
Where margin actually leaks in a group home, and how to close it before it reaches your claim file.
Most SIL providers donât lose money on big decisions. They lose it in small ones: a sleepover rostered where an active night was needed, a shift staffed at 1:2 when the plan funds 1:3, a room that sat empty for eleven days, a group shift split three ways when only two participants were actually there.
This playbook is for providers already running a SIL house, or about to open one. It covers where the money goes, how to build a roster from the plan instead of the group chat, how to claim group shifts under the 2026 rules, and the evidence chain an auditor follows from a shift to a claim line. There are no definitions of SIL or SDA here. For the software-selection view, read our SIL software features guide; for how households, group sessions and assets sit on one timeline, see SIL and group supports rostering.
Key takeaways
Every SIL operation runs on the same three measures. If you canât produce all three from your software this week, fix that before anything else on this page. (Our wider list of KPIs every NDIS provider should measure sits around these.)
Utilisation: funded support hours delivered as a percentage of available hours. When it sits well below your target, you are carrying staff cost the plan isnât paying for.
Ratio adherence: how often the shift ran at the ratio the plan funds, in both directions. A shift funded at 1:3 but staffed at 1:2, or funded at 1:2 but staffed at 1:1, is wage cost you absorb, because the funded arrangement sets what the plan pays. A shift staffed thinner than funded is a delivery shortfall you will have to explain.
Claim reconciliation: the percentage of verified shifts that became clean claim lines without manual correction. When it is low, someone is rebuilding the same data twice.
None of these are finance metrics. They only become finance problems a month later, when the claim is short and nobody can say why. Set your own targets from a three-month baseline; there is no NDIA benchmark for any of them.
Not in the price limits. In five operating habits that look normal until you put them on a timeline.
The Roster of Care is the contractual picture of support: which worker, at which ratio, in which window. Every provider starts with it. Most drift away from it within a quarter: a swap here, a shorter Sunday there, a participant who prefers mornings now. Each change is reasonable. Together they mean the roster you run is no longer the roster you quoted, and the difference is rarely reconciled before claiming. We cover the day-to-day causes in NDIS roster management: 7 common problems.
Rule of thumb: every departure from the pattern needs a written reason attached to that shift, not a verbal agreement in the office. If nobody can reconstruct why a shift looked different eight months ago, neither can an auditor.
A sleepover and an active night can look identical on a roster and are opposites on a cost sheet. Book a sleepover where the participant needs waking support and you get two problems: a quality risk you will eventually have to explain, and a claim that doesnât match the support actually delivered. Book an active night where a sleepover was funded and you have paid awake wages out of a sleeping budget.
Cancellations, early finishes, transport that took longer than planned, a participant in hospital for four days. The shift was in the system, the hours were planned, and the support didnât happen. If your roster and your verified delivery live in different tools, nobody notices until the claim is rejected or the plan is exhausted early and no one can explain where the money went. GPS-verified time tracking closes this gap; our time tracking compliance guide explains what evidence to capture.
One worker, three participants in the same house. The cost is shared. The claims are not: each participantâs plan funds their share of that workerâs time, at the arrangement their plan specifies. Split the shift evenly by headcount and you will over-claim one plan and under-claim another. Over time, the over-claimed plan hits its budget early and the participant loses support; the under-claimed plan quietly funds less than it should.
An empty room still costs rent, utilities and the staffing you canât flex down. It also canât be recovered from anyone else: the NDIA says it doesnât consider vacancy costs when making SIL funding decisions, and providers must not increase the price of supports or claim additional funds from other participantsâ plans to cover vacancy periods. Most providers know their vacancy rate. Far fewer track void days, the actual days a room sat empty per month. Multiply those days by the roomâs fixed daily cost and the number stops being abstract.
The fastest way to stop leaks one to four is to change the order in which you build the week. (More on the usual failure points in common NDIS rostering mistakes.)
This is also the difference between a roster that survives an audit and one that doesnât. An auditor isnât looking for a perfect week. They are looking for evidence that you knew what the plan said, and that when practice departed from it, you noticed and recorded why.
This single call changes both your wage bill and your claim. It should be a documented decision per participant, made at plan review, not a default applied by whoever builds the roster.
| Factor | Sleepover | Active night |
|---|---|---|
| Worker status | Present overnight, sleeping, available if needed | Awake and actively supporting |
| Typical trigger | Overnight presence for safety or reassurance | Regular overnight intervention: medication, health monitoring, behaviour support |
| Wage profile | Lower: sleepover allowance basis (confirm against the current SCHADS Award) | Higher: awake hours worked |
| Main risk | Used where active support was actually needed | Used where only presence was funded |
| What it needs | A written rationale in the support plan | Evidence of the interventions delivered, per shift |
Whichever applies, the claim has to match the support actually delivered, and the note has to explain what happened during the window. A sleepover note that says only âsleptâ is not evidence of anything; a note that records the two times the participant was checked and what was observed is.
