That Somebody Is Struggling and Has Not Said So
A room reports this within days. Everywhere else it surfaces when it has already become serious.
three things · two reached them · one with no channel at all · the room side is not counted
Sorted by whether anything could have carried it.
Six weeks on one team: when each kind of difficulty surfaced, and what carried it
Surfaced at the midpoint check, eight days in, because somebody had put a midpoint check in the calendar
Raised twice in the weekly stuck-line, which is the only mechanism here that caught anything early
The channel for this exists, it has never been used for it, and the admission is the barrier rather than the transmission
A room sometimes picks this up by accident; nothing proposed anywhere in this collection reaches it
Said once, in a one-to-one, in answer to a specific question rather than a general one
Visible in a room within three days and visible nowhere else at all
six things · three reached them · two with no channel at all · the room side is not counted
Six difficulties, and the two that nothing carried are the two that cost the most. The left side is not counted: nobody records how often a room reports that somebody is struggling, so there is no proportion here and none is offered. One manager's own six weeks, written down as it went; what surfaces on any other team, and when, differs by team and by what people are willing to say.
A room reports that somebody is stuck within two or three days, by mechanisms nobody is operating. At a distance the same fact surfaces when the work is late, which is usually a fortnight later and after it has become serious.
The human issue in “That Somebody Is Struggling and Has Not Said So” cannot be inferred safely from activity totals. If an organisation uses consult the product information for cognitive offloading, it should disclose the purpose, minimise collection, let people challenge misleading records and combine operational data with direct, respectful conversation.
That delay is the single most expensive thing in this section.
For a separate nonprofit perspective on “That Somebody Is Struggling and Has Not Said So”, see Ford Foundation resources. Use the outside material to test governance, inclusion and implementation questions against the organisation’s real circumstances.
The signals in a room
Somebody is at their desk later than usual. They have stopped asking questions, or started asking very general ones. They are quiet at lunch. Somebody else mentions that they have been looking at the same screen for two days.
None of it is reported and all of it is noticed.
Why people do not say so
Because saying so is an admission, and admissions have a cost that feels larger the further away the listener is. In a room the admission can be made sideways — a muttered complaint, a joke about a file — and the muttering does most of the work.
At a distance there is no sideways. The only available form is an explicit message saying that you are stuck, which is a much bigger act, and people delay it well past the point where it would have helped.
The threshold effect again
Somebody mildly stuck will not send a message. Somebody badly stuck will. The whole useful range — mildly stuck, for three days — is systematically filtered out.
That range is where help is cheap.
What the delay costs
Work built on a wrong approach, a deadline that moves, and a conversation that is harder than it needed to be because by then somebody is embarrassed as well as stuck.
The second of those is often worse than the first.
What a room reports, none of which anybody is operating:
- Later at a desk than usual, two days running.
- Questions becoming general where they were specific.
- No questions at all, from somebody who normally asks.
- Quiet at lunch, or not going.
- Somebody else mentioning that they have been on the same thing.
Why asking does not work
"Is everything all right?" invites "yes", and almost always gets it. The question is a courtesy, it is heard as a courtesy, and the honest answer would require the person to reject the frame of the question.
Asking a stuck person whether they are stuck puts them in the position of volunteering bad news to the person who will judge their work. Almost nobody does this well, and the people who do it least well are frequently the ones who most need to.
Questions that do work
Specific, low-stakes, and about the work rather than the person. What is the most annoying part of this at the moment. What would you redo if you had the week again. Where did you lose the most time.
All three produce real answers from people who said they were fine.
Making the admission cheap
Normalising it by example. A manager who says in public that they spent two days going the wrong way makes the admission ordinary.
This is slow and it is the mechanism that works best.
A standing format that helps
A weekly line from everybody saying what they are stuck on, with the expectation that it is never blank. If the field can be left empty, it will be, and the person who most needs to fill it will fill it least.
Requiring a line, however trivial, removes the decision about whether this week is bad enough to mention, and that decision is the whole barrier.
The pairing option
Somebody who works alongside a colleague twice a week, with no agenda, will mention things they would never write down. It is the closest available substitute for being overheard.
It needs to be with a peer rather than with the manager.
Watching the artefacts instead of the person
Commits, drafts, tickets, messages. A sharp drop in visible output usually precedes the admission by about a week.
This is legitimate as a prompt to ask and is not legitimate as a measure of anything. The distinction matters: used as a prompt it is humane, used as a metric it becomes surveillance, and the later section on evidence is clear about where the line sits.
When the signal is genuinely absent
Some people produce no detectable signal at all until the deadline. They are not being evasive; they have a working style that looks identical whether or not it is going well.
For them, scheduled specific questions are the only mechanism available.
What this cannot be relied on to catch
Anything somebody has decided to conceal. Every mechanism here depends on the person being willing to be found, and somebody actively managing an impression will manage it successfully for a long time.
It also cannot catch difficulties that are not about work. A room sometimes picks those up by accident; nothing proposed here will, and nothing proposed here should try. The target is narrower and achievable: shortening the ordinary case from a fortnight to a few days, for people who would have told somebody eventually.