Australia's NDIS market has matured. The off-the-shelf platforms that worked in 2019 are starting to creak — and providers are asking whether it's finally time to build their own. This is the honest framework we use with NDIS clients evaluating that decision in 2026.
We've built and rescued NDIS platforms for providers ranging from 20-staff regional teams to multi-state networks. CareUpp is our flagship case study, and the lessons below come straight from that engagement and four others.
Buy first, almost always.
If you're under 50 active participants and not running multi-program (NDIS + Aged Care + DVA), buy. Lumary, ShiftCare, Brevity and Carelink+ will all do the job for $80–$220 per user per month. The first sign you've outgrown them isn't usually a feature gap — it's the implementation cost of every new compliance tweak.
The four real reasons to build
- Multi-program billing complexity. When you're running NDIS, Home Care Packages and CHSP under one roof, off-the-shelf splits the work across three logins. A custom build collapses that into one workflow.
- Worker app UX matters more than you think. Generic mobile apps cause real cancellations. Workers stop logging shifts properly; payroll breaks; rosters slip. A bespoke app fixes that fast.
- Audit-grade reporting. NDIS Commission and Aged Care Quality Commission both want defensible audit trails. Off-the-shelf gives you reports; bespoke gives you the underlying event log.
- Integration with your accounting / payroll stack. Xero, MYOB, KeyPay and Employment Hero all have quirks. Custom platforms integrate cleanly; rented platforms charge per integration.
"We saved $14,200 a year on per-user fees within the first 12 months of going custom — but the bigger win was that our shift-claim error rate dropped from 9% to under 1%."
NDIS provider · 2025 build
What an honest custom NDIS MVP costs in 2026
- Phase 1 (8 weeks, $18k–$28k AUD): participant + worker management, plan budgets, shift claims, basic invoicing.
- Phase 2 (4–6 weeks, $9k–$15k): worker mobile app (offline-capable), shift logging with photo, expense receipts.
- Phase 3 (4 weeks, $6k–$10k): Xero/MYOB sync, NDIS price guide auto-update, BI dashboards.
Hosted on AWS Sydney for ~$280/month all-in. No per-user fees, ever. Compare that to $200/user × 30 users × 12 months = $72k/year and the build pays back inside 14 months.
Three honest reasons NOT to build
- You don't have a delivery team. Custom software needs ongoing maintenance — at minimum 1 day/month of dev time. If you don't budget for it, the system rots.
- Your processes change every quarter. If you're still figuring out how you operate, codify the wrong workflow and you'll pay to undo it.
- You want the certifications without the work. Building doesn't auto-give you ISO 27001, IRAP or SOC 2. If your enterprise clients require those, factor in another $40k+ of audit work.
The middle path: hybrid
The smart 2026 move for many providers is hybrid: keep your billing/invoicing on Lumary or Brevity, but build a custom worker app + roster engine that talks to it via API. We've done this for two clients and it's the lowest-risk way to get most of the value without rebuilding the whole stack.