How many office pods per employee? What a 50-person office actually needs
Ask how many office pods per employee you need and you will be handed a ratio — one booth per ten people, one per twenty, pick a number. Every one of those is a guess with a colon in it, because the demand a pod serves has almost nothing to do with headcount and almost everything to do with how many people are in the building at once and how much of their day is spent talking. For a 50-desk office the honest answer lands somewhere between three and seven booths plus around two enclosed rooms, and which end you sit at is decided by your call culture, not your payroll. Below is the arithmetic that tells you which.
The short answer, by office size
Ranges, not numbers, because the spread between a quiet office and a phone-heavy one is wider than the spread between office sizes. The method that produces these is set out further down, and you should run it on your own inputs rather than lifting the row.
| Office | 1-person booths | Enclosed rooms | Works out at |
|---|---|---|---|
| 25 desks | 2 – 4 | 1 | 1 booth per 6 – 12 desks |
| 50 desks | 3 – 7 | 2 | 1 booth per 7 – 17 desks |
| 100 desks | 4 – 12 | 3 | 1 booth per 8 – 25 desks |
| 250 desks | 8 – 25 | 7 | 1 booth per 10 – 31 desks |
Two things to read off this before anything else. The ratio gets better as the office gets bigger — a larger pool absorbs its own peaks, so you buy proportionally fewer booths per head. And the enclosed-room count is the shortfall on top of the meeting rooms you already have, not a total.
The ratios most offices get wrong
There is no published industry figure for this, and you should be wary of anyone who quotes you one as though there were. We do not have one either — what follows is arithmetic you run on your own numbers, not a benchmark we are asking you to trust. Three assumptions break most pod plans before the first unit is ordered.
Headcount is not attendance
A hybrid office of 50 people is rarely a building containing 50 people. What matters is the peak — the busiest hour of the busiest day, which in most hybrid patterns is a midweek morning. Size against that peak, because a booth that is free on a Friday and queued on a Tuesday reads to everyone as not enough booths. Your door-entry or desk-booking data already has this number and it is the single most valuable input you have.
Daily demand is not concurrent demand
Thirty calls a day across an office does not need thirty booths, or even ten. It needs enough booths to cover how many of those calls overlap. Concurrency is the only quantity that matters, and it is a product of how often calls happen and how long they last — two numbers that are easy to estimate and that nobody collects. Counting calls instead of concurrency is how offices end up buying for a demand that never exists at one moment.
A booth and a room are not interchangeable units
An office phone booth ratio tells you nothing about how many enclosed meeting rooms you need, because the two serve different demand. Booths absorb ad-hoc, unscheduled, short demand — someone needs privacy in the next thirty seconds. Rooms absorb scheduled, longer, multi-person demand that appears in a calendar days ahead. Adding booths does not relieve pressure on rooms and adding rooms does not stop the call queue. They are two separate sums.
And a ratio that never changes is wrong by definition
Whatever you buy first is a hypothesis. Attendance patterns move, teams change shape, and a sales hire shifts call volume more than twenty new desks do. The offices that get this right treat the first phase as a measurement instrument and size the second phase from what it tells them — which is the subject of the last two sections here.
Booth demand as a function of call culture and hybrid peaks
Booth demand is a queueing problem, and the useful thing about queueing problems is that the core relationship is arithmetic rather than opinion. The average number of booths in use at any moment is simply how often calls start multiplied by how long they last. Everything else is deciding how much headroom to put on top.
| Input | What it is | Where to get it |
|---|---|---|
| P | People on site at the peak hour — not headcount | Door entry, desk booking, or a manual count over two weeks |
| c | Calls per person per day that genuinely need privacy — not every call, only the ones someone would get up for | Ask three teams with different roles; the spread between them is the point |
| t | Average length of one of those calls, in hours | Calendar or telephony reporting; otherwise estimate and refine later |
| H | Hours in the working day over which that demand spreads | Usually 8, less if your office has a hard core-hours pattern |
Step one: average concurrent demand
Multiply the peak population by the calls each person makes and by how long each one takes, then divide by the length of the day: A = P × c × t ÷ H. That gives the average number of booths occupied at any given moment. It is a floor, not an answer — a fleet sized to its own average is busy all day and queued half of it.
