An office is asked to do two things that work against each other.
Collision: put people near each other so unplanned conversation happens. This is the argument for proximity and it is real.
For a practical reference on identifying mouse jigglers in remote teams, Monitask provides read more.
Concentration: let people work without interruption. Also real, and also most of what most people do.
A floor that optimises for one penalises the other. Most floors optimise for neither, and produce a compromise that is bad at both — which is the actual condition of most offices, and it is a design failure rather than an inevitability.
Why the compromise fails
The standard response is a large open area with a few enclosed rooms. This looks balanced and is not.
The collision benefit is smaller than assumed. Proximity produces incidental contact with the six people nearest you, not with the floor. Beyond a short distance the effect drops away sharply, which means a large open room does not produce more collisions than several smaller ones — it produces the same collisions plus more noise.
The concentration cost is larger than assumed. Intelligible speech is far more disruptive than louder undifferentiated noise, and a large room guarantees a supply of it. See what open plan actually did.
The enclosed rooms are always occupied. Two phone booths for eighty people is a statement of intent rather than a facility. If a quiet space is never available when needed, it does not exist.
What actually has to be separated
Sound, before anything else. The single highest-return intervention on an existing floor, and the least glamorous. Absorption, barriers, and separation between areas where people talk and areas where people work.
Talking from working. Not by policy but by geometry. If the route to the kitchen passes six desks, those six people are in a corridor.
Scheduled from unscheduled. Meeting rooms clustered rather than distributed, so that meeting traffic does not cross working areas.
Team from team. Most of the collision value is within a team; most of the noise cost is from other teams. Team-scale zones capture the first and reduce the second, and this is the highest-value configuration decision available. See hot-desking and desk booking.
What produces collision, and what does not
Does: shared routes to shared destinations. Coffee, kitchen, stairs, printers. The kitchen is the highest-traffic collision point in almost every office, and it is usually designed as a utility rather than as the social infrastructure it functions as.
Does: somewhere to stand and talk that is not in front of somebody working. The absence of this is why conversations happen at desks.
Does: the two minutes before and after a meeting, if there is anywhere for them to happen.
Does not: an open floor plan by itself. Removing walls does not make people talk more, and the best-known research on it found the opposite.
Does not: designated collaboration zones with beanbags. Space is used for what is convenient, not for what it was labelled.
A configuration that works
Not a universal answer, and a defensible default:
Team neighbourhoods, sized for a team, with acoustic separation between them.
Enclosed focus space, genuinely enough of it. The rule of thumb worth applying: if it is regularly unavailable, there is not enough.
Meeting rooms clustered, away from working areas, and more of them than the desk count suggests — on coordinated days the constraint is always rooms, never desks.
A social core where the routes cross, sized generously, with somewhere to stand.
Enough small call rooms. Video calls in an open area are the largest single source of noise in a modern office, and hybrid working has made this much worse than the open-plan research anticipated. See video parity.
Acoustic treatment throughout, which is the thing that gets cut first and returns the most.
The measurement
Occupancy tells you how full the building is, not whether it works.
Better:
Room availability at peak. If people cannot find a room on the busy day, the building is failing at its main job.
Failed searches for quiet space. Harder to measure and worth asking about directly.
Where people actually sit, versus where they were assigned. People vote with their feet and the pattern is informative.
One survey question: when you need to concentrate at the office, where do you go? If the common answer is "home," the floor is not doing one of its two jobs.
The framing that helps
An office is not a place where work is kept any more. It is a coordination device, and the concentration provision exists so that people who came in to coordinate can also do their actual work while they are there. See the office as a coordination device.
That ordering matters for design. Optimise the floor for the days people come in together, and provide enough concentration space that the day is not wasted — rather than designing a general-purpose workplace for a population that is no longer there five days a week.
For broader public guidance and background, consult the National Institute for Occupational Safety and Health.