager, communication is the work, and a high share is not a defect.

Harvard Business Review reported in 2017 that senior managers spent close to 23 hours a week in meetings, against under ten in the 1960s. It is old data about a specific population, and it is still useful for one thing: meeting load has a long history of growing without anyone deciding to grow it.

Taken together, these findings suggest that meeting time is large, disliked and rarely chosen deliberately. They do not establish what that time is spent on. Our interpretation, which is a hypothesis rather than a finding, is that in organisations where the load grew by accident, a substantial part of the hour goes to transferring information that could have been read. That is worth testing on your own calendar before believing it, and the next section is how to test it.

Moving information out of the room assumes people will read what replaces it. In teams with no habit of writing, or where several people are working in a second language and a dense written update is slower to absorb than a spoken one, that assumption does not hold, and the cost simply moves somewhere you are not counting it.

It also misleads for meetings whose purpose is not throughput. For a distributed company, a weekly start is often the only moment everyone is in the same place at the same time. That has a value the spreadsheet does not contain, and optimising it towards zero is a decision, not an efficiency.

So calculate the number, but do not let it decide alone. The goal is a meeting worth its price, not the cheapest possible meeting.

Shortening a meeting is easy to do once and hard to sustain, because the material that filled the hour still has to go somewhere. Moving the preparation is the change that usually survives: the week's developments are collected before Monday, the people closest to each one write their own two lines, and the meeting opens on the first decision rather than the first status update.

That preparation is real work, and without a structure it lands on whoever cares most. Curact exists to hold that structure — the week's material gathered, the right people asked to write their own part, the meeting starting on what actually changed. The arithmetic above works the same whether or not you use it.