Why business-as-usual work usually is not eligible R&D

If you run a business in Australia and are exploring the R&D Tax Incentive, one of the first things to understand is that not all technical work counts. In fact, the majority of day-to-day operational work is explicitly excluded from eligibility. This guide walks you through the principles that separate eligible R&D from business-as-usual, so you can honestly self-assess before approaching a registered tax agent. It is general information only, not tax or legal advice. Always confirm your position with a qualified professional.

The Australian R&D Tax Incentive is jointly administered by AusIndustry and the ATO, and the rules are designed to reward genuine experimentation, not everyday business activities. Many businesses, especially first-time claimants, come to GrantsMAX believing that any work involving technology or improvement must qualify. However, the program draws a firm line between routine development and true R&D. Understanding that line can save you from a rejected claim or, worse, an ATO review down the track.

What you need before you start

Before you begin self-assessing, make sure you have these building blocks in place:

• A working knowledge of the R&D Tax Incentive basics. Our plain-English guide to the R&D Tax Incentive explains the structure, rates, and registration requirements in clear terms.
• Access to your business's financial data, typically through Xero, MYOB, or QuickBooks. GrantsMAX's AI Application Pack Drafting can pull costs directly from your accounting software, but you still need to know which activities generated those costs.
• A clear list of the technical projects your team has undertaken. If you have not documented them, start with a rough timeline. Later, you will need contemporaneous evidence.
• An understanding that only a registered tax agent (or a tax agent with an R&D endorsement) can review and lodge the claim. GrantsMAX's Accountant Review & Lodge Workflow is designed with that division of responsibility in mind: the AI prepares an evidence-backed pack, and your accountant takes it from there.
• Awareness that the R&D Tax Incentive is not a rebate for all innovation spend. It is a targeted tax offset for activities that meet strict legislative criteria.

With these in hand, you can move through the steps below.

Step 1: Identify your core and supporting R&D activities

The first step is to categorise the work your business does. Under the program, activities are split into two types: core R&D activities and supporting R&D activities. Core activities are experimental activities whose outcome cannot be known or determined in advance, and that are conducted for the purpose of generating new knowledge. Supporting activities are those that have a direct, close, and relatively limited relationship to the core activities.

As an example, if you are developing a new machine-learning model to predict crop yields, the design and training of that model may be a core R&D activity. Contrast this with building a standard mobile app for your customers to order farm produce: that is likely business-as-usual, because a competent software developer could do it using well-known methods. The distinction matters because AusIndustry requires you to register your R&D activities annually, and they will look for a genuine hypothesis or technical gap.

To help you begin this categorisation, you might start with a list of all your current projects. Then ask: which of these involve attempting something that a professional in the field would not find straightforward? If the answer is "most of them", you probably need to be more selective. Our Eligibility Assessment & Risk Flags feature can flag areas that might meet the bar and those that need more evidence, based on your data.

Step 2: Determine if the work addresses a technical uncertainty

R&D, by its nature, must confront a specific technical uncertainty. This means you cannot know at the outset whether the solution will work, or how to achieve it within the constraints. The uncertainty cannot be about cost or scheduling, it must be about the underlying science or technology. According to the ATO, an activity is not experimental if you are simply applying established knowledge. The OECD's Frascati Manual 2015, which underpins many countries' definitions, describes experimental development as work that draws on research and practical experience to produce new or improved materials, devices, products, or processes. However, it also explicitly states that routine adaptations do not qualify.

In practice, you need to ask: "Did we face a genuine tech hurdle?" For instance, a manufacturing business might have tried to reduce material waste by trialling several novel process recipes whose outcomes were uncertain. That could be eligible. But if you simply purchased and installed a more efficient machine from an off-the-shelf vendor, that is business-as-usual. Even if the machine required fine-tuning, if the adjustments were standard for an engineer in that field, the work does not meet the threshold. GrantsMAX for manufacturers helps manufacturers identify which process improvement projects may cross into eligible territory.

Step 3: Check if the work follows a systematic experimental process

Genuine R&D is not ad-hoc troubleshooting or a random series of tests. It must follow a systematic progression of work from hypothesis to experiment, evaluation, and conclusion. You should be able to point to a plan, even if it is a simple document, and show that you iterated based on results. Without a systematic approach, the ATO may view the work as business-as-usual problem-solving rather than R&D.

