A schedule cost turnaround.
A rising roster cost brought back to its funded line, removing around $240K per month in unfunded labour without touching service quality.
A community care provider was building its roster well above what care plans funded, at the year's peak unit cost. Human Nexus measured the gap in the provider's own data, brought delivery back inside the funded envelope, and stood up an operating model that holds it there. Over servicing fell from 12.8% of funded hours to almost nil, with overtime never above 2% of wages.
Anonymised. Client, funder, and system names are removed. Figures are drawn from the provider's operational data over the engagement window, and each ties back to a single reproducible query.
- Window
- December 2025 to May 2026
- Scope
- Whole organisation, all rosters
- Attribution
- Prepared by Human Nexus
A roster realigned to what it is funded to deliver.
over servicing as a share of funded hours, cut to almost nil
unfunded labour removed from the run rate
loaded unit labour cost, held below the December peak
unassigned rostered hours per month, idle capacity converted toward billable
overtime as a share of wages, throughout
agency share of hours, a non lever ruled out in the data
The schedule was costing more than it was funded to.
The provider was rostering close to 13% more care than care plans funded, in a single month. That gap is labour with no funding behind it, and at the time it was being delivered at the year's peak unit cost.
The cause was structural, not a one off. The roster had drifted above the funded envelope and was filling reactively, which pulled in overtime, rest break breaches, and premium open shift cover. On a delivery model that runs on a thin margin, a gap like that moves the result almost dollar for dollar.
It was also invisible in the day to day. Funded hours, rostered hours, and the cost to deliver lived in separate systems, so no one number told the operator how far above funding the schedule actually sat, or what it was costing.
Four moves, each tied to the same number.
We did not cut services or chase penalty hours. We found the gap in the provider's own data, matched supply to what was funded, and made the alignment something the operating model maintains on its own rather than a one time clean up.
Align the roster to funding.
We measured funded versus rostered hours for every roster, every period, and brought the schedule back inside the funded envelope. Funding utilisation became the number the whole engagement is tied to.
Take out the penalty layer.
We moved fills from emergency to planned, which removes the rest break, overtime, and premium open shift loadings that reactive rostering creates. The saving comes from the model, not from cutting hours.
Use the capacity already paid for.
Unassigned rostered hours were converted toward real, billable demand instead of being carried as idle cost. Unassigned hours fell from 2,052 to 452 per month.
Make it self correcting.
We stood up continuous checks on roster to funding, cost movement, and award compliance over one reconciled dataset, so drift is caught before it reaches payroll.
The gap is almost always there. We find it in your own data.
Over servicing fell from 12.8% to 0.2%, and held.
Over servicing all but removed.
Over servicing fell from 12.8% to 0.2% of funded hours, so the roster now sits on its funded line rather than 12.8% above it. The turn follows the point we started work, and the line holds rather than spiking back.
Around $240K a month in unfunded labour removed.
The roster hours above funding, valued at the loaded rate, taken out of the monthly run rate as the gap closed.
Unit labour cost down about 9%.
Loaded cost per paid hour settled below the December peak from January onward and held there.
Idle capacity converted.
Unassigned rostered hours fell from 2,052 to 452 per month, moving cost that delivered nothing toward billable demand.
The penalty layer came down.
Overtime stayed under 2% of wages throughout, and the rest break and premium open shift cover that reactive filling creates was reduced as fills moved to planned.
Agency proved a non lever.
Agency ran under 0.5% of hours, so there was no rate arbitrage to chase. The gain sits in how permanent labour is deployed, a useful thing to know for certain.
A model that corrects itself.
Roster to funding, cost movement, and award compliance are now checked continuously over one reconciled dataset, so the gains stay banked instead of drifting back.
Held, not a one off.
This was an operating change rather than a clean up. The same data that runs operations runs finance and reporting, so drift is caught before it reaches payroll.
Three decisions made the difference.
Align to funding, not blanket cuts. We matched supply to what care plans funded rather than trimming services, so quality was never the lever. The roster came down to the funded line because that is where the funding was, and overtime stayed under 2% throughout.
Planned, not reactive. Reactive filling is what creates the penalty layer, the overtime, rest break breaches, and premium open shift cover. Moving fills to planned removed the loadings at the source instead of paying for them and explaining them later.
One source of truth, always on. Funded hours, rostered hours, cost to deliver, and compliance now reconcile in one dataset that runs operations, finance, and reporting from the same numbers. The model catches drift on its own, which is why the result holds rather than fading once attention moves on.
The provider asked us to bring a rising schedule cost under control. What they have is a roster that sits on its funded line by default, a unit cost that holds, and an operating model that keeps it that way.
Is your roster costing more than it is funded to?
We find the gap in your own data, fix the delivery model, and prove it with a number. Tell us about your roster and we will get back to you within one business day.
See more case studiesBook a cost review
Tell us about your roster and funding. We will reply within one business day.