By Angela Hucker, Founder and CEO, EPIC
Almost nothing in an exit interview is true.
Not because people lie exactly. Because we ask them why they are leaving at the precise moment they have the least possible reason to tell us. They still need a reference. They may want to come back one day. The industry is small and everyone knows everyone. So they give the answer that costs nothing and closes the conversation.
More money. Closer to home. Better opportunity. Time for a change.
Then we aggregate those answers, and they tell us construction has a pay problem and a market problem. And we build an entire national response on top of that, which is how the industry ended up spending a decade on attraction while the people already inside it kept walking out the gate.
One of the leaders I interviewed for the EPIC Construction Leadership Report 2026, an HR director with nineteen years in utilities, described the problem more precisely than any dataset has managed.
“Half my job is reading between the lines. People don’t tell you the real reason they’re leaving until they’ve already gone.”
1. The attraction argument, and the hole in the bucket
Australian construction is facing a workforce shortfall projected in the hundreds of thousands, and Jobs and Skills Australia continues to list construction trades among the country’s most persistent occupational shortages. That is real, and I am not going to minimise it.
The response has been overwhelmingly about getting more people in. School programs. Careers campaigns. Apprenticeship incentives. Migration settings. Programs to bring women into the trades. All of it useful, much of it overdue, and I support essentially all of it.
But the attraction argument has the same problem as every other explanation in this series. It is not wrong. It has become too complete. It explains the shortfall so neatly that it lets us skip the more uncomfortable question, which is what happens to people once they arrive.
Turnover across the construction industry is commonly reported in the range of 25 to 30%. You cannot campaign your way out of that. Every dollar spent attracting someone into construction is wasted if the supervisor they land under was promoted on a Friday and abandoned on a Monday.
People do not leave industries. They leave sites, crews and supervisors.
2. The decision was made months ago
Here is what I have watched happen more times than I can count, and it never looks like a decision while it is happening.
Somebody raises something and nothing happens. A few weeks later they get spoken to in front of the crew over something minor. The interesting job goes to the same bloke it always goes to. They ask about doing their ticket and are told the timing is not right, twice. Someone behaves badly at smoko and everyone looks at the supervisor, who says nothing. They start doing exactly their hours. They stop offering ideas at the toolbox talk. A mate mentions a job somewhere else and, for the first time, they actually listen.
Six months later they resign, and the exit interview records that they got a better offer.
Which is true. It is also the last event in a sequence that had nothing to do with money, and the sequence was made entirely of small leadership moments that nobody logged.
By the time a person resigns, they left months ago. What you are watching is the paperwork catching up.
“Retention is not an HR initiative delivered after someone becomes disengaged. It is the accumulated result of how people are led every day.”
3. The eight experiences that decide it
When you strip retention back to what a person actually experiences on a work front, it comes down to eight things. Every one of them is delivered or withheld by a frontline leader, usually without any deliberate decision being made.
Eight everyday experiences
Nobody resigns over any single one. People resign over the accumulation.
| What makes people leave | What makes people stay | |
|---|---|---|
| Expectations | Never sure what good looks like, then blamed when it is not delivered. | Outcome, standard, timeframe and owner, stated up front. |
| Fairness | Decisions depend on who you are and what mood the boss is in. | The same rules applied to the apprentice, the gun tradie and the subcontractor. |
| Accountability | Corrected in front of the crew, or never corrected at all. | Told early, told privately, told specifically, with a clear way back. |
| Growth | Same task for three years. The timing is never right for the ticket. | Stretch work, tickets supported, someone actively backing them. |
| Trust | Every decision routed through the supervisor. No authority to act. | Real authority with agreed boundaries and an agreed check point. |
| Behaviour | One person makes it unbearable and nothing is done about it. | The leader addresses it, early, without waiting for a complaint. |
| Being heard | Raised something once, watched it disappear, learned the lesson. | Concerns are acknowledged, acted on, and followed up within a week. |
| Access | The good work always goes to the same people, and everyone can see it. | Opportunity, mentoring and progression distributed on capability. |
Read the left column as a resignation letter written six months in advance.
That table is really the whole series in one visual. Each row is one of the previous nine articles: clear expectations, fair decisions, respectful accountability, delegation and growth, addressing harmful behaviour, being taken seriously when you speak up, and equal access to work and progression. Underneath all of them sits a leader who is not transmitting their own pressure into every interaction.
4. What it costs, honestly
I am wary of turning people into line items, but the financial argument is the one that gets budget approved, so here it is plainly.
The cost of every skilled departure in Australian construction is estimated at around $67,000 once you count recruitment, onboarding, lost productivity and the knowledge that walks out with them. Broader workforce research commonly puts replacement cost somewhere between 50% and 200% of an annual salary depending on the role.
One departure, or one trained leader
Estimated cost of losing one skilled worker against the cost of training the leader they report to
$67,000 One skilled departure $3,497 Leadership training, full fee $700 After Keystone rebate (Tas)Departure cost: Infrastructure Australia. Course fee excludes GST. The 80% rebate is Keystone Tasmania and eligibility applies.
One supervisor leads a crew. If training that supervisor prevents a single resignation across a year, the return is not marginal. And most businesses are not losing one person a year off a struggling work front. They are losing three or four, and describing it as a market problem.
Gallup’s long-running workplace research has consistently found that the manager accounts for the large majority of the variance in team engagement. That finding is not construction-specific, but on a site where a crew spends fifty hours a week within twenty metres of the same supervisor, it is not less true. It is more.
5. The generational question, asked properly
78% of the leaders in the report raised generational retention. In most industry conversations this becomes a complaint: young workers do not want to work, they will not do the hours, they expect everything immediately.
One of the leaders I interviewed, a general manager with twenty-five years in commercial, put it differently, and I have quoted him often since.
