The Coles underpayment bill landed in August 2026: $235 million, set aside to repay the people who run its stores. In the same results, Coles reported a $1.1 billion profit.
For years, salaried store managers at Coles worked under annual salaries meant to cover their overtime and penalty rates. A Federal Court found that arrangement left many of them underpaid.
The 2026 statistic every retail leader should read
DHR Global’s 2026 Workforce Trends Report found retail has the highest rate of moderate to extreme burnout of any industry it studied, at 62%. Across all sectors, the top two causes are overwhelming workloads (48%) and working too many hours (40%).
There’s a gap at the top, too. 62% of associates say burnout drags down their engagement. Only 38% of C-suite leaders say the same.
The people doing the extra hours feel it. The people setting the salaries often don’t.
Why this happens, and when
It happens when a salary becomes a ceiling on what an organisation notices. Once someone is on a fixed annual package, extra hours stop showing up anywhere leadership looks. The work still gets done. The cost just moves onto the person.
It usually starts in the busiest periods. A manager covers a shift, stays late for stocktake, comes in on a public holiday. Nobody logs it because nobody has to. After a few years, it’s a pattern. After a few more, it’s a court case.
What happened in the Coles underpayment case
Here is what has been publicly reported, with sources.
The Fair Work Ombudsman and a class action took Coles, and separately Woolworths, to court over salaried store managers on annual salaries meant to cover penalty rates, overtime, allowances and loadings under the retail award (Squire Patton Boggs, September 2025).
Coles had already tried to fix the issue, paying over A$7 million in remediation backdated to 2013. The regulator and class action applicants argued this wasn’t enough (Squire Patton Boggs, September 2025).
In September 2025, the Federal Court found the employers had not tracked or paid award entitlements like overtime and penalties in each pay period, and had relied on set-off clauses instead (BDO, 2026).
On 25 August 2026, Coles reported a $1.1 billion full-year profit, including a $235 million provision to cover remediation costs linked to the judgment (ABC News, 25 August 2026).
Coles’ response
Coles said it recorded the $235 million in significant items ($165 million after tax) as a result of the September 2025 Federal Court judgment in the Fair Work Ombudsman’s proceedings. The Epoch Times reported that Coles also cut short-term incentives for executives it held accountable. Before the ruling, Coles had made remediation payments and relied on its salary arrangements.
The MoodyBot probability model: what Coles could have saved
MoodyBot, the AI layer inside Moody At Work, reads anonymous daily mood check-ins across stores and teams. It isn’t a payroll system and can’t calculate award entitlements. What it tracks is how people feel at work, every day, without their names attached.
In a business like Coles, the signals MoodyBot could have surfaced include sustained exhaustion scores among salaried store managers compared with hourly staff, check-in reasons clustering around long hours, unpaid extra time or feeling undervalued, falling mood in peak trading periods, and a gap between how store managers and head office feel about workload.
Those signals point at exactly the problem the court later found: salaried people working more hours than their salary covered. Spotting that earlier could have prompted a payroll review years before the regulator did.
| Risk | Publicly reported loss | Model assumptions | What MoodyBot could have saved |
|---|---|---|---|
| Underpayment provision | A$235m (A$165m after tax), FY26 | 10% to 25% avoidable through earlier action (most of the provision is back pay Coles always owed staff); 30% to 50% chance early signals lead to action | A$7.1m to A$29.4m |
| Penalties | The court has yet to decide | Not yet known | Not quantified |
| Executive accountability | Short-term incentives reportedly cut | Internal cost, not a saving | Not counted |
| Legal and remediation work | Years of proceedings and a class action | Not publicly disclosed | Not quantified |
| Total quantified | A$7.1m to A$29.4m |
How the numbers were built: Most of the A$235m is back pay Coles always owed its managers, so the model doesn’t count it as a saving. It treats only 10% to 25% as avoidable, covering interest, legal costs and years of the liability growing. A$235m x 10% x 30% gives A$7.1m. A$235m x 25% x 50% gives A$29.4m. The bigger win is harder to put a number on: managers paid properly, sooner.
What every retail leader can take from the Coles underpayment
Salaried doesn’t mean invisible. If your store managers are exhausted, the hours they’re working are real, whether or not your payroll system records them. The sooner you know, the smaller the bill.
Silence costs money in other ways too, as we saw with the KPMG Australia whistleblower scandal.
Frequently asked questions
Why did Coles set aside $235 million? Coles recorded a $235 million provision in its 2026 results to cover remediation costs after a September 2025 Federal Court judgment found it had underpaid salaried store staff.
Which industry has the highest burnout? According to DHR Global’s 2026 Workforce Trends Report, retail has the highest rate of moderate to extreme burnout at 62%.
How can employers spot overwork in salaried staff? Anonymous daily check-ins surface exhaustion and workload pressure that payroll systems miss. Moody At Work turns those signals into early warnings leaders can act on.
Sources: DHR Global, Workforce Trends Report 2026; ABC News (25 August 2026); Squire Patton Boggs (September 2025); BDO (2026); The Epoch Times (25 August 2026).
All factual claims are sourced from publicly available reporting. Coles’ representatives have stated that the $235 million provision was recorded as a result of the September 2025 Federal Court judgment in the Fair Work Ombudsman’s proceedings. MoodyBot projections represent an independent illustrative probability model based on publicly reported information and are not guaranteed outcomes.