Against the nearly private room

Most enclosed rooms in most offices are acoustically ornamental. The partition stops at the suspended ceiling, the void above it is continuous to the next room, the door has a three-millimetre gap at the threshold, and the glazing is single-glazed because it looked better in the visual.

Such a room reduces sound by perhaps twenty-five decibels. That is enough that the conversation is muffled and not enough that it is private. Every word is recoverable by anyone who cares to listen from the desk outside.

This would merely be a shame if people understood it. The problem is that they do not, because the room looks private, and so they use it as though it were.

What we found at Selby Court

Selby Court is a law firm. Before their move we spent three weeks logging what actually happened in their existing premises: how many calls, how long, requiring what degree of confidentiality, at what time of day.

They occupied sixty-one cellular offices for ninety-six staff. Peak simultaneous demand for genuinely private conversation — the kind where a name and a circumstance must not be overheard — was fourteen.

Fourteen out of sixty-one. The other forty-seven rooms were, most of the time, single-occupancy offices being used for email.

Fourteen rooms, built properly

The new floor has fourteen enclosed rooms. Every one is built to a real standard: partitions full height to the structural slab, an isolated head detail, acoustic doors with drop seals, independent ceiling to each room, and no shared void anywhere.

Tested performance is 45 dB Dn,f,w. At that level a raised voice inside is inaudible as words outside.

Per room, this costs roughly two and a half times what a standard partition costs. In total it cost dramatically less than sixty-one rooms of any specification, and it produced fourteen rooms that do the job rather than sixty-one that do not.

The uncomfortable part of the argument

Reducing sixty-one offices to fourteen bookable rooms is, to most of the people affected, a demotion. Cellular offices in professional practices are not primarily acoustic devices; they are seniority made visible.

We did not pretend otherwise, and we would advise against any consultant who does. The occupancy data was useful because it settled a factual question — how many private conversations happen at once — and completely useless for the status question, which is a governance decision the firm had to make itself.

They made it before we started detailed design, which is the only reason the project worked. Had that decision drifted into Stage 4, we would have ended up with thirty compromised rooms, which is the worst of both.

Where the ceiling matters most

With the private conversations properly contained, the open floor turned out to need far less acoustic treatment than we had budgeted for. Most residual noise was incidental talking in circulation, so the absorption is concentrated in a raft over the central route rather than spread evenly across the ceiling.

That concentration is more effective per square metre and considerably cheaper, and it only became the obvious answer once the difficult conversations had somewhere else to go.

The rule we now apply

If a room is described as private, it must be built to a tested standard, and if the budget will not carry that, the room should not have a door.

An open booth that is visibly not private is honest. People modulate what they say in it because they can see that they are in the open. A sealed-looking room that leaks invites exactly the conversation it cannot protect, and the failure is invisible until it is not.