There is a whole category of web page that will multiply attendees by salaries and tell you a recurring meeting costs your company $47,000 a year. They are free, they are well built, and most of them exist to collect your email address.
The arithmetic is not wrong. The strategy is. Bringing a dollar figure to the person who owns a meeting is one of the least effective things you can do, and it fails in a predictable way that is worth understanding before you spend your credibility on it.
Here is the formula anyway, because you should know it. Then the two numbers that actually move a decision.
The meeting cost formula
Cost per occurrence = Σ (attendee hourly rate × fully-loaded multiplier × duration in hours)
Then multiply by occurrences per year.
Three things people get wrong:
Duration is not the meeting length. It is the meeting length plus preparation plus the recovery on the other side. A weekly review that runs an hour and requires an hour of deck-building from two people is not a one-hour meeting. If you are going to compute this at all, compute it honestly, because a deflated number is worse than none.
The fully-loaded multiplier. Salary is not what an employee costs. Payroll taxes, benefits, equipment, software, and facilities sit on top of it. Finance teams commonly work with a multiplier somewhere between about 1.25 and 1.4 on base salary, but the figure varies by country, benefits structure and how your organisation allocates overhead — it is an accounting convention, not a universal constant. If the number matters, ask your finance team what multiplier they use. If it does not matter enough to ask, it does not matter enough to put in a slide.
Attendee rates. You do not know your colleagues’ salaries and you should not attempt to. Use a band midpoint for the level, or use a single blended rate for everyone. Precision here is fake and pursuing it makes people uncomfortable for no gain.
Worked, for a weekly one-hour meeting with eight people at a blended £60/hour and a 1.3 multiplier, with 15 minutes of preparation each:
- Attendee time per occurrence: 8 × 1.25 hours = 10 hours
- Cost per occurrence: 10 × £60 × 1.3 = £780
- Occurrences per year, allowing for holidays and cancellations: ~45
- Annual: about £35,000
That is a real number, computed honestly. Now watch it fail.
Why the dollar figure does not persuade
You take £35,000 to the meeting’s owner. Four things happen, usually in this order.
The number is not spendable. Cancelling the meeting does not produce £35,000. Nobody gets a budget line back, no headcount is freed, no one can point at a saving on any report. Everyone in the room knows this, which makes the figure feel rhetorical rather than financial — and once a number reads as rhetoric, it argues against you.
It attacks the meeting’s existence, which means it attacks the owner. A cost argument says: this thing you created is destroying value. The owner is now defending themselves rather than evaluating the meeting, and people defending themselves do not cancel things. This is the main failure and it is almost always avoidable, because you rarely needed the meeting killed — you needed to not be in it.
The estimate becomes the argument. “Where did you get £60 an hour?” “It’s not really an hour of prep for everyone.” “We don’t count time that way.” Twenty minutes later you are litigating a multiplier instead of discussing a meeting, and you have lost, because the person who has to defend their assumptions is losing.
It implies the alternative is free. The counter-question is always the same: what does not having this meeting cost? If the answer is “a coordination failure we would find out about in three weeks”, then £35,000 might be cheap, and you have made your opponent’s case for them.
Cost arguments do work in one narrow place: setting a policy about meeting sizes or defaults, where nobody’s specific meeting is on trial and the number is doing illustrative work. That is a different conversation, usually with a different person, and it belongs to whoever owns team-level norms.
The two numbers that work instead
1. Headcount-weeks.
Take the same meeting and convert the annual attendee-hours into full-time weeks.
10 attendee-hours per occurrence × 45 occurrences = 450 hours. At a 37.5-hour week, that is 12 person-weeks a year. Roughly a quarter of one person’s year, spent in one recurring meeting.
Say it that way:
“This meeting is about twelve person-weeks a year across the eight of us. That’s a quarter of a headcount. I’m not saying we shouldn’t do it — I’m asking whether it’s worth a quarter of a headcount, because that’s the shape of it.”
This lands where the money does not, for a concrete reason: headcount is the currency managers actually operate in. They fight for it, they lose it in planning, they know exactly what a quarter of a person is worth. And unlike pounds, person-weeks are not obviously fake — they are just the hours, rearranged.
It also invites the right response. “Twelve person-weeks for this?” is a question the owner can answer with yes, and sometimes yes is correct. You have made the trade visible without making it a verdict.
2. What it displaces.
The strongest version is not a cost at all. It is the specific work that does not happen.
“The Thursday review sits in the middle of the only morning I get without meetings. The forecast rebuild needs about six hours of uninterrupted time and it’s been slipping for three weeks. If I come off this or it moves to the afternoon, that gets done.”
No estimates, no multiplier, nothing to argue with. It names a thing your manager already wants and a specific obstacle to it. And it offers a move rather than a cancellation, which as the fragmentation post makes clear is usually the cheaper ask and often the real fix.
When to compute the cost anyway
Three cases where the arithmetic is worth doing even though you will not lead with it.
For a meeting you own. No politics, no defensiveness, and the number tells you whether to cut the invite list — which is nearly always the highest-return change to a meeting, since cost scales with attendees and value usually does not.
For a policy conversation. When a leadership team is deciding whether to cap standing meetings or set a default duration, aggregate cost is legitimately useful. It is describing a system, not judging a person.
For your own decision. Before you spend social capital on getting out of something, work out what it is worth. Some meetings cost you 45 minutes a fortnight and the argument would cost more than the meeting does. Knowing when not to fight is most of what makes the fights you do pick work.
Cutting the invite list beats cutting the meeting
Cost scales linearly with attendees. Value does not — past a certain size a meeting’s output is set by three or four people and everyone else is present for information they could have read.
Which means the highest-return change available is almost never cancellation. It is going from eleven attendees to five. That change is cheap to propose, does not threaten the owner, and takes 55% of the cost out in one move.
“Does the whole group need to be in this, or could it be the four of us who have actions and the notes go out after? Happy to write the notes.”
Offering to write the notes is what makes it a contribution rather than a complaint. It costs you fifteen minutes and it makes the proposal impossible to read as trying to get out of something.
Common questions
Should I use one of the meeting cost calculators? For your own understanding, sure — the arithmetic is trivial and the calculators do it fine. Just do not paste the output into a message. The number’s problem is not its accuracy, it is what quoting it does to the conversation.
What if my company already tracks meeting cost? Then use their number, not yours. An internal figure with finance’s multiplier behind it cannot be argued with in the way that your estimate can, and half the failure mode disappears.
Isn’t it dishonest to hide the cost? You are not hiding it. You are converting it into a unit the other person can act on. Person-weeks and displaced work are the same hours as the pounds — they are the hours without the implication that someone has been wasting money, which is the part that stops people listening.
My manager loves numbers. Then bring both, with person-weeks first and the money as a footnote you can produce if asked. Leading with the money still invites the assumption argument; leading with the trade does not.