The Rostering Puzzle: Strategic Workforce Planning for Variable Demand
By Tim Fagan, Senior Manager, Trace Consultants
If you’ve ever walked into a morning handover and felt the tension before anyone’s said a word, you’ll know exactly what I mean by the rostering puzzle.
The phones are already ringing. A few people have called in sick. Demand has spiked in a way last week’s roster didn’t predict. Someone’s asked for compassionate leave. A new compliance rule has landed. And the service target everyone is measured on—response time, waitlists, patient ratios, visit windows, incident coverage—doesn’t care that humans are not Lego blocks.
In Australia and New Zealand, this problem has become harder, not easier. Labour markets have been volatile. Costs have moved. Industrial relations and leave compliance are evolving. Australia’s Annual Wage Review increased the National Minimum Wage and modern award minimum wages by 3.5% from 1 July 2025, adding real pressure to already tight service budgets. And in sectors like aged care, award changes flowing from the Aged Care Work Value Case have introduced staged changes and pay rate increases (with key changes taking effect from January 2025 and additional increases from October 2025). Meanwhile, broader workplace reforms under the “Closing Loopholes” laws have rolled out in phases between late 2023 and August 2025, changing the IR landscape that rostering teams operate within.
New Zealand organisations have their own complexity curve. Leave compliance has been a long-running headache for employers, and Cabinet decisions in August 2025 signalled a move toward new employment leave legislation intended to replace the Holidays Act 2003 with a simpler model. (If you’ve lived through Holidays Act remediation projects, you don’t need me to explain why “simpler and more workable” gets attention.)
So here’s the key takeaway I want to land:
Strategic workforce planning and rostering is about meeting service targets while managing the specific complexities of shifts and labour costs in a volatile Australian market—across aged care, hospitals, and other labour-intensive organisations.
This isn’t a “get better at scheduling” pep talk. It’s a practical guide to lifting rostering out of reactive firefighting and into a controlled, measurable capability—without losing the human element that service organisations run on.
Why rostering feels harder than it “should” be
Most organisations treat rostering as an operational function: build the roster, fill gaps, approve swaps, manage overtime, repeat.
But in reality, rostering is the point where three forces collide:
- Demand volatility (what the service needs)
- Workforce supply constraints (who is available, skilled, and safe to deploy)
- Cost and compliance (awards, fatigue, leave rules, budgets, industrial arrangements)
When those three are stable, rostering is straightforward.
When they’re not—and in many sectors they aren’t—rostering becomes a daily puzzle with moving pieces. The “best” roster on paper can fail in the first two hours of the shift if it doesn’t reflect how demand actually behaves.
Demand isn’t just volume. It’s shape.
Aged care demand isn’t evenly spread across a day. Hospitals don’t experience “average” ED arrivals. Facilities teams don’t get breakdowns according to neat weekly patterns. Call centres don’t receive a perfectly smooth call rate. Hospitality doesn’t staff to the average Friday; it staffs to the Friday where a major event is on, plus a bus tour turns up early.
The shape of demand—peaks, troughs, seasonality, event-driven surges—matters as much as the total volume.
Supply is more than headcount.
In labour-intensive services, capacity isn’t “number of people on payroll”. It’s:
- skills and scope (who can do what safely)
- fatigue limits (who shouldn’t do more)
- location and travel (who can get where, and when)
- shift preferences and retention risks (who will quit if this becomes unsustainable)
- compliance (minimum breaks, maximum hours, award rules)
- reliability (availability, sickness trends, unplanned leave)
If you’ve got the headcount but the wrong skill mix, you still don’t have capacity.
Cost is no longer just “overtime vs base”
Cost is now a broader system problem. Wage increases and award changes flow through base rates, allowances, penalties, classifications, and career structures. In some sectors, cost pressures are compounded by agency reliance, backfill practices, and the hidden cost of turnover. And reforms that alter the employment landscape can shift how organisations think about casualisation, labour hire, and workforce models.
The difference between rostering and workforce planning
A useful distinction:
- Rostering is the weekly (or fortnightly) act of allocating people to shifts.
- Strategic workforce planning is the ongoing discipline of ensuring you have the right workforce capacity, composition, and deployment model to meet demand—now and into the future.
If you only roster, you’re forever reacting.
If you plan strategically, rostering becomes the execution arm of a deliberate design.