Group shifts are where the arithmetic gets uncomfortable, because three things have to agree at once: the workerâs verified hours, each participantâs funded arrangement, and the claim line per participant.
The shift is the source of truth, not the invoice. Verified start and finish times, with location evidence, should drive every claim line in the batch.
Allocate by funded arrangement, not by headcount. If one participantâs plan funds 1:1 overnight and the other two are funded for shared support, an even three-way split is wrong for all three.
Use the right code. For SIL delivered from 1 July 2026, claim lines use support items in registration group 0138; delivery before that date stays under 0115 (NDIS: SIL claims and payment changes). Keep the two apart in any batch, and check quotes and service agreements reference the current items.
Flag the exceptions, donât absorb them. A shift where a participant was absent, hospitalised or supported by a second worker needs to surface before the batch is lodged, not as a rejection three weeks later.
Claim inside the window. From 1 December 2026 a claim must be made within 90 days of the support being delivered, not two years. Month-end claiming leaves little margin for a rejected line. Weekly reconciliation of exceptions is the safer rhythm; see how to reduce NDIS claim rejections and the NDIS claims process explained.
Keep the audit trail attached. Every claim line should point back to the shift, the worker, the location evidence and the note. If those four canât be produced together, the claim is a guess.
Providers who do this well still reconcile weekly. The difference is that reconciliation is a review of exceptions, not a rebuild of the whole week.
Void days are usually discovered in the monthly accounts, which is far too late to influence them. The lead indicators are operational and visible weeks earlier. (Forecasting demand against supply is covered in NDIS workforce planning.)
Track void days per room per month as a standing number, next to utilisation. It changes the conversation from âwe have a vacancyâ to âthis room has cost us this much since Marchâ, which is the sentence that gets a referral pipeline built.
An audit of a SIL house is rarely a surprise inspection of a single shift. It is a walk along a chain. The auditor picks a participant, picks a date, and follows the thread to the claim. Every link has to hold. Use this table alongside our audit preparation checklist.
| Link in the chain | What it proves | How it usually breaks |
|---|---|---|
| Service agreement and plan | The support was funded and agreed | Agreement signed but not updated after a plan review |
| Roster | Who was scheduled, at what ratio, when | Roster edited after the fact with no change history |
| Shift verification | The worker was actually there | Paper timesheets, or GPS without a matching note |
| Progress note | What support was delivered | Generic notes that could describe any participant on any day |
| Medication record | Medication was administered as prescribed | PRN given without a recorded indication |
| Incident record | Issues were identified and acted on | Incident logged, no follow-up or closure date |
| Claim line | The money matches the support | Claim built from the roster, not from verified delivery |
Read that table as a checklist for your next internal review. Most findings arenât caused by one broken link. They are caused by two links that never quite matched, unnoticed for a year.
SIL moved from the lighter verification route onto the certification pathway, and the claiming rules changed with it. The dates below are the ones that touch your roster and your claim file. Confirm current detail on the NDIS Commission reform hub before acting on any of them.
| When | What changed | What it means for you | Source |
|---|---|---|---|
| 1 July 2026 | Mandatory registration for SIL providers begins. New registration group 0138 and the SIL Practice Standards take effect. | Certification audit is the route in. Evidence has to exist from the start of the cycle. | NDIS Commission |
| 1 July 2026 | SIL delivered from this date is claimed under new support items in group 0138. Earlier delivery stays under 0115. | Check quotes, agreements and claim templates. Keep pre- and post-July lines apart in a batch. | NDIS |
| Next audit after 1 July 2026 | Providers already registered under 0115 are audited against the new SIL Practice Standards at their next mid-term audit or renewal. | Treat the next audit as a SIL Practice Standards audit, not a repeat of the last one. | NDIS Commission |
| 1 October 2026 (passed) | Unregistered providers who delivered SIL before 1 July had to apply to keep delivering. Plan managers reject SIL invoices from providers that are neither registered nor applied. | Registration status is now a payment issue, not only a compliance one. | NDIS |
| 1 December 2026 | Claims must be made within 90 days of the support being delivered, replacing the two-year window. | Reconcile and lodge weekly. Unclaimed verified shifts need to be visible and aged. | PWDA summary |
The practical implication is timing, not paperwork. Whether you are mid-term or renewing, evidence collection has to be a monthly habit from the start of the cycle, not a project that begins two months before a visit.
