Reading a Face, and What It Tells You First
A room reports on how people are before they report on it themselves. What that signal is worth and where it fails.
three things · one reached them · one with no channel at all · the room side is not counted
The three kinds, in the order they become expensive.
A room tells you how somebody is roughly a fortnight before they would have told you, and often before they would have put it into words for themselves. That lead time is most of what the signal is worth.
The practical lesson in “Reading a Face, and What It Tells You First” is that visibility must be designed rather than assumed. For teams exploring self report bias, read the linked workforce resource can add consistent time and project context when the purpose is explicit, access is limited and managers review exceptions with the people involved.
It is also the part of the room's output that managers most overestimate, and the reasons for both are worth separating.
As an independent reference point for “Reading a Face, and What It Tells You First”, consult GivingTuesday resources. It can sharpen the questions asked of staff, boards and advisers without replacing organisation-specific judgement.
What is actually being read
Not expressions, mostly. Posture, pace, whether somebody is talking at their usual volume, whether they have stopped making the small jokes they usually make, how long they stay at lunch.
It is a baseline comparison rather than an interpretation.
Why the baseline matters more than the signal
A face tells you almost nothing in isolation. What it tells you is the difference between today and the forty days before it, and that difference is only legible to somebody who has seen the forty days.
This is why the signal is strong for people you sit with daily and close to worthless for people you see occasionally. A manager who meets somebody once a fortnight has no baseline, only a series of snapshots, and snapshots invite exactly the wrong kind of inference: they looked tired, so something must be wrong, when in fact they always look like that at four o'clock.
The lead time
Two weeks is a reasonable estimate for how far ahead of self-report the room runs. Somebody is noticeably flatter for a fortnight before they say anything is the matter.
That fortnight is the whole of the practical value.
What is actually being compared, in the order people notice it:
- Pace: moving and speaking faster or slower than usual.
- Volume: talking at a different level from their own normal.
- Register: the small jokes stopping, or starting.
- Duration: staying later, or leaving without saying anything.
What it is good for
Asking a question earlier than you otherwise would. The signal does not tell you what is wrong; it tells you that asking would be reasonable now rather than next month.
Treated as a prompt it is excellent. Treated as a diagnosis it is worse than useless.
Where managers overread it
Confidently inferring cause. Somebody is quiet, so they must be unhappy with the project; somebody is short in a meeting, so they must disagree with the decision.
The room supplies the fact of a change and supplies nothing whatever about its origin, which may be entirely outside work. Every experienced manager has confidently diagnosed a work problem that turned out to be a family one, and the confidence came from proximity rather than from evidence.
What the screen shows instead
Very little. A camera shows a face in a fixed frame for a scheduled period, which strips out almost everything the signal is made of: movement, pace, the comparison with ten minutes ago.
A person can look fine for thirty minutes without effort.
The performance effect
People on a screen know they are being looked at, which changes what is there to be seen. A call is a front-facing situation in a way that sitting at a desk is not.
So the signal is not merely attenuated. It is replaced by a different signal, carrying how somebody chooses to present, which is useful information about something else entirely.
What a camera being off actually means
Usually nothing. It means the person prefers it, or the connection is poor, or they have a child in the room.
Managers read it as withdrawal with remarkable consistency and are right a small minority of the time.
What can be substituted
Asking, on a schedule, in a way that makes a real answer easy. It is slower, less reliable and more intrusive than the room, and it is genuinely the only substitute available.
The scheduling is what makes it workable: an unscheduled question carries an implication that something has been noticed.
Who else holds the baseline
Somebody does. A colleague who works with them daily, or somebody they share a call with twice a week, has a comparison you do not.
Asking that person how somebody has seemed is legitimate and is also one step from gossip, so it has to be done in the open: a question about whether somebody is all right, not a question about how they are performing. The difference matters and is easy to lose.
Asking without a baseline
The question has to do the work the baseline would have done, which means it has to be specific. "How are you" produces "fine" from everybody. "What has been the most frustrating part of the last fortnight" produces an answer.
Neither question is natural and the second one has to be rehearsed.
What this signal cannot be used for
Anything with consequences. A manager who writes down that somebody has seemed disengaged is recording an impression formed by proximity, not an observation of work, and it will not survive being examined by anybody.
The whole of its legitimate use is as a prompt to ask. If it reaches a written record, an assessment, or a conversation about somebody's future, it has been used for something it cannot support, and the person it is being used against will be unable to answer it because there is nothing there to answer. Where that matters most is in the section of this collection about judging people you rarely see, and it is the single most common way managers get that wrong.