In practical terms, strategic workforce planning answers questions like:
- What demand do we need to meet, and how variable is it?
- What service targets are non-negotiable (and which are negotiable)?
- What workforce mix do we need (full-time, part-time, casual, agency, contractors)?
- Where are the structural bottlenecks (skills, locations, time windows)?
- What should “good” look like in terms of overtime, leave cover, and utilisation?
- Which constraints are real, and which are self-imposed by legacy policies?
Then rostering becomes less about guesswork and more about optimising within known rules.
A practical framework for variable demand rostering
When demand is volatile, you don’t win by chasing perfection. You win by building a system that is resilient, measurable, and fair.
Here’s a framework we use across labour-intensive environments.
1) Start with the service promise (and be brutally clear)
Service organisations often carry a quiet contradiction:
- “We will meet our service targets.”
- “We must reduce labour costs.”
- “We must protect wellbeing and retention.”
You can pursue all three, but only if you define what “service targets” actually mean.
Examples of service promises:
- Aged care: visit windows, continuity of care, safe ratios, response times for incidents
- Hospitals: safe staffing, surge capability, theatre utilisation, ward coverage, ED flow
- Local government: response times for critical maintenance and emergencies
- Facilities: compliance coverage, after-hours readiness, fault response
- Contact centres: answer times, abandonment rates, quality scores
Once you define the promise, you can translate it into demand drivers and coverage requirements.
If you can’t articulate the service promise in plain language, your roster will always be a compromise that nobody owns.
2) Understand demand properly (not just averages)
Most rosters are built off historical averages or manager intuition.
That’s not enough when demand is variable.
Better demand understanding includes:
- peak day vs average day analysis
- intra-day patterns (hourly shape, not daily totals)
- event impacts (public holidays, school holidays, major events, planned outages)
- seasonal shifts
- lead indicators (referrals, bookings, call volumes, admissions forecasts)
- “failure demand” (rework and repeat tasks caused by poor first-time resolution)
A useful question to ask: What proportion of today’s work was predictable a week ago?
If the answer is “most of it”, your issue might be process, not forecasting. If the answer is “hardly any”, your workforce model likely needs more flexibility and faster redeployment mechanisms.
3) Translate demand into workload, and workload into staffing
Demand signals only become useful when converted into workload.
Workload is the combination of:
- task volumes
- task durations
- travel time (where relevant)
- variability (standard deviation matters)
- non-productive time (handover, admin, safety checks)
- escalation overheads
From there, staffing needs can be modelled by:
- coverage requirements (minimums, ratios, “always-on” coverage)
- productive hours per shift after breaks
- skill mix rules
- fatigue and safety rules
- allowance for unplanned absence (shrinkage)
This is where a lot of rosters fall over: they roster to headcount, not workload. The result is either chronic understaffing (and overtime) or chronic overstaffing (and frustration about cost).
4) Design a workforce mix that matches volatility
Variable demand is not solved purely by “better rostering software”. It’s solved by aligning the workforce model to demand characteristics.
Common levers include:
- more flexible part-time structures (instead of forcing full-time patterns everywhere)
- planned “float” roles (not as a punishment, but as a designed capability)
- split shifts or staggered starts (where appropriate and agreed)
- internal banks/pools for backfill
- cross-skilling to increase redeployability
- clearer rules for overtime and call-in (to avoid ad hoc inequity)
- agency use as a last resort, with controls and learning loops
In aged care and hospitals, this can also include explicit surge capacity models—because pretending surges won’t happen doesn’t make them disappear. It just shifts the impact onto people at the worst possible moment.
5) Build rosters around constraints, but challenge the right ones
Constraints are real in service environments. The trick is distinguishing:
- constraints that protect safety and fairness, and
- constraints that exist because “we’ve always done it that way”.
Common constraints worth testing:
- overly rigid shift start times that don’t match demand peaks
- “everyone rotates equally” rules that ignore skill needs
- blanket rules about weekends/nights that drive perverse outcomes
- cumbersome approval processes for swaps and minor adjustments
- policies that force managers to hoard staff “just in case”
- lack of clarity on what can be flexed during surge conditions
When constraints are unclear, rosters become negotiation documents instead of operational tools.
