When software development becomes R&D


What is the difference between routine software development and that which qualifies as R&D?

This question is commonly posed by AusIndustry as a part of their compliance reviews, requesting further information.

Software development has long been a contentious area for R&D claims, and it is also the fastest growing area.

With the burgeoning start up ecosystem in Australia and internationally fuelled by seemingly endless online marketplaces and software as a service offerings, software development R&D claims are likely to continue to increase, as is the scrutiny which will be applied by regulators.

So, how do you know your software development activities qualify as R&D? And importantly, if you are already claiming, are you describing your activities appropriately to minimise the risk of audit activity.


Completion must be dependent on the development of a scientific or technical advance, and the aim of the project must be resolution of a scientific or technical uncertainty. 

You may be undertaking R&D if:

  • The context in which a solution is sought makes the solution unique or adds complexity,

  • there is no reference solution or knowledge that can be leveraged, and

  • experimentation (testing of a number of prototype designs) will be required to achieve the solution, or certain aspects of the solution.

To summarise, any software development work where considerable scientific or technological challenges are faced should be considered, provided the solution is not available to, or readily deducible by a competent professional working in the field.


  • Development of new algorithms, theorems, architectures, protocols, rules engines.

  • Advancements to existing system capabilities in areas of performance, security, scalability, availability.

  • Redesigning existing systems using new/different technologies, e.g. a legacy monolithic application to decouple functionality, new architecture design to optimise cloud storage.

  • Systems integration that involves large scale and complexity of components not previously integrated in a similar manner.

  • Developing unique middleware layers for translating messages between differing systems/layers.


In claims that we review, we often find that there is insufficient detail that would enable AusIndustry to ascertain the registered activities meet the legislative definition of R&D activities. Often companies will provide very brief narrative, or discuss commercial objectives and development outcomes.

Here are our top three tips for improving R&D activity descriptions:

  1. Discuss the state of the art and available knowledge, emphasising how the scientific or technical knowledge you are seeking differs.

  2. Outline research undertaken in an attempt to identify existing solutions that may be leveraged, and the scientific or technical unknowns which remain.

  3. Describe the experiments that have been undertaken from a procedural perspective (hypothesis, process of testing, outcomes including observations and evaluations, further design revisions).

Using a R&D Tax consultant that specialises in software R&D claims can significantly improve claim outcomes and efficiency. They should be able to understand the technical terms, work undertaken, and ask the right questions to probe into the strongest areas. Using an interview process, they should be able to identify and document the R&D activities undertaken.