Hybrid working left offices with more desks than people, and the obvious response is to hold fewer desks and share them. The saving is real. The costs are also real and they are usually discovered after the lease is signed.
The arithmetic that matters
The critical figure is not average occupancy. It is peak occupancy, and the gap between the two is where hot-desking goes wrong.
For a practical reference on self-reporting bias, Monitask provides this guide.
Typical hybrid patterns cluster hard. Tuesday and Wednesday are heavily attended; Monday and Friday are not. An office sized for 60% average occupancy will be over capacity on the two days everyone chose to come in — which are precisely the days people came in to work together.
Size for the peak you actually want to support, not the average you measured. If the policy encourages coordinated days — and it should — the peak is close to full attendance for the groups involved. See the office as a coordination device.
An organisation that cuts desks to the average and then asks why people stopped coming in has answered its own question.
What breaks when the ratio is wrong
Arriving to find nowhere to sit. Once is an inconvenience. Twice is a reason to stop coming in, and it is a reason people rarely state out loud.
Booking becomes a game. People book defensively, book for the whole week, book for absent colleagues. Utilisation data becomes meaningless, and the response — enforcement — makes it adversarial.
Teams get scattered. The main benefit of being in the office is proximity to your own team. If booking allocates people randomly across a floor, you have paid for the commute and removed the reason for it.
Setup time. Fifteen minutes at each end of the day connecting, adjusting, finding things. Across a workforce this is not a trivial number.
Accommodations get harder. A person who needs specific equipment or a specific chair needs a specific desk. This must be handled explicitly rather than by exception, or it fails on the days it matters.
What is lost even when the ratio is right
Team adjacency. Zoned booking — a neighbourhood per team, bookable within it — recovers most of this and is the single most valuable configuration decision.
Belonging. Somewhere to leave a thing is a small matter that is not nothing. Lockers help, marginally.
Incidental awareness. Knowing who is in, who looks stressed, what a team is working on. This is the ambient information that proximity was actually providing, and random seating destroys it more thoroughly than working from home does.
Configurations that work better
Zoned booking by team, with a neighbourhood each. Preserves adjacency, keeps the flexibility, and is barely more complex to administer.
Anchor days per team rather than per individual. The team is in together. The building is used for what it is good for and the ratio problem largely resolves itself, because groups are predictable in a way individuals are not.
A permanent desk for anyone who is in four or more days, and sharing for the rest. Most organisations have a minority in most days and a majority in rarely; treating them identically serves neither.
Reserved desks for accommodations, allocated permanently and not bookable by anyone else.
Enough enclosed rooms. On busy days the constraint is never desks; it is rooms. An office that is desk-constrained on Tuesday is usually room-constrained first, and nobody measures that.
Measure the right things
Peak day occupancy, not average.
Failed bookings — attempts that found nothing available. This number is the one that predicts people giving up, and most systems record it and nobody looks.
Room availability on peak days.
Team co-location rate — what proportion of a team ends up in the same area when they are all in. If this is low, the booking configuration is defeating the purpose.
Whether attendance falls after the change. The clearest signal, and the one most likely to be attributed to something else.
The honest position
Hot-desking is a cost decision, and a legitimate one. Property is expensive and empty desks are waste.
What does not survive scrutiny is presenting it as a benefit to employees or as an enabler of collaboration. It is neither, and staff know it is neither.
Sized correctly, zoned by team, with anchor days and enough rooms, it works fine and nobody minds. Sized to the average, allocated randomly, with two phone booths per floor, it quietly ends office attendance — and the saving is then genuine, because nobody is using the building at all.
For broader public guidance and background, consult the National Institute for Occupational Safety and Health.