Core R&D activities vs supporting R&D activities: how to tell them apart

Introduction

Any Australian business claiming the R&D Tax Incentive must correctly classify its activities as either core R&D activities or supporting R&D activities. The distinction is not academic: only costs attributable to an eligible core activity (or a directly linked supporting activity) can feed the notional R&D deduction and the resulting tax offset. Getting it wrong can mean an inflated claim, an ATO review, or expenditure left off the table. This guide walks through the official tests, gives practical examples, and highlights where a registered tax agent’s oversight becomes essential. Always remember: the information here is general only and does not constitute tax or legal advice. You should confirm your own position with a registered tax agent who knows your business.

GrantsMAX prepares an evidence-backed application pack from your accounting data, but the final review and lodgment rests with your registered tax agent, and you own the claim. For a plain‑English overview of the incentive itself, see What is the R&D Tax Incentive?.

Why getting the distinction right matters

Under the R&D Tax Incentive, registered companies may be entitled to a tax offset for eligible R&D expenditure. The offset calculation starts with the aggregated expenditure on core R&D activities and on supporting R&D activities that are directly and substantially related to those core activities. The ATO’s overview of the incentive emphasises that eligibility hinges on the activity meeting the legislative definition, not just on a general notion of “research.”

If you purposefully or inadvertently include costs that relate to non‑core or unrelated supporting activities, the claim may be overstated, exposing the company to compliance risk. Conversely, if you fail to identify a genuine supporting activity that is directly integral to core R&D, you miss out on legitimate expenditure. Many R&D‑active businesses in software, manufacturing, biotech, agtech, and clean energy discover that careful categorisation both strengthens the claim and simplifies the substantiation the ATO expects.

The Income Tax Assessment Act 1997 sets out the statutory tests, and AusIndustry (within the Department of Industry, Science and Resources) provides detailed guidance through the eligibility‑of‑activities fact sheet. The fact sheet is one of the most practical official resources available, and it directly addresses how to tell core from supporting. It is a document worth keeping open while you work through this guide.

Prerequisites: what you need before you begin

Before you attempt to categorise activities, make sure you have:

• A clear project or program description that records the overall R&D objective.
• Detailed records of the tasks, experiments, and iterative work undertaken during the period.
• Time sheets, work logs, or project management data that link specific tasks to people and costs.
• Access to your accounting system (Xero, MYOB, QuickBooks, Microsoft 365, Google Workspace) so you can later trace expenditure to activities.
• The involvement, or at least the availability, of a registered tax agent who will ultimately be responsible for the technical accuracy of the R&D claim lodged with the ATO.

GrantsMAX works for R&D‑active startups, scale‑ups, manufacturers, and technology companies, and indeed for the accountants and bookkeepers who serve them. While it connects to your live accounting data to surface possible R&D activities, it is not a substitute for your own understanding of the business’s technical work. The business, not software, owns the characterisation, and the accountant verifies it.

What are core R&D activities?

The Australian Government defines core R&D activities by three interrelated tests found in the tax legislation and explained in the business.gov.au eligibility fact sheet. A core activity must:

1. Be experimental, the work involves a systematic progression of work based on testing a hypothesis, not just applying known techniques.
2. Have an outcome that cannot be known or determined in advance on the basis of current knowledge, information, or experience.
3. Be conducted for the purpose of generating new knowledge (including new or improved materials, products, devices, processes, or services).

The statutory test for core R&D

The Income Tax Assessment Act 1997 describes core R&D activities as those that meet the above criteria and are conducted in a scientific or technological field. The activity must pursue a specific, technically uncertain outcome. Routine development, debugging of a known system, market research, management studies, or cosmetic changes ordinarily do not satisfy the experimental and new‑knowledge requirements.

What counts as experimental

“Experimental” does not mean every trial qualifies. The activity must involve testing a hypothesis, evaluating alternative approaches, or creating a prototype to resolve technical uncertainty. For a manufacturer experimenting with a new composite material, working through different resin mixtures and curing conditions to see what achieves the target strength, with no off‑the‑shelf answer, is likely a core activity. But switching to a cheaper, readily available raw material without any technical novelty is not.

Outcome cannot be known in advance

If a competent professional in the field could determine the outcome from publicly available information or from standard professional practice without undertaking a systematic investigative process, the activity is not a core R&D activity. This test weeds out activities that merely apply an established methodology to a known problem. The uncertainty must be technical, outcome unknown in a way that requires investigation, not unquantified commercial risk.

The generation of new knowledge

The purpose of the activity must be to create new knowledge. “New” here means new to the world, not just new to your business. Incremental improvements can qualify if the underlying technical approach is novel, but a simple configuration change that any skilled engineer would make does not meet the threshold.

These definitions align with the OECD Frascati Manual, which Australia uses as a primary reference for measuring R&D, and with international practice as reflected in the WIPO Intellectual Property Handbook. The Frascati manual clarifies that R&D must be “creative and systematic work undertaken in order to increase the stock of knowledge.”