This is a common stumbling block for software companies that constantly tweak their platforms. A typical situation is a tech firm that runs A/B tests on user-interface elements to improve conversion. Such testing, while data-driven, is usually not R&D because the outcomes are not fundamentally uncertain in a scientific sense; they are about commercial optimisation. On the other hand, if your team is developing a new compression algorithm that has no known solution and they go through a documented cycle of hypothesis, testing, and refinement, that may qualify. GrantsMAX for technology companies shows how to frame these narratives for a claim.

The importance of systematic experimentation is also reflected overseas. The IRS, for example, requires a similar level of rigour in its Form 6765, Credit for Increasing Research Activities instructions. While the Australian rules are separate, the global expectation is that eligible R&D leaves a paper trail of hypotheses and results.

Step 4: Assess whether the outcome was unknown at the start

A hallmark of R&D is that you did not know the result before you began. If your team knew it would work, and especially if they had done it before, the activity fails the test. This rule filters out many "business-as-usual" projects. For example, a construction company that regularly builds bridges knows that it can build a certain type of bridge using its standard methods. Even if a bridge is longer than usual, if the engineering principles are well-established, the work is routine. Only if the company attempted a design that pushed the boundaries of known structural engineering would the R&D criteria come into play.

Pro tip: Write down the scientific or technical challenge you are trying to solve before the work begins. If you cannot articulate a clear unknown, it is probably business-as-usual. This practice also creates contemporaneous evidence, which is invaluable for substantiation.

On the other side, many businesses mislabel well-understood digital transformation projects as R&D. Moving data to the cloud, implementing a standard CRM system, or automating invoice processing with off-the-shelf tools is not R&D. You may derive business benefits, but the technical challenge is minimal. The ATO regularly reminds claimants that "mere commercialisation or style, presentation, or minor modifications" are not eligible (see their guidance on business.gov.au). Always ground your assessment in official sources, and if you are unsure, discuss with a registered tax agent before lodgement.

Step 5: Compare with business-as-usual work: is it routine or incremental?

Now the crux of this guide: the direct comparison with business-as-usual work. Business-as-usual is any activity that you would perform in the normal course of running your operations, using standard practices. It may be technically skilled work, but if a competent professional in that field could achieve the result without experimentation, it is not eligible R&D.

Take a food manufacturer developing a new flavour variant of a snack. If the process involves tweaking ingredient percentages and the formulation approach is known (e.g., adjusting sugar and salt levels within a known range), that is routine product development. However, if the company had to invent a new encapsulation technique to preserve a probiotic strain under high-heat processing, and there was no guarantee the technique would work, that could be core R&D. The difference is the leap into the unknown.

Business-as-usual also includes activities like routine testing, data collection, customer feedback analysis, and market research. These are not experimental. The UK government's Corporation Tax: Research and Development (R&D) Relief spells out similar exclusions: routine testing or analysis, and simple modifications to existing products, do not qualify. The British Chambers of Commerce also warns businesses against over-claiming for marketing, design, or normal operational improvements.

Closer to home, AusIndustry's registration form asks you to describe the hypothesis and the experimental activities. If your answer reads like a standard project plan, design, build, test, you are likely looking at business-as-usual. To strengthen your self-assessment, you might read real-world innovation case studies, such as those highlighted by Harvard Business Review, which often outline the messy, uncertain process behind genuine R&D breakthroughs. Moreover, data from the National Science Foundation's The State of U.S. Science and Engineering 2024 shows that the bulk of reported R&D activity is concentrated in a few sectors, not spread evenly across every business function.

Another litmus test: would you have done this work even if there were no R&D Tax Incentive? If the answer is yes, and the work is essential to your business's ongoing commercial operations, it is almost certainly business-as-usual. The incentive is designed to reward additional risk-taking, not to subsidise your standard operating model.

Step 6: Evaluate the "dominant purpose" rule for supporting activities

Supporting activities are often misunderstood. For example, keeping the lights on in a lab is not automatically eligible. A supporting activity must have a dominant purpose of helping a core R&D activity. The "dominant purpose" test looks at whether the main reason you undertook the activity was to progress the R&D. If the activity would have occurred anyway, say, because the business needs a functioning production line, it fails.

Consider an agtech firm that builds a prototype drone for crop monitoring. The direct design and testing of the drone's novel sensor array is core R&D. Calibrating the sensor in a field trial might be a supporting activity if directly linked to resolving a technical uncertainty. But sending a technician to replace batteries is likely business-as-usual maintenance, even if the data from the drone feeds into R&D. The line is fine, and the ATO expects clear documentation linking each supporting activity to the core R&D.

