A plain-English guide for Australian businesses building a custom app, web app or SaaS product. What actually qualifies, how much comes back, and where the catch is. No jargon, no overclaiming.
If you have an idea for a software product, the first question is rarely “is it any good.” It is “what is this going to cost me to build.” App development in Australia is not cheap. A lean minimum viable product commonly lands somewhere around $40,000 to $90,000, and a fuller custom platform can run well past $200,000 once you account for design, build, infrastructure and the inevitable rounds of change.
So here is the part a lot of founders never get told. If you are building genuinely new software through an Australian company, a meaningful slice of that spend may be claimable back through the Research and Development Tax Incentive. For smaller companies that can mean a refundable offset of up to 43.5% of your eligible costs. On a $60,000 build, that is a number worth understanding before you write a line of code.
This guide walks through what the incentive is, how much it is actually worth, whether software qualifies, and the honest catches that the slick “get money back” ads tend to skip.
The Research and Development Tax Incentive (often shortened to RDTI) is an Australian Government program. It is jointly run by the Department of Industry, Science and Resources and the Australian Taxation Office. The idea is simple: the government wants local businesses to take on risky, innovative work, so it offsets some of the cost of that work through the tax system.
It is not a grant you apply for up front and wait to be approved. It works through your company tax return. You do the eligible work, you register the activities, and you claim an offset based on what you spent. Because it runs on self-assessment, the responsibility to get the eligibility right sits with you.
The benefit depends on the size of your company.
To make that concrete, here is roughly what the offset can look like at different levels of eligible spend for a small company. Treat these as illustration only, because the real figure depends on your tax rate and profit position.
| Eligible R&D spend | Indicative offset |
|---|---|
| $40,000 | $17,400 |
| $60,000 | $26,100 |
| $100,000 | $43,500 |
| $250,000 | $108,750 |
Figures are illustrative. Your actual benefit depends on your company tax rate, profitability, and which of your costs are genuinely eligible. This is not a quote or a promise.
This is where most people get it wrong in both directions. Some assume software never counts because it is not done in a lab. Others assume every hour their developers bill is claimable. Neither is true.
The incentive separates your work into two buckets. Core R&D activities are experimental: the outcome cannot be known in advance, you are working from established science or engineering principles, and you resolve the unknown through a systematic progression of work, building a hypothesis, testing it, observing the result and drawing a conclusion. Supporting R&D activities are the work directly tied to those core activities.
For software, the line tends to fall like this.
Australian companies that have used the incentive for software include Moodle, intelliHR and DUG Technology, per business.gov.au. The common thread is real technical risk, not just the act of writing code. The plain test to keep in your head: were you solving a problem where you honestly did not know in advance whether your solution would work? If yes, you may be in scope. If you were assembling known parts in a known way, probably not.
A few hard requirements decide this before anything else.
Put the pieces together and the maths shifts. A $60,000 build that involves real technical experimentation is not really a $60,000 decision for an eligible company. If a good chunk of that work qualifies, the effective cost after the offset can look very different. That is the difference between an idea staying a spreadsheet and an idea becoming a product.
The catch is that the incentive rewards a specific kind of work, and it rewards you for being able to prove you did it. This is exactly where founders trip up. They build something genuinely novel, then have nothing on paper showing the uncertainty they faced, what they tried, what failed and why. A claim built on memory after the fact is a weak claim, and the ATO does review them.
We are not a registered tax agent or R&D tax consultant, and nothing here is tax, financial or legal advice. Eligibility is self-assessed and genuinely technical, and the ATO and AusIndustry can review or audit a claim. Some operators are aggressive about what they say qualifies. Be careful with that.
Before you rely on any of this, speak to a registered tax agent or an R&D tax specialist about your own situation. The figures above are illustrations, not a promise of what you will receive.
We build custom apps for service businesses. The kind of work where an off-the-shelf tool does not fit how you actually operate: a client adherence app for a physio practice, an accountability tool for a coaching business, a between-session workbook for a psychology clinic, an operations hub for an agency. Often these builds involve solving a real problem that nobody has solved quite this way before, which is the same territory the incentive is interested in.
Here is the part that matters for a future claim. Because of how we work, we document the build as we go. The technical questions we were unsure about, the approaches we tested, the things that did not work and what we changed. That record is not just good practice. It is close to the evidence a registered adviser needs to support an R&D claim, and it is the thing most founders wish they had kept.
So the simple version: we build the product and document the journey. Your registered R&D adviser handles the claim itself. We are happy to work alongside them and hand over what we have. People first, process second, technology third, and a paper trail that earns its keep.
Generally no. The incentive is only open to companies incorporated in Australia, meaning a Pty Ltd or Ltd. Sole traders, partnerships and trusts cannot claim it directly. If you are serious about building software, this is worth raising with your accountant early.
No. Routine coding, debugging, configuration and building to a known specification usually do not count as core R&D. The work needs to involve a genuine technical uncertainty that you resolve through systematic experimentation.
It varies widely. A lean MVP often falls around $40,000 to $90,000, and a fuller platform can run past $200,000 depending on features, integrations and complexity. The incentive can change the effective cost for eligible companies, which is why it is worth understanding before you commit.
Within 10 months of the end of your financial year. For a 30 June year end, the deadline is 30 April. It cannot be extended, so the work and the paperwork need to keep pace with each other.
No. We are based in Melbourne’s northern suburbs and we work with businesses across Australia over Zoom. The build process is the same wherever you are.
Let’s talk about what you want to build, whether a custom app is the right call, and how to set the work up so it stands on solid ground from day one.
Explore custom app builds