What are supporting R&D activities?

Supporting R&D activities are not themselves core, but they are directly related to core R&D activities and must be undertaken for the dominant purpose of supporting those core activities. The legislation requires that a supporting activity be, in effect, integral to the core activity.

The business.gov.au eligibility fact sheet provides a clear explanation: an activity will be a supporting R&D activity only if it is directly related to core R&D activities and is necessary for them to be carried out. For example, cleaning a cleanroom, calibrating instruments used exclusively for experiments, or fabricating a test rig can be supporting. But ordering standard office supplies is not, because it lacks the direct, necessary link.

The integral‑to test

The phrase “directly related” is often clarified by asking: would this core activity be impossible or significantly impeded without it? If the task is indispensable and closely targeted at the core experiments, it is likely a supporting activity. The New Zealand R&D tax incentive guidance on supporting activities illustrates the same principle, the activity must be integral to the core R&D, not merely useful or coincidental.

Activities directly related to core R&D

In Australia, the law specifies a non‑exhaustive list of activities that may be supporting if they satisfy the direct‑relationship test, including:

• setting‑up and maintaining equipment used exclusively for core R&D
• data collection and analysis performed solely as part of a core experiment
• trials and prototyping, where the direct purpose is to test a core hypothesis
• documentation and patent searches that inform the technical direction of core R&D

Each of these must be attributable to specific, defined core activities and not to general business operations.

Examples of supporting activities

• A biotech startup performs a core activity of experimenting with gene‑editing reagents to improve yield. The daily calibration of the flow cytometer used only in those experiments is a supporting activity. Washing standard lab glassware that is used across many projects, commercial and R&D, would not be.
• A software company runs a core activity of testing novel algorithms for latency reduction. The process of creating the test data sets and simulating high‑load conditions is supporting, because without it the algorithm testing could not proceed. General unit testing of existing features, however, is not.
• A clean‑energy firm develops a new hydrogen electrolyser stack, a core activity. The fabrication of the experimental membrane‑electrode assemblies and the maintenance of the test rig are supporting, provided they are used exclusively for the core experiment.

Whenever you identify a potential supporting activity, document precisely which core activity it serves, why it is necessary, and when you would cease to incur it if the core activity stopped. That record becomes central to the audit‑ready evidence trail GrantsMAX builds for your accountant to assess.

How to tell them apart: a step‑by‑step guide

Now that the conceptual framework is in place, you can apply a structured process. The following steps walk you through the categorisation, and at each stage the involvement of a registered tax agent is critical, AI or a checklist alone cannot make the definitive call.

Step 1: Map out all activities

Start with a plain list of everything your team did during the relevant period that might relate to R&D. Use timesheets, Jira tickets, lab notebooks, or internal progress reports. Do not filter yet, just capture the tasks. GrantsMAX’s discovery engine pulls in financial transactions and project descriptions from your accounting data to help compile this initial map, but the human understanding of why a task was done is indispensable.

Step 2: Apply the core activity tests

For each item on the map, ask three questions, referencing the ATO’s guidance:

1. Is it experimental? Did you test a hypothesis? If the work simply repeated a known process, score it no.
2. Could the outcome be determined in advance? If a technical expert would have had no reason to doubt the result, the activity fails.
3. Was the purpose the generation of new knowledge? The activity must be aimed at creating knowledge that did not previously exist in a scientific or technical sense, not just solving a commercial problem.

If the answer to all three is clearly yes, the activity is a core R&D activity. Flag it as such and move to the next.

Step 3: Identify activity purpose, experimental or operational?

Many borderline activities involve testing or validation. Ask: is the testing designed to resolve a scientific or technological uncertainty, or is it routine quality assurance? For example, durability testing of a new alloy formulation to see whether it meets a previously unknown performance threshold is core. Durability testing of a batch of well‑known steel to satisfy a customer specification is not. Draw a bright line: if the uncertainty disappears and the work is simply verifying a predetermined outcome, it is not core.

Step 4: Check for a substantial direct link

If an activity did not pass the core tests, examine whether it might be supporting. Look for a substantial, direct, and inseparable connection to a specific core activity. Could the core activity be carried out without it? If no, it is likely supporting. Document the linking rationale: “Task X was performed solely to enable core activity Y, consuming Z hours.” This linkage is where GrantsMAX’s eligibility assessment and risk‑flagging feature highlights activities that may warrant a closer look, but again, the registered tax agent must confirm the classification.

Step 5: Document your rationale

For every core and supporting activity, write a short narrative that explains:

• The hypothesis or technical aim (for core)
• The exact relationship to a core activity (for supporting)
• The key personnel and time spent
• The contemporaneous records that back it up

Later, when GrantsMAX drafts the application pack, it assembles these narratives into a structured, evidence‑backed guide. But the pack is only as good as the underlying logic the business provides.