Important: Do not assume that all lab or workshop time automatically qualifies. You must show the direct relationship. Your registered tax agent will be reviewing this linkage, and AusIndustry may also question it. Our Audit-Ready Evidence Trail is built to tie every line item in the application pack to its source, giving your accountant a strong basis to confirm the dominant purpose.

Step 7: Document evidence and substantiation

Business-as-usual work rarely produces the kind of technical documentation that the ATO might ask for in an R&D review. Eligible R&D, on the other hand, should generate contemporaneous records of the hypothesis, the experiments, the failures, and the conclusions. Without this contemporaneous evidence, even a genuine R&D claim can be denied. The ATO's view is that if you did R&D, you would have the records.

That means you need more than just a high-level description of what you did. You should have dated notes, emails discussing the technical challenge, design diagrams, test data, and time-tracking records for staff who dedicated their hours to R&D. This is where many businesses fall short: they have the activity but not the paper trail. GrantsMAX was built to help with this gap. By connecting to your Xero data and scanning your business emails and documents (with read-only access), GrantsMAX drafts a complete application pack that includes a cost structure linked to specific R&D activities.

The difference between business-as-usual records and R&D records is striking. A typical operational invoice for materials might just say "500kg aluminium, job #1234." For R&D, you would need to show why that aluminium was used for a particular prototype trial and how the outcome of that trial fed into the next iteration. Our Audit-Ready Evidence Trail helps index these pieces so your accountant can see the full picture before lodgement.

For comparison, look at the detail required by the IRS in the United States on Form 6765. The U.S. credit expects a breakdown of research expenditures by project and a description of the information the research was intended to discover. While Australian rules differ, the principle is the same: if you cannot explain what you were trying to achieve and how you tried to achieve it, the ATO will likely treat the costs as ordinary business deductions, not R&D offsets.

Step 8: Consider the role of a registered tax agent

Finally, remember that only a registered tax agent, not the business owner acting alone, can prepare and lodge the R&D claim on your business's behalf with the ATO. The tax agent reviews the eligibility assessment, the narratives, and the cost calculations, and must be satisfied that the claim is supported. This professional gate-keeping helps protect the integrity of the program.

When you use GrantsMAX, the platform prepares the evidence-backed pack, but it does not lodge. That step is deliberately left to your accountant. We have designed the Accountant Review & Lodge Workflow to give the accountant full visibility and control, and the business remains the owner of the claim. This division of responsibility means you get a thorough pack at a fraction of the cost of traditional consultants, but you still have the professional assurance of a registered tax agent's review. For first-time claimants, GrantsMAX for first-time claimants provides extra guidance to smooth the path, while the Annual Refresh & Accountant Channel keeps the process efficient in subsequent years.

Common pitfalls and pro tips

Pro tip: Build a one-page summary of each project before you approach your accountant. Include the hypothesis, the systematic experimental process, the uncertainties, and the key results. This summary will quickly separate candidate R&D from business-as-usual.

Important: Do not be tempted to "dress up" routine work as R&D by using technical jargon. The ATO and AusIndustry are sophisticated reviewers and will look past the language to the substance. If it walks and talks like business-as-usual, they will see it.

Pro tip: Use the "test of a competent professional." Ask: "Could a professional with ordinary skills in this field have done this without experimentation?" If yes, it is not R&D. This test is a practical shortcut endorsed in many industry guides.

Key takeaways

1. Business-as-usual work is any activity that can be performed using existing knowledge and routine practices. It is not experimental, does not involve a genuine technical uncertainty, and is expected to succeed.
2. Core R&D activities must involve a systematic, experimental approach where the outcome cannot be known in advance.
3. Supporting activities are only eligible if they directly contribute to a core R&D activity and their dominant purpose is to support that R&D.
4. Document everything. Contemporaneous records are not just nice-to-haves; they are essential for a defensible claim.
5. A registered tax agent must review and lodge the claim. They are your safeguard, and their comfort with the evidence is key.
6. The rules are not static, keep an eye on proposed changes, such as the planned increase in the refundable offset turnover threshold to $50 million, which may affect growing companies, though it is not yet law.

What to do next

If you are still uncertain whether your activities fall on the right side of the line, the best step is to get an expert review. At GrantsMAX, we connect to your accounting data to give you a data-driven eligibility snapshot, then prepare the pack your accountant needs to review and lodge. It is a faster, more affordable way to assess your position without sitting through weeks of manual consulting. To learn more, visit our founders and CFOs page or see how we compare to traditional R&D consultants. And if you are ready to take the next step, join the GrantsMAX waitlist, we are onboarding new users now.