Answer these honestly. Each one you canât answer from your current system is a leak youâre carrying.
| Question | If the answer is no |
|---|---|
| Can you produce funded versus delivered hours for any participant, for any month, in under five minutes? | Utilisation is being estimated, not measured |
| Does every shift carry its planned ratio and its actual ratio? | Ratio adherence is invisible |
| Can you show every departure from the Roster of Care with a reason attached? | Roster drift is undetected |
| Is the shift type (sleepover or active night) recorded per participant with a rationale? | Cost and claim mismatch risk |
| Can you split one group shift across participants by funded arrangement, not headcount? | Systematic over- and under-claiming |
| Does verified delivery, not the roster, drive the claim batch? | Claims are built on intention |
| Can you see claim lines that will fail validation before you lodge? | Rejections are discovered late |
| Can you list every verified shift not yet claimed, with its age in days? | The 90-day window canât be managed |
| Is there location and time evidence for every completed shift? | Service verification is anecdotal |
| Do medication and PRN records sit with the participant, not in a separate folder? | eMAR evidence is fragmented |
| Does every incident have an owner, an action and a closure date? | Continuous improvement canât be shown |
| Do you track void days per room per month? | Vacancy cost is unknown |
| Could you hand one participantâs full evidence chain to an auditor today? | The audit is a project, not a state |
In a well-run SIL operation, the four systems that usually live apart behave like one system with four views.
That is the shape Ausvanta is built around. The SIL scheduler shows households, group sessions and assets on one timeline. Workers clock in with GPS and a selfie. In claims and billing, claim lines are validated against their source evidence, failed lines are fixed and re-validated, and shifts that canât be claimed yet are listed as blocked. Audit packs export from the compliance view. Ausvanta prepares and exports claim batches for you to lodge through your usual NDIA process; it doesnât lodge claims directly. If your current setup needs four tools and a monthly spreadsheet to answer the thirteen questions above, itâs worth seeing one system against your own house.
See it against your own house: book a walkthrough and weâll go through your roster shape, your ratios and your claiming workflow with you, or start a 14-day free trial with unlimited users.
From 1 July 2026, SIL supports are claimed under support items in registration group 0138, Assistance with Supported Independent Living. SIL delivered before that date stays under the earlier 0115 group. Registered providers, newly registered providers and those who have applied to register all use 0138 for supports delivered from 1 July.
Allocate the verified worker hours by each participantâs funded arrangement, not by headcount. Use verified start and finish times and location evidence as the source, flag absences or second-worker shifts before lodgement, and keep every claim line tied to its shift, worker, location evidence and note. Reconcile exceptions weekly.
No. The NDIA does not consider vacancy costs when making SIL funding decisions, and providers must not raise prices or claim extra funds from other participantsâ plans to cover vacancy periods. Vacancy is a business cost to manage, which is why tracking void days per room per month is worth doing.
In a sleepover the worker is present and available overnight but sleeping. In an active night the worker is awake and actively supporting, at a higher wage cost. The choice should be a documented decision at plan review, and the claim and progress notes must match the support actually delivered.
From 1 December 2026, a claim must be made within 90 days of the support being delivered, replacing the previous two-year window. Late claims may be rejected. For SIL, that means reconciling verified shifts and lodging claim batches weekly rather than waiting until month end.
Yes. Since 1 July 2026, providers delivering SIL must be registered, with certification audits against the SIL Practice Standards. Unregistered providers who delivered SIL before 1 July had until 1 October 2026 to apply. Those who did not apply can only claim for SIL delivered up to 30 September 2026.
The Roster of Care sets out which worker supports which participants, at what ratio and when, and it informs SIL funding. Drift between the roster you run and the roster you quoted creates unrecovered cost and weak audit evidence, so every departure needs a reason recorded against the shift when it happens.
Auditors follow a chain from the service agreement and plan, through the roster, shift verification, progress note, medication and incident records, to the claim line. Every link must hold and agree with the others. Most findings come from two links that never quite matched, such as a roster edited with no change history.
Reviewed November 2026. Operational guidance only. Confirm current NDIS Commission requirements and the NDIS Pricing Schedule against the official sources before relying on them.
Start your free 14-day trial â unlimited users, no credit card required.