Why your most experienced supervisor is the most likely to quit next quarter

The same supervisor role has been re-listed twice in twelve months. The exit interviews come back with "looking for a new challenge" — clean, polite, and oddly consistent across people who have nothing else in common. Meanwhile replacement and onboarding costs keep landing on the same cost line quarter after quarter, and the margin on those sites is quietly thinner than the ones where the supervisor has stayed put for six years. You have read the exit notes. They don't add up to the pattern you are watching.

They don't add up because the stated reason and the actual reason are different things. People leaving a role that is grinding them down very rarely write that down on the way out. "New challenge" is the socially safe answer. The real one is closer to: the job asks more of me emotionally than it returns, and no one ever named that as part of the job.

Why does the most capable supervisor go first?

Because capability is exactly why you load them up. The supervisor who is good with people becomes the person every difficult conversation routes through. The contractor dispute, the apprentice whose dad is sick, the crew member who is one bad week from a safety incident, the client who rings at 6am — all of it lands on the same desk. The better they are at it, the more of it they get.

The data tells the same story from the top down. Burnout among immediate managers and supervisors sits at 54% — the highest of any career level, and up six percentage points since 2022 (CMHAA Leading Mentally Healthy Workplaces Survey, 2024). Operational and middle management isn't far behind at 50% (CMHAA, 2024). These are not the most junior people and they are not the executives. They are the layer that connects the two — the conduit between the work and the people doing it, between the regulation on paper and the reality on the ground.

That layer carries the highest emotional load in the business, and in most organisations it carries that load without a single framework or boundary to manage it. Your supervisors are already doing the emotional labour. They are simply doing it unnamed, unstructured, and without anything that shows up on a balance sheet.

Isn't this just a hard job that some people aren't cut out for?

That's the wrong question — and it's the question that keeps you re-listing the same role. If it were a selection problem, you would not see the pattern repeat across different people in the same seat. You are not hiring the wrong individuals. You have built a position that systematically converts good operators into accidental counsellors, then expresses surprise when they burn out.

Consider what we ask a multi-site supervisor to hold. Robin Dunbar's work on social cognition put a practical ceiling — roughly 150 — on the number of stable relationships one person can actively maintain. A supervisor across two or three crews, plus contractors, plus client contacts, is often well past that line and absorbing the personal disclosures of all of them. There is no version of "be more resilient" that fixes a structural overload. Resilience training is the thing organisations reach for here, and it is precisely the thing that is not enough — it asks the individual to better withstand a load the operating model should never have put on them alone.

What does this actually cost the operation?

It costs you twice. First in turnover — the recruitment fee, the onboarding ramp, the productivity dip while a new supervisor learns the site, the institutional knowledge that walks out the gate. Second, and larger, in everything that supervisor was silently preventing: the near-misses they defused, the contractor relationships they held together, the early-warning signs they caught before they became incidents. When that person breaks, a chunk of your operating model breaks with them, because so much of it was running on their unrecognised judgement.

There is a compounding effect underneath. Those same supervisors are the people you would expect to support a struggling team member — yet 38% of leaders and managers say their organisation has not provided training to support their team's mental health, or are unsure whether they have received any (CMHAA, 2024). So the layer carrying the most load is also the layer least equipped to carry it. And the people most exposed beneath them are the youngest: burnout among under-25 employees runs at 54%, the highest of any age bracket (CMHAA, 2024) — the apprentices and first-jobbers who lean hardest on a good supervisor for their first taste of how work is meant to feel.

Doesn't the EAP already cover this?

The EAP is necessary, and it is downstream. It activates after harm — it is reactive support that someone reaches for once they are already in trouble, and it relies on the individual recognising they need it and choosing to call. By the time your supervisor is dialling that number, the cost has already been incurred: the disengagement, the quiet job-hunting, the resignation that is three weeks from your inbox. The EAP sits at the bottom of the cliff. The supervisor went over the edge somewhere up the slope, in the months no one was looking.

The gap is everything that happens before that call — and it is where the actual exposure lives. This is the part most organisations cannot see, which is why it goes unmanaged. Making that invisible risk visible is the whole point. And here is the part operations leaders tend to miss: building structure in that gap does two things at once. It reduces the need for the EAP, because fewer situations escalate to crisis. And it increases appropriate uptake of the EAP, because people get routed to it earlier, when it can still do something useful. Done properly, prevention upstream makes your existing reactive support work better, not redundant.

If it isn't a morale problem, what is it?

It is a psychosocial risk problem, and under Australian WHS law that reframing matters. High job demands, low support, and exposure to traumatic or emotionally charged material are named psychosocial hazards. You already manage physical hazards with a risk assessment, a risk register and a hierarchy of controls — you identify the hazard, you control it as far as is reasonably practicable, and your officers exercise due diligence to make sure that is happening. Psychosocial hazards sit under the same duty. The supervisor's emotional overload is not a wellbeing nicety; it is an uncontrolled hazard on a register that probably doesn't list it yet.

That is the shift from a "mental health" framing to an operational and regulatory one. The question stops being "how do we make people feel supported?" and becomes "where is the unmanaged psychosocial load in our operating model, and what controls have we put in place?" That is a question a COO, an HSE manager and a risk lead can all act on — because it has a register, controls and an evidence trail attached, the same as any other hazard you'd be asked about after an incident.

What does structured prevention look like in practice?

Not another course. The supervisor doesn't need to be taught to absorb more — they need the load distributed and a structure underneath them. In practice that means structure built in the gap upstream of the EAP: a psychosocial safety framework that names the hazard, a network of trained wellbeing champions so the emotional traffic isn't all funnelled through one or two accidental counsellors, clear boundaries on what a supervisor is and isn't expected to hold, and a defined route — the Pathway — that moves someone from "noticing a problem" to "the right support" without the supervisor having to improvise it alone at 6am.

This is wellbeing infrastructure, not a campaign. It is the difference between hoping your best supervisor stays, and building an operating model that doesn't quietly consume them. The first option keeps re-listing the role. The second keeps the person — and the margin, the knowledge and the incidents-that-never-happened that go with them.