Australian startup grants and R&D incentives

Every round is priced against risk and runway. Non-dilutive funding attacks both. A company that recovers a meaningful share of its engineering spend through the R&D Tax Incentive can reach the same milestone on materially less equity, which compounds directly into founder ownership at exit.

Investors read grant activity as a secondary signal too. A successful competitive grant means an independent panel reviewed the technology, the commercialisation plan and the team, and backed it. That is not a substitute for revenue, but it de-risks the technical story — particularly in deep tech and applied AI where the engineering claim is hard for a generalist investor to verify.

The failure mode is the opposite: a company that has spent two years optimising for grants rather than customers. Investors notice. Government funding should accelerate a commercial plan, never substitute for one.

The R&D Tax Incentive (RDTI) is jointly administered by AusIndustry and the ATO and is the single largest source of innovation funding in Australia. For companies under the aggregated turnover threshold and in a tax-loss position, it delivers a refundable offset — a cash refund rather than a credit carried forward. Larger companies receive a non-refundable offset tiered to R&D intensity.

What actually qualifies. The test is not "did we build something new to us". A core R&D activity must have an outcome that cannot be known in advance by a competent professional, and must be resolved through a systematic progression of hypothesis, experiment, observation and conclusion. Supporting activities are claimable where they directly relate to that core work.

What usually does not. Routine feature development, UI and UX work, configuration of off-the-shelf systems, standard integrations, data migration, bug fixing, and internal business software built for your own administration. Many software claims fail on review not because the work was unimpressive but because it was documented as delivery rather than experimentation.

AI-specific note. Fine-tuning, retrieval architecture, evaluation harnesses and novel model-serving approaches can qualify where there is genuine technical uncertainty about whether a target performance threshold is achievable. Calling an existing model API and shipping a product around it generally does not.

Cash-flow reality. The refund arrives after the financial year ends and the claim is lodged and processed. In a model, treat it as a lagging receivable, not as monthly income. Specialist R&D lenders will advance against a forecast claim, which is useful for bridging but carries a real cost of capital — assess it as debt.

Registration. Activities are registered with AusIndustry after the end of the income year, within the statutory deadline, before the offset is claimed in the company tax return. Missing that deadline forfeits the year.

Accelerating Commercialisation (Industry Growth Program). Matched funding for taking a novel product from validated prototype to first commercial sales, delivered alongside adviser support. It suits companies with defensible technology and a concrete route to market, and is deliberately competitive. Because funding is matched and paid on milestone reimbursement, you need working capital to spend first and claim after.

MVP and early-stage state grants. The New South Wales MVP Ventures program and equivalent state schemes — Victoria, Queensland and South Australia each run their own — fund the step from proof of concept to a minimum viable product a customer will pay for. Amounts are smaller and turnaround is faster, which makes them a good fit ahead of a pre-seed or seed round. Eligibility is usually tied to where the company operates and employs, so check the current state guidelines rather than assuming last year's rules.

Export Market Development Grant (EMDG). Reimburses a portion of eligible costs of promoting an Australian product overseas. For a SaaS company making its first push into the United States or Southeast Asia, this partially offsets marketing, travel and market-entry spend that boards often resist funding.

CSIRO Kick-Start and research vouchers. Matched dollar-for-dollar funding for research services delivered with CSIRO, useful where a technical claim needs independent validation before an investor will underwrite it.

Early Stage Innovation Company (ESIC) status. Not a grant, but adjacent and often overlooked. Qualifying companies allow their investors to access a tax offset and a CGT exemption on eligible shares. It does not put money in the company, but it materially improves how an angel round prices and closes.

Programs, amounts and eligibility change between funding rounds. Confirm the current guidelines on business.gov.au and the relevant state agency before you build a plan around any of them, and take professional tax advice on the RDTI.

The programs are complementary if you order them correctly. A typical path for a technical software company looks like this.

Pre-product. Establish the R&D record-keeping discipline from the first sprint — timesheets against activities, written hypotheses, experiment logs. Retrofitting evidence a year later is where most claims weaken. Consider a state MVP grant to fund the build to a payable product.

Post-product, pre-revenue scale. Lodge the first RDTI claim and use the refund to fund the next engineering cycle. Assess ESIC status before your angel round so investors get the benefit.

Early commercial traction. Accelerating Commercialisation becomes viable once you can evidence validated demand. This is also the point where matched funding is affordable, because you have the working capital to spend ahead of reimbursement.

Expansion. EMDG offsets the cost of the first offshore market push, alongside a Series A that funds the go-to- market team.

In diligence, the grant question is rarely "how much did you get". It is closer to the following.

Is the runway calculation honest? Does the model treat the RDTI refund as a lagging receivable with a realistic processing time, or as certain monthly income? An overstated refund hidden inside a runway number is one of the fastest ways to lose credibility in a data room.

Is the claim defensible? Would the registered activities survive an AusIndustry review? Contingent liability from an aggressive historical claim is a diligence finding, and it can reduce a valuation or hold up a close.

Is the business grant-dependent? If removing all government funding kills the plan, that is a structural risk. If it slows the plan by two quarters, that is leverage.

Does the founder understand the instrument? A founder who can explain, unprompted, which activities are core versus supporting, when the refund lands, and how it changes the milestone plan is demonstrating exactly the operating discipline investors are trying to price.

Before your next raise, be able to answer each of these in a sentence:

— Which R&D activities are registered, and what is the technical uncertainty each one resolves?

— What is the forecast refund, and in which month does it land in the bank?

— What evidence exists, written at the time, for each activity?

— Which grants have you applied for, which succeeded, and what did each one change about the product or the market position?

— Does the company qualify for ESIC, and has that been documented?

— Stripped of every government dollar, what does the plan look like?