Step two: headroom, because demand is lumpy
Calls do not arrive evenly. They cluster on the hour and the half hour, after stand-ups, and in the window before lunch. Queues form at well under full occupancy, so plan the fleet to run at roughly half utilisation in a small office: booths = A ÷ 0.5, rounded up. Below that headroom people stop walking to the booth because they expect it to be taken, and the utilisation figure you later measure looks healthy for the wrong reason.
Step three: let bigger fleets run hotter
Headroom is proportionally cheaper at scale. With three booths, one person overrunning takes out a third of your capacity; with fifteen, it is noise. So the target utilisation can climb with the size of the pool — around 0.5 for a handful of booths, closer to 0.7 once there are a dozen or more on one floor. This is why the ratio improves with office size, and it is the one genuinely good piece of news in fleet planning.
Call culture moves the answer more than office size does
Run the sum for a design studio where people are mostly heads-down and for a sales floor on consecutive calls, and the two answers differ by a factor of three or four at the same headcount. That is the whole reason a single industry ratio cannot work. If your office is a mix — most are — run the sum per team and add the results, rather than averaging the behaviour of people who behave nothing alike.
Worked examples: 25, 50, 100 and 250 desks
The inputs below are illustrative placeholders, chosen to show the arithmetic working rather than to describe any real office — including ours. Peak attendance is taken at 70% of desks, calls average 24 minutes, and the day is 8 hours. The two columns differ only in call culture: a quiet office at 0.6 private calls per person per day, a phone-heavy one at 2.0. Substitute your own four numbers and the table rebuilds itself.
| Desks | Peak on site | Quiet office — concurrent / booths | Phone-heavy — concurrent / booths |
|---|---|---|---|
| 25 | ~17 | 0.5 → 2 booths | 1.8 → 4 booths |
| 50 | ~35 | 1.1 → 3 booths | 3.5 → 7 booths |
| 100 | ~70 | 2.1 → 4 booths | 7.0 → 12 booths |
| 250 | ~175 | 5.3 → 8 booths | 17.5 → 25 booths |
The 50-desk office, step by step
Fifty desks at 70% attendance is 35 people on site at the peak. In the phone-heavy case: 35 people × 2.0 calls × 0.4 hours ÷ 8 hours = 3.5 booths occupied on average. Divide by a 0.5 target utilisation and you need 7. In the quiet case the same office needs 3. Same building, same payroll, more than double the fleet — which is why the question "how many pods for 50 people" has no answer until somebody asks what those fifty people do all day.
Why 250 desks does not need ten times what 25 does
Concurrent demand scales linearly with people, but the headroom on top does not. At 25 desks you are buying two booths to cover half a booth of average demand, because you cannot buy a fraction and because one occupied booth out of two is a coin flip. At 250 the pool smooths itself and the same quality of service costs proportionally less. The practical consequence: small offices should expect a worse ratio than the figure a larger neighbour quotes them, and there is nothing wrong with their plan.
Where the sum stops being reliable
It assumes demand spreads over the working day. Two patterns break that. A hard core-hours culture compresses the same calls into fewer hours, so reduce H and the fleet grows. And a single large team on a synchronised rhythm — everyone off a stand-up and onto client calls at the same moment — produces a spike no average captures. If that describes you, size on the spike hour alone rather than the day.
Mixing 1-person booths with 4- and 6-person rooms
The classic failure is not buying too many pods or too few. It is buying six single booths when the floor needed two four-person rooms — a fleet that is the right size and the wrong shape. It happens because booths are cheaper per unit, easier to place and simpler to justify, so a budget under pressure quietly converts rooms into booths and nobody notices until the booths sit empty while people still cannot find anywhere to hold a meeting.
Rooms are a scheduled-demand sum, not a call sum
The meeting room ratio hybrid office planners need is driven by calendar load: how many meetings a week genuinely need an enclosed space, multiplied by their average length, spread over the bookable hours in a day. Divide by a target utilisation of around 0.65 — rooms tolerate a little more pressure than booths because people book them ahead rather than walking up — and round up. On the same peak-attendance basis as above, and taking the illustrative figures of 1.5 such meetings per person per week at 45 minutes each, that is roughly one room at 25 desks, two at 50, three at 100 and seven at 250. Those two inputs are the ones to replace first: they vary more between organisations than anything on the booth side.
