Back to blog

Are Software Projects Missing Out On R&D Tax Credits?

R&D tax credits are one of Ireland’s generous and widely used tax incentives, supporting innovation in Ireland and boosting the economy.

The IT sector is the second largest industry claiming R&D tax credits by value but has the largest       number of claimants, according to Revenue’s statistics published in 2024.

So, are some developers are missing out on claiming their full entitlement?

The basic requirements for a qualifying project are:

• Overcoming scientific or technological uncertainty
• Seeking to achieve a scientific or technological advance
• Conducting systematic, investigative or experimental activities

These requirements are intentionally vague, to cater to projects across many fields. However, Revenue also released specific guidance for software projects.

Developers need to keep abreast of the state of the art. In such a fast-moving sector, like software, it’s important to identify when you believe you are pushing boundaries and make notes of your research into the wider field at the time.

To qualify as R&D, work must be systematic (organised and structured) and aim to make real advances. Revenue has clarified that Agile approaches are still systematic, even though they don’t follow a linear development path.

Claimants need to demonstrate which phases of the software development lifecycle meet the qualifying criteria. The boundaries of R&D begin and end with technological uncertainty; once the uncertainty is identified, R&D begins, and ends with the resolution of the uncertainty.

In large projects with formal project management, it’s pretty straightforward to track qualifying R&D expenses by phase. Smaller projects, or projects working iteratively, may have less formal tracking, so it’s important to clearly document any activity claimed as R&D while you’re doing it.

For agile projects, qualifying and non-qualifying activities may happen together, making it harder to separate them. In these cases, project managers should carefully estimate and allocate costs to reflect the effort spent on each type of activity.

It’s important to catalogue your qualifying activities as they happen, not only in the event of an audit, but also to help you accurately claim for staff and subcontractor costs.

Using or implementing existing tools, methods, or processes doesn’t usually qualify as R&D because it doesn't push the boundaries of what's known or tackle any real technological uncertainty. Routine analysis, copying, or minor updates to existing products aren’t considered R&D either.

It’s also important to consider the phases of R&D and the nature of the work you’re doing, as not all work in a project will qualify. Some projects will be largely qualifying, with small bubbles that are not, and others will be the opposite.

Examples of non-qualifying activities found in R&D projects include:

• Standard framework set up
• User acceptance testing to confirm that users are satisfied
• Packaging or preparing a product for market without any new technical hurdles
• Adding features that don’t require solving any new scientific or technical challenges

We’ve written up some example projects below to show the kinds of advances and uncertainties which might qualify.

Data Analysis Project

This example project centres around Company A’s need to recognise millions of data points input by hand by many people and process them into usable datasets for analysis.

Technological advance

Company A set out to create a data validation and orchestration framework to handle large volumes of interconnected data with different outputs. Their goal was to improve data analysis tools to make them 90% less dependent on manual input from operators.

Technological uncertainty

Company A wasn’t sure if it was even possible to automatically apply data logic across millions of datasets. No existing tools could handle automatic business logic for these unique data scenarios.

Metadata Connectivity Project

This example project centres around understanding data flows from external systems, which contain metadata. Company B builds metadata models for these systems. The metadata received from these systems is unintelligible and needs to be analysed and put into some kind of order to understand the data flow through these systems.

Technological advance

Company B wanted to upgrade its connector system to extract only the essential data needed for successful visualisations. They wanted to build an end-to-end visualisation tool for multiple databases by developing a multi-database connector. This connector could target a single instance while pulling in cross-database relationships across the whole system, giving users a clear, detailed view of data connections. This approach offers much more flexibility and detail than current solutions, helping users fully understand and make the most of their data environment.

Technological uncertainty

For SQL parsing, handling multiple databases introduced size issues. Pulling all metadata from various databases and storing it in memory wasn’t possible due to the sheer volume, with some systems having thousands of schemas and tables.

Additionally, some third-party products do not expose much connectivity information, which means that Company B would not know whether it is technologically feasible to integrate with them.

Company B could not determine if a connection was possible as the formats are unknown and undocumented.

As with any R&D project, software developers should keep continuous, contemporaneous records of R&D work done. This ensures that your project is qualifying at the outset, but also accurately calculated.

These records may include:

• Timesheets for staff
• Investigations into the state of the art
  • Competitor analyses
  • Results of a literature review
  • Results from industry research
• Evidence of the systematic investigation, such as:
  • Code repositories
  • Technical specifications
  • Experimental work and associated results
  • Project plans and logs
  • Milestone reports
• Staff qualifications and experience
• Internal communication discussing the projects

Revenue’s position is that all projects should have defined project deliverables, but they recognise that smaller-scale projects may have less formal governance methodologies in place.

The guidance makes a point that any electronic records should clearly indicate the author/creator, and the date of creation.

An R&D tax credit claim is made using the company’s Corporate Tax Return (CT1). However, companies should prepare their claim carefully, as they may need to defend it in the event of an audit.

This includes identifying eligible projects, only claiming for eligible expenditure and, ideally, writing up a full report that shows how your claim meets the eligibility criteria.

At Tax Cloud, we take the stress out of R&D tax claims. Our online portal allows to prepare robust claim reports with the assurance that they have been checked by R&D tax experts. Our team reviews your projects to make sure they’re qualifying and inspects your claimed costs for accuracy. We’ll even prepare the submission to Revenue with you.

Sign up to Tax Cloud now and take advantage of our no-win no-fee service.