How Many Meeting Rooms Per Employee? A Planning Method

Walk through most offices at 10:30 on a Wednesday and you will see the same thing: every wall-mounted display shows red, two people are standing in the hallway waiting for a room to free up, and someone is taking a video call from a phone booth that was never meant for a 45-minute planning session. Then walk through the same office on a Friday afternoon and half the rooms sit empty. The building did not change. The demand did.

That gap is why the question "how many meeting rooms per employee" does not have a clean answer. A common rule of thumb is roughly one meeting room per 10 employees, and that is a fine place to start. But the number that actually matters is not your total headcount. It is your peak in-office headcount, because hybrid work concentrates people onto the same two or three days. Plan for the average and you will be short exactly when everyone needs a room. This post gives you a meeting room ratio to start from, then a method to adjust it to your building.

Why "conference rooms per employee" is the wrong unit

The instinct is to take your total staff, divide by some number, and call it done. A 100-person company at one room per 10 employees needs 10 rooms. Simple. It is also wrong for most hybrid offices, for one reason: not everyone is in on the same day.

If your 100 people average three days a week in the office, you do not have 100 people to plan for. You have whatever your busiest day looks like, and that busiest day is almost never an even slice. It clusters. So the useful unit is not conference rooms per employee across your whole payroll. It is meeting rooms sized against the most people you ever have in the building at once.

The midweek mountain

Attendance in hybrid offices is not flat across the week. It forms a shape people have started calling the midweek mountain: Tuesday and Wednesday run hot, Monday and Thursday are moderate, and Friday drops off a cliff. It is common for a midweek peak day to have two or three times the in-office headcount of a Friday.

This matters for room planning because your rooms get consumed on the mountain, not on the average. If you size for a typical day, Tuesday and Wednesday will feel broken (people circling for space, meetings bumped, calls taken at desks) even though your spreadsheet says you have "enough" rooms. Pull a few weeks of badge data or calendar data and find your real peak. That number, not your headcount, is what you plan against. Our 2026 utilization benchmarks go deeper on reading these patterns.

A meeting room ratio worth starting from

Once you know your peak in-office headcount, apply a ratio. One meeting room per 10 people at peak is a reasonable starting point for a collaboration-heavy office. Sales and consulting teams that live in calls may want to plan closer to one per 8. Heads-down engineering or design teams that meet less can stretch toward one per 12 to 15.

The ratio is a starting estimate, not a target you defend to the death. The real check is utilization. Global average meeting room utilization sits around 38 to 43 percent, according to workplace sensor data from firms like XY Sense and other industry benchmarks. That is low, and it is a trap: a low average usually hides rooms that are jammed at peak and dead the rest of the time. A healthy, well-planned room stock tends to run 60 to 75 percent utilization during core hours. If your average is far below that and people still complain about not finding rooms, your problem is almost always the wrong mix of room sizes, not the total count.

Split the rooms by size, and weight toward small

This is the step most offices get wrong, and it is the most expensive mistake to fix later. The classic error is to build one or two large, impressive boardrooms and a handful of medium rooms, then wonder why people cannot find space. The data on how meetings actually happen tells a different story: about 68 percent of meetings involve only one or two people. Most of what happens in your rooms is a one-on-one, a quick sync, or a single person taking a video call in private.

So the split should lean heavily toward small. A workable default breakdown of your total rooms:

Small (1 to 4 people): roughly 60 to 70 percent of your rooms. These absorb one-on-ones, calls, and quick syncs. You almost cannot build too many.
Medium (5 to 8 people): roughly 20 to 30 percent. Team standups, project meetings, small workshops.
Large (9 or more): roughly 10 percent, often just one or two rooms total. All-hands, big client presentations, workshops.

The failure mode to avoid: over-investing in big boardrooms and under-investing in 2-to-4-person rooms. A boardroom that seats 16 and gets used twice a week is a very costly way to hold a two-person call. If you are choosing between one more large room and three more small ones, the small ones will almost always serve more meetings.

A worked example: 60-person hybrid office

Say you have 60 employees, hybrid, averaging three days a week in the office.

Find peak, not total. You pull four weeks of badge data. Tuesday and Wednesday peak at about 42 people in the building. Friday bottoms out near 15. You plan against 42, not 60.
Apply the ratio. At one room per 10 at peak, 42 people gives you about 4 rooms. This is a collaboration-heavy team, so you round up and plan for 5.
Split by size. Of those 5 rooms: 3 small (1 to 4 people), 1 medium (5 to 8), 1 large (9-plus). If you can only afford 4 rooms, drop the large one before you drop a small one. Take the occasional all-hands offsite or in the largest room you have.

Notice what this avoids. Planning against 60 total headcount would have suggested 6 rooms, over-built for a day that never comes. And the naive instinct to build "a proper boardroom plus a couple of meeting rooms" would have left the midweek crowd fighting over the two small spaces that see the most real demand.

Then measure and adjust

Any plan is a hypothesis. The number that tells you whether you got it right is per-room utilization by size class, measured over a few weeks. If your small rooms run 80 percent at peak and your large room runs 15 percent, convert. If everything is jammed and your average utilization is above 70 percent, you are genuinely short and need more rooms or staggered in-office days. A ghost meeting (a room booked and never used) will also quietly eat 20 to 30 percent of your apparent capacity, so auto-releasing no-shows can recover rooms without building any.

This is the honest buy-side note. If you want the utilization data to plan and re-plan with, a low-power e-ink door display like Lobby reads your Microsoft 365 or Google calendars, shows whether a room is free at the door, lets people check in, and auto-releases rooms when nobody shows. That gives you real occupancy signal by room and by size class, which is exactly what you need for the "measure and adjust" step above. It will not, on its own, tell you your peak badge headcount (that comes from your access system), and it will not solve a genuine shortage of physical rooms. It tells you honestly how the rooms you have are actually used.

TL;DR

There is no universal "how many conference rooms" number. Start from a meeting room ratio of about one room per 10 people, but apply it to peak in-office headcount, not total headcount.
Hybrid concentrates people onto a midweek mountain (Tuesday and Wednesday can run 2 to 3x a Friday). Plan against your busiest day.
Split rooms heavily toward small: roughly 60 to 70 percent small (1 to 4 people), 20 to 30 percent medium, 10 percent large. About 68 percent of meetings are 1 to 2 people.
Do not over-build boardrooms. The expensive mistake is too few 2-to-4-person rooms.
Global average room utilization is about 38 to 43 percent; healthy is 60 to 75 percent at core hours. Measure per-room by size, then adjust.

Sources

XY Sense, workplace occupancy and meeting room utilization sensor data.
Leesman, workplace experience research on office attendance patterns.
U.S. General Services Administration, space utilization and workplace planning guidance.

How Many Meeting Rooms Per Employee? (A Planning Method, Not a Magic Number)