Absenteeism Was Not the Problem. Not Knowing Where It Sat Was.
It is 6:15 AM on a Monday. Three support workers call in sick across two SIL houses, and you have ninety minutes to fill them. At one NDIS provider, a My Offshore Staff Business Analyst noticed those calls were not landing randomly. He pulled in his rostering colleague and went looking.
Multi-site NDIS provider, Australia. Client name withheld. Figures are organisation level only.
The provider delivers Supported Independent Living (SIL) and Specialist Disability Accommodation (SDA) across multiple sites, alongside personal care, in-home supports and positive behaviour support. As it grew, unplanned absence stopped being an HR issue and became an operational one: broken rosters, agency backfill, continuity of support at risk.
The data was already there, every call-out and every backfill, sitting in the rostering and payroll systems. Nobody had analysed it, because filling a shift and working out why it came free are two different jobs, and only the first one was anybody’s problem on the day.
So the Business Analyst and the Rostering Officer, both already embedded in the provider’s operations, worked through 48,000 shifts across the first half of the year.
Absence was not spread evenly across the workforce. It was concentrated, predictable, and clustered into windows you can plan around. Ten people, two days, one three-hour window. None of it random, all of it plannable.
What MOS Did
Absenteeism usually lands on HR. Treating it as a rostering problem with an HR tail changed the fix.
The Business Analyst broke absence down by day, call-out time, site and individual, so the provider could see where it actually sat. The Rostering Officer turned that into a standing Rapid Response Team, critical rosters anchored around the most reliable staff, and check-in triggers that fired before anything formal did. Attendance then moved onto a standing leadership dashboard. The provider’s team owned and ran the changes; ours supplied the analysis, the roster design and the monitoring.
Most emergency cover is spread evenly across the day. Demand here was front-loaded, so three of the provider’s best attenders were shifted to match where the calls came.
Sizing it was the easy part. Making it work across a spread-out SIL portfolio took three rules:
- Roster on real travel time, not map distance. Cluster the houses by proximity, assign each responder the cluster they reach fastest under actual traffic, and keep long cross-suburb runs out of peak hours.
- Match skill to house. Complex participant needs get higher-skilled responders. Filling a gap is not the same as covering it.
- Anchor on reliability. Both shifts built around staff with proven attendance in emergencies.
What Improved in Three Months
The changes went in at the start of the second half. Three months later, every measure had moved, and the biggest shift came where the fix was most direct: the 6 AM to 9 AM window the Rapid Response Team was built to absorb. These are early numbers, not a finished result, so the provider set twelve-month targets and now tracks them continuously.
Each figure is that segment’s share of total sick leave hours, except engagement and satisfaction, which are survey scores. Change is stated in percentage points. Baseline drawn from six months of rostering and payroll data. Targets are internal goals agreed with the provider, not guaranteed outcomes.
Do you know when your sick calls actually come in?
Find out in 30 minutes →Wider Outcomes Across the Second Half
Several things changed at this provider that year, so we are not putting all of this down to the rostering work. But across the full second half against the first:
- Delivery rose while the workforce stayed the same size. Service hours up 3.2% with headcount flat, and total overtime down 15%. More hours delivered, the same number of people, less expensive cover.
- Overtime was concentrated, exactly like the sick leave. A small number of area managers carried a disproportionate share, pointing at roster design and vacancy cover in particular portfolios rather than an organisation-wide problem.
- Workforce stability improved. Terminations down 12.5% while hiring rose.
These come from payroll and HR systems, not the rostering data: a separate set of records, kept by different people for different purposes, moving the same way.
What to Do If Your Rosters Look Like This
- The data is already in your system. The problem is that nobody has the hours to read six months of it properly.
- Segment before you act. Ten people drove 13.5% of sick leave hours. A blanket policy would have missed them and annoyed everyone else.
- Size emergency cover to the demand curve. Even coverage wastes capacity when 62% of call-outs land in one eight-hour block.
- Watch overtime and delivered hours, not just absence. Absence rate alone will not tell you whether the fix worked. Hours delivered per head and overtime concentration by manager will.
Questions We Get Asked
Any shift a worker cannot attend without enough notice to fill it normally, usually a sick call. It matters more in SIL and in-home supports than in most industries, because the participant needs the support regardless, so an unfilled shift is a continuity of support risk rather than a productivity loss.
In this provider’s data, 25% of sick leave hours fell on a Monday and 22% on a Friday. Clustering at the ends of the week is common across shift-based workforces. The useful point is not why it happens but that it is predictable, so emergency cover can be scheduled for it rather than improvised.
Analyse your existing rostering and payroll data to find where absence is concentrated, by day, call-out time, site and individual. Then size emergency cover to that demand curve rather than spreading it evenly, anchor critical rosters around staff with proven attendance, and trigger supportive check-ins before formal processes. Track it on a standing dashboard.
Often, but not automatically. Unplanned absence is usually covered by overtime or agency staff, so when call-outs become predictable and cover is sized to match, overtime tends to fall. At this provider it fell 15% while service hours rose 3.2% and headcount stayed flat. Overtime also reflects vacancy levels and how work is spread between managers.
Builds and maintains rosters, fills unplanned gaps, manages availability and compliance requirements such as SCHADS conditions, and coordinates emergency cover. In a well-run service they also design the roster structure, deciding how cover is sized and shifted against actual demand rather than filling gaps reactively.
Yes. Rostering coordination, workforce analytics, attendance reporting and dashboard maintenance are rules-based and need consistency rather than someone on site. Participant-facing support cannot be offshored; the coordination and analysis behind it can. My Offshore Staff places rostering officers and business analysts into Australian NDIS providers to do exactly this.
Put a rostering officer on your data
We place offshore rostering officers, business analysts and workforce coordinators into Australian NDIS providers. Book a free consultation and we will scope what an attendance analysis would surface in your own rosters.
Book a free 30-minute call →Related: NDIS offshore support · How offshore staffing works · All services