The rostering realities in aged care
Aged care rostering is particularly complex because demand is both:
- predictable (regular care needs), and
- unpredictable (incidents, acuity changes, staff absences, family needs)
Add to that:
- workforce shortages in many markets
- wage and award changes flowing through classifications and pay structures
- heavy compliance load and documentation requirements
- high emotional labour and burnout risk
- continuity of care expectations (relationships matter)
In this environment, “just add casuals” isn’t a strategy. It can erode continuity and increase training burden.
What tends to work better:
- designing a stable core roster that covers predictable workload
- building a structured flexibility layer (bank/pool, planned floats)
- using demand signals to plan rather than react (e.g., known high-risk periods)
- simplifying admin so frontline staff spend time on care, not paperwork
- making fairness explicit (so overtime and unpopular shifts don’t become political)
There’s also a wider funding and compliance context. Government guidance has been issued to support providers in implementing wage increases linked to the Aged Care Work Value Case. This is a reminder that rostering is not just “operations”—it’s tied directly to funding, compliance, and workforce sustainability.
The rostering realities in hospitals
Hospitals are where rostering meets complexity at speed.
Demand is volatile (ED arrivals, bed flow, theatre overruns), and staffing is constrained by:
- strict skill requirements
- patient safety and minimum coverage needs
- fatigue and overtime limits
- industrial arrangements and agreements
- the operational reality that handovers, breaks, and supervision take time
The common trap is to treat rosters as fixed artefacts. But in hospitals, the real game is dynamic capacity management—knowing what levers exist when the day deviates from plan.
Effective approaches often include:
- rosters built with explicit surge rules (what triggers escalation, who is contacted, what gets deferred)
- better use of predictive indicators (bed management, elective schedules, known demand drivers)
- clear governance around overtime and additional shifts (safe, fair, and consistent)
- skill-based allocation rather than purely “equal rotation”
- realistic staffing models that include non-patient-facing workload
And importantly: a shared view of demand, workforce, and constraints—not separate spreadsheets living in silos.
“Other labour intensive organisations”: where the same puzzle shows up
The rostering puzzle isn’t limited to healthcare.
You see the same structural dynamics in:
- disability and community services (variable demand, travel, skills)
- hospitality and venues (event-driven peaks, late changes, casual workforce)
- manufacturing and processing (shift coverage, overtime, fatigue, absenteeism)
- warehousing and logistics (seasonality, cut-off times, volume spikes)
- facilities and maintenance (reactive work + compliance coverage)
- contact centres (volume variability + quality requirements)
- emergency services support functions (readiness + cost control)
The specifics differ, but the underlying challenge is the same:
How do we meet service targets when demand moves—and do it without blowing out labour cost or burning out the workforce?
The cost side: labour costs aren’t just “too much overtime”
Overtime is the visible symptom, but the underlying cost drivers are usually structural:
- mismatch between roster shape and demand shape
- poor shrinkage planning (sick leave, training, admin time)
- excessive reliance on agency due to lack of internal flexibility
- rework and inefficiency that inflate workload
- turnover that creates permanent training overhead
- fragmented scheduling practices across sites/teams
- lack of real-time visibility, leading to “just in case” staffing
Wage increases and award changes amplify these issues. A 3.5% wage increase flows through every hour you roster. If your roster design is inefficient, that inefficiency just got more expensive.
This is why strategic workforce planning matters: it targets the system, not just the symptom.
Technology: helpful when it supports decisions, harmful when it replaces them
Rostering technology can be powerful. It can also become shelfware if it’s not aligned to the operating model.
Good rostering tech supports:
- demand-based staffing models
- skill and credential constraints
- fair allocation rules
- shift bids/swaps with governance
- fatigue management and compliance checks
- real-time coverage visibility
- reporting that managers actually trust
But technology won’t fix:
- unclear service targets
- broken handovers
- inconsistent role expectations
- lack of accountability for staffing decisions
- policies that incentivise gaming
- poor data definitions (e.g., what counts as “productive time”)
In New Zealand, leave complexity has been a recurring pain point for payroll and rostering, and reforms are explicitly trying to reduce uncertainty and compliance burden. That context matters when selecting systems and designing processes—because “leave” is never just an HR function in a shift-based environment. It’s a rostering constraint that affects coverage and cost every day.
The human side: fairness is a performance lever
Rostering discussions often get stuck on numbers: hours, FTE, overtime cost.
But fairness is a major driver of performance, retention, and discretionary effort.
If staff believe the roster is:
- inconsistent,
- biased,
- opaque, or
- constantly changed without consultation,
you’ll pay for it in absenteeism, turnover, and conflict.
