Systematic progression of work in R&D claims

Systematic progression: the method a core R&D activity must follow

A systematic progression of work is the method a core R&D activity must follow: based on principles of established science, it proceeds from hypothesis to experiment, observation and evaluation, and leads to logical conclusions. Unstructured trial and error does not meet it, however much work is done, and each step needs records made at the time.

Reviewed by George WalchFacts last verified 25 October 2026

Example

For example, an illustrative battery firmware team hypothesises that a stepped charging profile keeps cell temperature under 45 degrees at a 2C charge rate. It varies the current ramp, holds ambient temperature at 25 degrees, logs cell temperatures, compares them with the limit and concludes the profile holds only up to 1.8C, which sets its next hypothesis.

Legislative reference: Section 355-25(1)(a)(i) and (ii), Income Tax Assessment Act 1997

The Act asks for two things about method: the work is based on principles of established science, and it proceeds from hypothesis to experiment, observation and evaluation to logical conclusions. Each step leaves a trace, and tribunals look for the traces.

The five steps and their records

Step	Record that shows it	When it should exist
Hypothesis	A note or plan stating the proposition and its basis	Before testing
Experiment	A test plan naming variables changed, fixed and measured	Before or during
Observation	Run logs, raw data, timestamped notes or tickets	During
Evaluation	Analysis comparing results with the hypothesis	After each run
Conclusion	A written outcome, including a negative one, and the next question	End of each cycle

The method applies the same way to code, chemistry and clinical work. What changes is the record: a pull request and benchmark output for software, a lab notebook for a formulation.

What fails it

Active Sports Management [2023] AATA 4078 is the clearest example. A company modified basketball shoe designs to suit one player, but nothing showed how hypotheses were formed, tested and recorded, and the Federal Court dismissed the appeal ([2024] FCA 1346). In Royal Wins [2020] AATA 4320, the absence of any record that a hypothesis was developed or tested was enough, however much work was done. AusIndustry's own guidance puts it in one line: "just exploring by trial and error isn't science".

Two patterns cause most failures:

unplanned trial and error, where changes are not designed or recorded in a way that lets a result be traced to a cause (a planned design that varies several factors at once can still be systematic);
selecting the best of known options by trial and error, such as scale-up settings, which AusIndustry says is not core R&D.

Where it shows up

Controlled tests can meet it when the thing measured is technical or clinical rather than commercial: see A/B tests and performance and scale engineering. The records come from your contemporaneous evidence.

Frequently asked questions

Does agile or sprint-based development count as a systematic progression?

Not by itself. AusIndustry's software guide says a development methodology neither makes an activity eligible nor rules it out. A sprint can contain an experiment, but the records still need a stated hypothesis, a planned test and an evaluation against it.

Do we need a positive conclusion?

No. A logical conclusion can be that the hypothesis failed, and that often leads to the next hypothesis. An experiment still running at year end is not a failure either: register the year's work and record the conclusion when it comes. What fails the test is work abandoned without its results ever being evaluated.

The R&D Tax Incentive is a self-assessment program. This page is general information, not tax, legal or financial advice; eligibility depends on your circumstances and you should seek independent advice for them.