“Young workers don’t want what we had. And honestly, what we had wasn’t that good. We just didn’t know any different.”
That is one of the most honest things anyone said to me across fifty interviews.
What actually changed is not resilience. It is tolerance for being led badly. My generation absorbed a great deal because we believed there was no alternative and because leaving looked like failing. Younger workers do not carry that belief, and in a labour market this tight, they are correct not to. They will leave a bad supervisor in four weeks.
You can call that a problem with them. It is more useful to treat it as a measurement instrument. Younger workers are a fast, accurate sensor for leadership quality, and they report their findings by resigning.
6. The people who leave first are the ones you most want
There is a pattern in poor culture that makes it far more expensive than the turnover percentage suggests.
Your most capable people have the most options. They are the ones who can walk into another job this month. So when a work front becomes difficult, they go first, and the people with the fewest alternatives stay longest.
Over two or three years that quietly re-sorts your workforce, and no report will ever show it to you. The turnover number stays roughly the same while the composition of who stays gets worse. Businesses in this position usually describe themselves as having a recruitment problem.
The same dynamic applies to anyone already in a minority on that site. A tradeswoman on a crew where behaviour is not addressed does not need to weigh up much. There are easier places to work. That is the mechanism behind three good women leaving in two years, described in unconscious bias in leadership decisions and in Designed Without Her, and it is a retention problem long before it is a diversity problem.
7. What should change
Five things that shift retention from the exit side to the daily side
None of these require a new programme or a restructure
-
1
Stop trusting exit interviews on their own
Ask instead at six months, at twelve months, and at any handover. People tell you the truth while they still have something to gain from the relationship. Also ask the people who stayed why they stayed, which almost nobody does. -
2
Look at turnover by supervisor, not by site
Site-level numbers average away the signal. Break it down by work front and by leader over twelve months and the pattern is usually obvious within about ten minutes. This is uncomfortable, which is why it is rarely done. -
3
Treat the first ninety days as leadership work
Most avoidable departures are decided in the first three months, before anyone has built a reason to stay. Someone should be accountable for whether a new starter has been shown the standard, given real work, and spoken to properly. -
4
Put retention on the supervisor scorecard
If frontline leaders are measured only on program and cost, leadership behaviour is the first thing sacrificed under pressure. That is not a character failure, it is a rational response to the incentives. Add retention and team feedback and the behaviour follows. -
5
Train the layer that actually decides it
Retention initiatives usually land in HR, one or two levels away from the person who determines the daily experience. Train the supervisors, and give them support afterwards while the habits form.
Every one of these moves the intervention earlier. That is the entire principle.
Where this series ends
This series began with the shift from accidental leader to intentional leader: a supervisor made a leader on a Friday and given a crew on a Monday, and the finding that 94% of fifty construction leaders named the leadership capability gap as a core industry challenge, ahead of mental health, ahead of gender, ahead of safety.
Everything since has been an attempt to describe what closing that gap actually requires. Managing your own pressure so you stop transmitting it. Making success clear before holding anyone to it. Raising things while they are still small. Being firm without becoming personal. Handing over real work and coaching rather than rescuing. Telling the difference between an argument and someone being diminished. Noticing who you give the good jobs to. Responding well the first time someone tells you something hard.
None of it is complicated. All of it is learnable. Almost none of it is currently taught to the people who need it most.
And it all shows up in the same place in the end. Not in an engagement score. In whether the good ones are still here in two years.
One of the CEOs I interviewed, thirty-one years in a tier one contractor, said something at the end of our conversation that I have not been able to improve on. If we keep losing the good ones because we will not change, we deserve the workforce we end up with.
He is right. And the change he is describing is not an industry-wide transformation programme. It is one supervisor, on one work front, doing eight small things differently. That is where retention is built, and it is the only place it has ever been built.
Start where retention is actually decided
Explore EPIC Elevate, two days plus four months of support for the frontline leaders who determine whether your people stay, with up to 80% rebate available through Keystone Tasmania. Read the full evidence in the Construction Leadership Report 2026, or talk to us about your turnover by supervisor.
Be Kind Always.
Frequently Asked Questions
Why do construction workers actually leave?
Rarely for the reason recorded in the exit interview. People are asked why they are leaving at the moment they have least incentive to be honest, because they still want a reference in a small industry, so they give the safe answer of more money or a better opportunity. The real sequence is usually months of small experiences: unclear expectations, unfair decisions, public correction, no growth, and behaviour nobody addressed.
How much does it cost to lose a skilled construction worker?
Around $67,000 per skilled departure once recruitment, onboarding, lost productivity and lost knowledge are counted. Broader workforce research commonly puts replacement cost between 50% and 200% of annual salary depending on the role. Set against the cost of training the supervisor that person reports to, the arithmetic is not close.
Is construction’s retention problem really about pay?
Pay matters and the market is genuinely tight, so it would be unserious to dismiss it. But turnover across the industry is commonly reported at 25 to 30%, and no pay settlement or attraction campaign closes a gap that size. Every dollar spent attracting people in is wasted if the supervisor they land under was never taught how to lead.
Are younger construction workers less committed?
What changed is not resilience, it is tolerance for being led badly. Older generations absorbed more because leaving looked like failing and alternatives seemed limited. Younger workers hold no such belief and in a tight labour market they are correct not to. They are best treated as a fast, accurate sensor for leadership quality who report their findings by resigning.
How can a construction business measure whether leadership is driving turnover?
Break turnover down by supervisor and work front rather than by site, over twelve months. Site-level numbers average the signal away. Then ask at six and twelve months rather than only at exit, and ask the people who stayed why they stayed, which almost nobody does. The pattern is usually obvious within about ten minutes, which is why it is rarely looked at.