Fairness doesn’t mean “everyone gets the same”. It means:
- the rules are clear,
- the reasoning is understood,
- the trade-offs are transparent,
- and people feel respected.
In practical terms, that can look like:
- published rules for overtime allocation
- transparent shift swap processes
- clear escalation triggers for surge rosters
- genuine consultation on roster changes
- rostering decisions supported by data, not favouritism
A “fair enough” roster is often more sustainable than a theoretically optimal one that nobody trusts.
A simple maturity model: where are you today?
If you want a quick self-check, consider these stages:
Stage 1: Reactive rostering
- built on intuition
- constant last-minute changes
- overtime and agency are default fixes
- limited visibility
Stage 2: Rules-based rostering
- basic compliance is embedded
- shift templates exist
- still relies heavily on manual intervention
- demand is loosely understood
Stage 3: Demand-informed rostering
- demand patterns drive shift design
- workload modelling exists
- shrinkage is planned
- clear governance on changes
Stage 4: Integrated workforce planning
- workforce mix is designed for volatility
- skills and training pipelines align to demand
- scenario planning is routine
- technology supports decisions end-to-end
Most organisations sit somewhere between Stage 1 and Stage 3. The good news is you don’t need to jump straight to Stage 4 to see benefits. Often, the biggest gains come from moving one stage forward with discipline.
What “good” looks like: measurable outcomes that matter
Rather than promising magic, it’s better to focus on outcomes you can track credibly:
Service outcomes
- coverage reliability
- response times / waitlists
- missed visits / delayed jobs
- patient or client experience measures (where applicable)
Workforce outcomes
- unplanned absence rates
- overtime hours (and distribution fairness)
- roster stability (how often it changes last minute)
- retention signals and turnover
Cost outcomes
- total labour cost per unit of service
- agency usage and reasons
- premium hours (penalties, overtime) as a proportion of total
- rework cost (repeat visits, repeat calls)
These measures help you move the conversation from “rostering feels bad” to “here’s what’s driving it, and here’s what we’re changing”.
How Trace Consultants can help
Rostering is one of those capabilities that touches everything—service, finance, people, compliance, technology. That makes it easy for it to fall between organisational cracks.
Trace Consultants helps clients solve the rostering puzzle by bringing structure to that complexity—without losing sight of the realities on the floor, in the ward, on the road, or on the phones.
Here’s how we typically support organisations across Australia and New Zealand:
1) Demand and workload analysis for variable environments
We help you understand demand patterns properly (including peaks and volatility), translate demand into workload, and build staffing models that reflect reality rather than averages.
2) Workforce mix and operating model design
We assess how your workforce is currently structured (full-time/part-time/casual/agency), identify where rigidity is creating cost or service risk, and design a mix that can absorb volatility more sustainably.
3) Rostering governance, rules, and fairness
We help define clear rostering principles—overtime allocation, surge triggers, swap rules, approval pathways—so rostering becomes consistent and defensible, not a weekly negotiation.
4) Technology selection and enablement (without the shelfware)
We support requirements definition, vendor evaluation, and implementation planning in a way that prioritises adoption, usability, compliance, and measurable outcomes—rather than feature lists.
5) Benefits tracking and continuous improvement
We establish performance measures tied to service outcomes, workforce wellbeing, and cost-to-serve, then embed routines so improvements stick beyond the project.
We don’t make up numbers. We don’t sell silver bullets. We help you build a workforce planning and rostering capability that matches the complexity of your service environment—and holds up on Monday morning, not just in a PowerPoint.
Final thought: the puzzle doesn’t disappear, but it can become solvable
If demand were stable, rostering would be admin.
But demand isn’t stable—and in Australia and New Zealand, the cost and compliance environment adds an extra layer of pressure. Wage changes, sector-specific award adjustments, and broader reforms mean labour decisions carry more weight than ever.
The goal isn’t to build the “perfect roster”. The goal is to build a system where:
- service targets are clear,
- demand is understood,
- staffing is modelled realistically,
- workforce flexibility is designed (not improvised),
- fairness is visible,
- and decisions are supported by good data and governance.
That’s when the rostering puzzle stops feeling like chaos—and starts feeling like a capability.
If you’re ready to move from reactive rostering to strategic workforce planning, Trace Consultants can help you define the pathway and deliver it in a way that suits your workforce, your customers, and your budget.
