The stress signals managers miss until it is too late

What do early stress signals actually look like?

Early stress shows up as a change in someone’s normal pattern, not as visible distress. That is the whole difficulty. The person heading for trouble in two months is rarely the person in tears today. They are the person who used to push back in planning meetings and has stopped, or the one whose work is still good but has quietly lost its edges.

Most managers watch for the wrong thing. They watch for distress, which is loud, obvious and very late. The useful signals are quiet, early, and almost all of them are changes rather than states:

• Withdrawal from optional contact. They still attend everything compulsory. They stop turning up to the things with no agenda: the coffee, the five minutes before the call, the informal catch-up. Optional contact is the first thing a stretched brain deletes.
• A thinning at the edges of the work. What they own tightly is still fine. What sits at the periphery gets thinner: shorter handovers, less documentation, fewer proactive flags about things that will bite in six weeks. Nothing has broken, so nothing gets said.
• Over-availability. Replies at 23:10. Volunteering for the extra piece. Leave that gets booked, then moved, then moved again. This is the signal most often missed, because it looks exactly like commitment.
• Defensiveness about small things. A minor correction comes back with three paragraphs of justification. Under sustained load, small challenges start to land as verdicts.
• Decisions coming back upward. Things that used to be settled in the corridor now arrive in your inbox, or simply wait. Weighing options is expensive, and a depleted person stops paying for it.
• The social texture going. The jokes, the asking about someone’s weekend, the willingness to be slightly unproductive for two minutes. It goes early and it goes without comment.

No single item on that list proves anything. People are quiet for a hundred reasons, and someone who works late on Tuesday is just someone who worked late on Tuesday. What matters is the combination and the duration: two or three of these, in the same person, sustained past a fortnight, where the pattern used to be different.

Why do capable managers miss them?

Managers miss these signals because each one has a perfectly reasonable innocent explanation, and because noticing carries a cost the manager has not budgeted for.

The first trap is reading behaviour as character. Once you have decided that someone is “just quite private” or “not a meetings person”, every subsequent withdrawal confirms the label rather than raising a question. The label is doing the work that observation should be doing. The test is simple: were they like this in March. If you cannot remember, that itself is worth knowing.

The second trap is that the strongest signal looks like your best outcome. A person who answers everything within four minutes, absorbs a departed colleague’s workload without complaint and has not taken proper leave in eight months is, from a delivery perspective, a gift. Organisations reward that behaviour openly and then express surprise when it ends abruptly. If you only investigate the people whose output is dropping, you will systematically miss the ones who are burning capital to keep it up.

The third trap is the threshold that never arrives. Managers hold an unspoken bar for when a concern becomes raiseable, and that bar is usually set at “obviously serious”. So you wait for one more data point, and while you wait the situation normalises. Two months later the signal is not subtle any more, but by then the person has already decided you were not going to ask.

There is a fourth reason, and it is the one people admit to privately: managers avoid the conversation because they do not know how to end it. They imagine opening something they cannot close, being told about a divorce or a diagnosis, and having no idea what to say next. That fear is reasonable and it is also solvable, because the conversation is smaller than they think.

What does a useful check-in actually sound like?

A useful check-in names one specific observed change, offers no diagnosis, asks a question with no adjectives in it, and then stops talking long enough for an answer to arrive.

Here is a constructed example, the sort of situation that turns up in most Lead workshops. Someone on your team runs a small integration workstream. For three years they have been the person who says “that timeline is fiction” in planning meetings. Over about six weeks, they stop. Their delivery is still on schedule. Their handover notes, which used to be famously thorough, are now three bullet points. They reply to messages on Sunday evenings. They have moved their October leave twice.

The version of the check-in that fails goes like this. “Everything alright? You seem a bit stressed. Just shout if you need anything.” Three things are wrong with it. It offers a diagnosis, so the only available answers are to accept the label or deny it, and most people deny it. It is vague, so it can be met with “yes, fine, just busy” at no cost. And it hands the next move to the person with the least energy, which means there is no next move.

The version that works is more specific and less sympathetic:

“I want to put an observation to you, and it is an observation rather than a criticism. Over the last month or so, two things have changed. You have stopped challenging timelines in planning, which you always used to do, and your handovers have got much shorter. Both of those are unlike you. What is going on with the work at the moment.”

Then silence. Not a comforting noise, not a follow-up question to fill the gap. Silence long enough to be slightly uncomfortable, because an honest answer takes a few seconds to assemble and most managers fill the gap before it arrives.

What that phrasing does, in order: it puts the manager’s perception on trial rather than the person’s competence, so nobody has to defend themselves. It cites two concrete, dated, observable changes, which cannot be waved away with “just busy”. It avoids the word stress entirely, so the person is not required to accept a label to answer. And it ends on a question about the work rather than about their feelings, which is the question a manager has standing to ask.

What do you do once someone tells you the truth?

Ask the question only if you are prepared to change something, because a check-in with no follow-through is worse than no check-in at all. You have just taught someone that honesty costs them and buys nothing.

Managers usually have three levers, and it is worth knowing which ones are genuinely yours before you open the conversation:

• The load. What comes off the list, not what gets reprioritised on it. Reprioritising is what managers offer when they cannot remove anything, and everybody in the room knows it.
• The clock. A deadline you can move, or a decision you can take now rather than in three weeks so that someone stops carrying it.
• The decision rights. Sometimes the pressure is not volume, it is having responsibility without the authority to resolve anything. That one is cheap to fix and rarely spotted.

Then agree something specific and put a date on it. “Let us see how it goes” is not an agreement. “You are off the migration workstream from Monday, and we will look at this again on the fifteenth” is. Write it down, and be the one who reopens it, because the person under pressure will not chase you.

It is also worth being clear about what is not your job. You are not a therapist and you should not try to be. If what surfaces is a health matter or something outside work, the useful move is to say plainly that you are out of your depth, point at whatever professional support the organisation provides, and stay focused on the part you control, which is the work itself. Knowing where your role ends makes the conversation safer to start.

Building the habit before you need it

The reason these conversations go badly is almost never that the manager does not care. It is that they are attempting a difficult, unrehearsed conversation live, with a colleague, in the middle of a busy week, with no practice and no feedback. That is a hard way to learn anything.

It is also the reason we build our stress and resilience work around practice rather than awareness. Bounce Back From Stress is aimed at the people carrying the load: it maps each participant’s own early warning signs in the order they actually appear, then spends the harder half of the day on saying no, later, or less, out loud, with feedback, until it sounds like the person speaking rather than a borrowed script. It runs as a one-day workshop or two half-days, in person or virtually, for 6 to 14 people.

The part that matters most for managers is the optional briefing that comes afterwards. A boundary agreed in a workshop and never mentioned to the manager tends to last about a fortnight, because the whole system it has to survive is unchanged. Teams where the manager knows what was agreed, and has practised asking about it without flinching, get a different result. If you want to talk through what your team is showing you at the moment, Louisa Savva runs these sessions personally and will tell you straight if training is not the answer.