Count the shortfall, not the requirement
Almost every office already has some enclosed space. The pod number is what is missing after you subtract the rooms you have — and after you check that those rooms are the right size. A floor with one twelve-person boardroom and nothing else has a real shortfall even though its total room count is not zero, because two people will not book the boardroom for a 1:1 and will hold it at a desk instead.
Hybrid pushed demand downward in size, not upward
The meetings that moved into the building are small ones with a remote half — two or three people in the room, two or three on a screen. That is four-person pod demand, and it is the size most floors are shortest of, because the rooms they built were sized for a pre-hybrid world where everyone attending was physically present. Large rooms sitting half empty next to a queue for anything small is the signature of this, and more booths will not fix it.
A sanity check before you commit the budget
Take your planned fleet and ask what it does for the three things people actually complain about: nowhere to take a call, nowhere to hold a 1:1, nowhere to run a small hybrid meeting. If every unit in the plan answers the first one, the shape is wrong whatever the total says. Our size guide works through which activity needs which size, and shop by size shows what each one costs to acquire.
Measuring utilisation before you scale the fleet
Office pod space planning is far easier the second time, because by then you have real demand data instead of estimates. The mistake is not collecting it — a first phase goes in, works, and nobody writes down how it is used, so phase two is another guess. Two weeks of deliberately crude measurement beats any model on this page.
Count occupancy, not bookings
Walk the floor at four fixed times a day for two weeks and record how many pods are occupied. It is unglamorous and it is the most reliable number you will get — booking systems record intent, and intent and occupancy diverge badly on walk-up space.
Log the turnaways
The decisive figure is not how full the pods are but how often somebody wanted one and could not have it. Those people do not complain, they go back to their desk — so ask, or put a sheet on the wall. A fleet at 60% occupancy with daily turnaways is under-supplied, not comfortable.
Separate the two demands
Record whether each occupied unit holds one person or several. If the single-person booths keep being used by pairs, you have room demand showing up as booth demand, and the next purchase should be rooms however busy the booths look.
Watch the peak, not the mean
Average weekly utilisation hides everything that matters. Pull out the busiest hour of the busiest day — that single figure is what your fleet is really sized against, and it is usually a midweek morning.
This is also a decent argument for renting the first phase rather than buying it. Renting turns a sizing assumption into a reversible one: if two weeks of data say the mix is wrong, you change it rather than living with it. Our buy, lease or rent comparison sets out what each route commits you to — though note that rental is not offered on every model, and thins out on the larger rooms.
Phasing a rollout across floors and sites
A fleet of twenty pods is not one project. It is a sequence, and treating it as a single order is how a rollout ends up with the wrong mix installed everywhere at once and no way back.
Pods pool per floor, not per building
This is the planning rule that gets broken most often, and the arithmetic above quietly depends on it. Nobody takes the stairs to find a phone booth — if it is not on this floor it does not exist. So run the sum once per floor using that floor's own population, not once for the building. A ten-pod building fleet split across five floors behaves like five fleets of two, with all the poor headroom economics that implies, and that is the real reason some multi-floor offices feel short of pods while their total count looks generous.
Phase one is a measuring instrument
Install enough to serve the floor with the clearest demand, deliberately slightly under the modelled number, and measure it properly for a month. You will learn more about your real mix from that than from any amount of planning, and the undersupply is informative — it is what produces the turnaway data that tells you what phase two should be.
Lead time and access compound at fleet scale
One pod is a day's work. Twenty is a programme, and the constraints multiply rather than add: lifts are shared, out-of-hours windows are finite, and the larger rooms carry longer lead times than the booths. Sequencing beats parallelism here, because the lift is the bottleneck whatever the crew size. The installation and access checklist covers the survey that has to happen per floor before any of this is schedulable.
Standardise across sites, but size each one separately
Multi-site operators should settle on a small, consistent set of models — it simplifies spares, finance and the internal conversation about what a "pod" is. What should not be standardised is the count. A sales office and a back-office site with identical headcounts have genuinely different demand, and giving them the same fleet because it is tidier guarantees one of them is wrong. We cover how multi-unit rollouts are coordinated on the business solutions page.
Four numbers, then a shortlist
Peak attendance, private calls per person per day, average call length, and how many meetings a week genuinely need a door. Those four decide the fleet, and three of them can be estimated well enough in an afternoon. Bring them to a conversation about models rather than the other way round, and the quote you get back is for the fleet you actually need.
More reading: matching pod size to how teams work, or browse all guides.