I ran a one-on-one last March with an engineer who told me she was fine. She was also on her second late-night Slack message of the week, she had stopped pushing back in standup, and her code review times had stretched from same-day to three days. She was not fine. She had simply stopped asking for anything, and my team was about to lose her.
I have coached remote managers across three continents, and the same pattern keeps showing up. Burnout does not appear as an admission. It appears as behavior that no one is looking for. The most damaging number in every dataset I have seen is not the burnout rate, it is the manager response rate. The person is exhausted, they open their mouth, and 42% of the time the person on the other end does nothing.
This is a response protocol. Not a wellness program, not a survey, and not a slide about boundaries. It is a set of moves a remote manager can run in a twenty-minute one-on-one when the signals are already there.
Why the burnout rate is the wrong number to watch
Eagle Hill Consulting and Ipsos published Burnout in the Workplace 2026 in November 2025, drawing on 1,400 US employees. 55% of the US workforce reports burnout, and the split by arrangement is close: 61% fully remote, 57% hybrid, 55% in-office.
Burnout is not a remote problem. It is a distributed problem that everyone has agreed to fix by adding process, which is the exact opposite of the fix.
Gallup’s State of the Global Workplace 2026 reports that 67% of employees experience burnout, and burned-out employees are 2.6 times more likely to actively job hunt. That multiplier is what I actually lose sleep over, because it is the number that shows up in your attrition forecast next quarter.
What the data is telling me to change
- Do not run another wellness survey.
- Do not add another ritual to the calendar.
- Do fix the manager response rate, because that is where the intervention actually happens.
Burnout collapses engagement through a specific mechanism: someone who is running on empty still performs for a while, then stops asking for anything, then leaves. The window between performing and leaving is where a twenty-minute conversation saves you a replacement cost.
Behavioral signals, not self-report
Burned-out people rarely volunteer the diagnosis. My work with remote managers taught me to stop asking how they feel, and start tracking what they do.
| Signal | What you observe | Early-warning threshold | Manager action |
|---|---|---|---|
| Workday drift | Start and end times spread beyond the agreed block | Two consecutive weeks outside the pattern | Open the energy conversation in the next 1-on-1 |
| Off-hours work | Messages sent or read outside the working window | Three times per week, sustained | Confirm whether the pattern is a preference or a symptom |
| Switching-off difficulty | Late-night activity, weekend commits, refusal of PTO | One month of overlap | Reduce scope by one committed deliverable |
| Engagement decline | Quieter in standup, deferred opinions, thinner reviews | Three consecutive reviews without substantive comment | Ask what is making participation feel like cost |
| Decision avoidance | Pushes choices back to the manager, blocks on approvals | Two decisions held past the natural deadline | Pair on the decision rather than asking for the answer |
What the signals share
- They are observable without a conversation.
- They persist across more than one week before they become a burnout event.
- They are visible from the manager side of the screen, which is exactly where most managers do not look.
Burnout is not a feeling, it is a set of behaviors that a manager is trained to ignore.
That last line is the whole protocol. If you can observe the signal without asking, you can intervene without waiting for a confession.
The twenty-minute conversation
I do not open a burnout conversation with ‘how are you doing’. That is a small-talk opener that a tired person will answer to protect the relationship. I open with energy.
‘I want to talk about energy, not effort. Effort is what you put in. Energy is what you have left after it. Can you describe where yours has gone this quarter?’
Structure of the conversation
- Two minutes on what you have observed, without a diagnosis.
- Eight minutes on what they describe as unsustainable.
- Eight minutes on what you are prepared to remove from their plate.
- Two minutes on a check-in date.
Do not diagnose. Do not label. Say what you saw, and let them tell you what it means. In my experience, the person who is exhausted already knows what they need; they have just stopped expecting it to be granted.
Cut load, do not add process
The most common failure I see from managers is adding a wellness ritual to someone who is overloaded. A standing ‘wellness check-in’ costs the same minutes as the meeting the person is trying to escape. A monthly mood tracker adds admin on top of overload.
Load reduction is what actually works. It is also what most managers are quietly afraid to do, because cutting scope looks like weakness in the moment.
What I actually cut
- One low-value commitment before anything else.
- One meeting that could have been a Slack thread.
- One deliverable that has been stretching the timeline for two cycles.
- One approval loop that has accumulated three people where one would do.
What I never add
- Any new recurring ritual aimed at wellbeing.
- Any new reporting line about mood or energy.
- Any recommendation to ‘take time off’ without removing a work obligation first.
The person does not need another check-in. They need one less thing.
The burnout risk dashboard
I moved from gut feeling to a dashboard when I realized I was missing burnout events in the second half of the year and catching them in the first half. The same manager cannot feel everything about a distributed team, so the signals need to be surfaced.
A minimal dashboard has four to six metrics, refreshed on a two-week cadence.
Metrics that belong on it
- Workday drift count per person per week.
- Off-hours message rate over the past fourteen days.
- Meeting load trend over the past six weeks.
- PR or task review latency for a role where it applies.
- Days since last PTO.
- Open decisions held by the individual.
The dashboard is not a surveillance tool. It is a manager review view, refreshed on the same cadence as the operational review already on the calendar. Two weeks is the window between ‘starting to drift’ and ‘about to leave’.
A burnout dashboard is a review view, not a watchtower.
If you have to invent a new meeting to look at it, you do not have a dashboard, you have a ritual. Attach it to the recurring operational review and let it live there.
Making intervention a team-level training
Individual managers trained individually will drift. The pattern holds only when every manager on the team has the same protocol, the same vocabulary, and the same dashboard.
NAMI’s 2025 Workplace Mental Health Poll found that mental-health training reduced the share of employees reporting burnout-related productivity loss from 38% to 21%. That is not a wellness number. It is a productivity number, and it is the argument for training managers as a team, not one at a time.
| Dimension | How training changes it | Reported impact |
|---|---|---|
| Productivity loss | Manager recognizes and intervenes before engagement collapses | NAMI reports a reduction from 38% to 21% of employees reporting burnout-related productivity loss after training |
| Escalation rate | Signals get named at the manager level instead of the HR level | Directional; not tied to a specific sourced figure in this protocol |
| Perceived support | Person sees the manager remove something concrete, not add a ritual | Directional; not tied to a specific sourced figure in this protocol |
What the training actually covers
- The five behavioral signals and their thresholds.
- The twenty-minute conversation structure.
- The load-reduction checklist, rehearsed once per manager.
- The dashboard cadence and how to read it.
- A debrief that names the escalation path, so the manager is not alone when a case is serious.
What I have stopped doing
I used to run quarterly burnout surveys. They gave me a slide that agreed with the annual report. I used to add wellness meetings to manager calendars. They filled a gap in the ritual calendar and added a meeting to a full calendar.
The two things I kept were the twenty-minute conversation and the dashboard. Everything else was theater around a gap in the response process. The burnout rate is not the number to fix. The manager response rate is.
If you take one thing from this protocol, take the conversation opener. Ask about energy, name what you saw, and let the person tell you which demand is unsustainable. Then cut something. That single sequence has saved more of my team than any wellness initiative I have ever run.