ⓘ Pro tip: Build the rationale in real time. Trying to reconstruct why you undertook a piece of work twelve months later is difficult and can weaken your substantiation if the ATO asks. Keep a simple log, even a shared spreadsheet, that maps tasks to the R&D purpose.

Step 6: Have your accountant review before lodgment

The registered tax agent’s professional judgement is the final safeguard. They will test your categorisation against the legislation and ATO interpretations and may reclassify activities or request additional evidence. This review step is integral to the Accountant Review & Lodge Workflow that GrantsMAX supports: the system prepares the pack, but the accountant controls what gets lodged with the ATO. No claim should go forward without that sign‑off.

Common pitfalls to avoid

• Calling everything “experimental” to inflate the claim. This is the fastest way to attract ATO scrutiny. Activity descriptions must be honest and technically defensible. Traditional R&D consultants often rely on a back‑and‑forth discovery process, but the tests remain the same.
• Treating all prototype builds as core. Fabrication of a prototype is core only if the experimentation occurs during the build itself. Building a copy of an already‑tested design for additional trials is often supporting, not core. The distinction is subtle but critical.
• Assuming that government grant‑funded activities are automatically R&D. A grant may fund a project that includes non‑experimental work; each activity must still satisfy the statutory tests.
• Failing to exclude ineligible costs. Even if an activity is core or supporting, ordinary overheads, interest expenses, and capital expenditure are generally not eligible. Make sure the costs feeding the notional deduction are specifically attributable.
• Using the wrong activity period. The expenditure must relate to the same income year in which the activity occurred, and the registration must be lodged with AusIndustry within ten months of year‑end.

Warning: This article cannot detect whether your specific activity qualifies. The tests require factual interpretation of your technical work, and the ATO regularly updates its compliance focus. Always involve a registered tax agent.

Pro tips for accurate categorisation

• Start with a single, well‑defined project. Trying to categorise all activities across the entire business at once can lead to loose judgements. Pick one R&D project, map its tasks, classify them, and then replicate the discipline for other projects.
• Use the business.gov.au fact sheet as a checklist. The eligibility‑of‑activities fact sheet includes a flowchart and a table of examples that can ground your thinking. Keep it open while you work.
• Don’t overthink “new knowledge.” Innovations do not need to be world‑first breakthroughs. If you are tackling a genuine technical barrier that no published solution resolves, you are in the right space.
• Remember annual repeatability. The R&D Tax Incentive repeats each financial year. Systems that catalog your activity‑to‑cost link once make the next year’s claim far easier and more consistent. Many R&D‑active businesses record activities in a way that maps to the core/supporting split from the start, using a tool like GrantsMAX’s accountant‑friendly platform to keep the pipeline clean.
• When in doubt, seek a second opinion. A quick call with your tax agent before you commit time to narrative drafting can save hours of rework.

Where GrantsMAX fits in

GrantsMAX does not lodge your claim, and it does not give advice. It connects to read‑only accounting data from Xero, MYOB, QuickBooks, Microsoft 365, and Google Workspace, and it uses that data, alongside your activity descriptions, to surface possible R&D activities and supporting tasks you might otherwise overlook. For R&D‑active startups, manufacturers, and technology companies, this initial matching step is often the hardest part, because it requires labour‑intensive back‑tracing through invoices and timesheets.

Once you approve a set of candidate activities, GrantsMAX drafts a complete evidence‑backed application pack. The pack includes an R&D activity narrative, a cost structure pulled from your accounting system, and an index tying each activity line to its supporting evidence (emails, test reports, supplier invoices). The pack is then handed to your registered tax agent inside a shared workspace. The agent reviews, refines, and ultimately lodges the claim, and you remain the owner of the claim throughout.

This division of roles, AI‑assisted preparation, professional review, business ownership, is designed to help you get the activity classification right without ever substituting for the independent judgement that the Tax Practitioners Board expects of registered agents.

Key takeaways

• Core R&D activities involve experimentation, outcome uncertainty, and new knowledge generation. They are the engine of an R&D claim.
• Supporting activities are those directly and substantially related to core R&D, without which the core work would be impracticable.
• Accurate categorisation is not a box‑ticking exercise: it determines what expenditure is eligible, and it shapes the substantiation the ATO may request.
• Always document the rationale and evidence contemporaneously; a retrospective effort is much harder and less reliable.
• A registered tax agent’s review is non‑negotiable, the tax system places the responsibility on them, not on software or on the business filing in isolation.

If you are preparing for your first R&D claim or looking to refine the categorisation in an existing program, GrantsMAX’s discovery and preparation tools can surface the right activities and build the evidence pack, leaving your accountant with a clean, well‑organised file to review. The platform is currently in controlled release; you can join the waitlist at GrantsMAX to be among the first Australian businesses to use it.