Ireland's R&D tax credit lets you recover up to 35% of qualifying expenditure. It's one of the reasons Ireland has become one of the world's most attractive fintech markets: the country now hosts over 400 global financial firms, and Enterprise Ireland's fintech clients alone generated over €600 million in R&D between 2018 and 2023.
But volume of spend isn't the same as qualifying spend. This post explains what counts as R&D in a fintech context, what costs you can include, and where companies most often get it wrong.
Why Irish fintech is R&D-intensive by nature
The sectors where Irish fintech has built real strength (regtech, payments, embedded finance, ESG, and AI-driven financial tools) all involve genuine technological problems; that's meaningful for Revenue. The R&D tax credit explicitly centres on scientific or technological advancement, and most serious fintech work sits comfortably within that frame.
It's worth noting that Ireland's R&D tax credit is regularly cited alongside the country's pro-business environment as a key reason global firms choose to set up here. Revenue knows fintech companies are active claimants.
What counts as R&D in fintech?
Revenue's definition of qualifying R&D centres on three things: scientific or technological uncertainty, systematic investigation, and advancement in the field.
The uncertainty test is the one most fintechs trip over. Technical complexity isn't the same as technological uncertainty. Building something difficult doesn't automatically make it R&D. Revenue's Manual states that qualifying activities must involve "the resolution of scientific or technological uncertainty." This essentially means that if a competent professional in your field would know, at the outset, how to achieve the technical goal you're working towards, you're not in qualifying territory. If the answer to that question is genuinely unclear, you probably are.
However, alongside this uncertainty, your project must also have an advance you’re looking for. This needs to go beyond the state of the art and the capabilities of the general field, not just in your company. The field will be specific to your project and what it is attempting to do.
Qualifying activity typically includes:
- Novel fraud detection or AML systems that go beyond existing approaches and require you to solve problems without established solutions
- AI or machine learning models for risk management or back-office automation where no off-the-shelf approach exists and the model architecture itself involves genuine uncertainty
- Regtech compliance tools built to navigate new regulatory frameworks like MiCA or DORA, where the technical standards are still being established and the engineering solutions aren't settled
- Embedded finance infrastructure that requires solving new data integration or architecture problems that can't be resolved using existing methods
- ESG data modelling where there's no accepted industry methodology and you're developing new approaches from the ground up
What typically doesn't qualify:
- Replicating a competitor's feature using known methods
- Rebuilding an existing system on a new technology stack without any genuine technical uncertainty
- Integrating third-party APIs, even complex ones, where the integration path is documented and known
What costs can you include?
Once you've identified qualifying activity, you need to link your costs to it. Eligible cost categories include:
- Staff costs for employees directly involved in R&D activities
- Subcontractor costs (subject to a cap; more information in our guide to claiming R&D tax credits for subcontracted work)
- Consumables used directly in the R&D process
- Software licences used directly in qualifying R&D work
- Plant and machinery used in R&D (our guide to claiming R&D tax credits for buildings, plant and machinery covers this in full)
The documentation linking each cost to a qualifying project is everything. Claims with broad cost categories need to be able to demonstrate that each cost was actually incurred and that they were properly apportioned to only include the R&D portion.
Do you need to be a dedicated R&D company to claim?
No. Most Irish fintech companies run R&D alongside live commercial products: payments platforms, lending tools, compliance software. That's absolutely fine. What you need to do is identify and separate the qualifying R&D component from the routine commercial activity sitting alongside it.
Consider a payments company operating an established B2B platform while simultaneously developing a novel cross-border settlement engine using a new cryptographic approach. The platform maintenance doesn't qualify. The settlement engine work, where there's genuine uncertainty about whether the technical approach will work, almost certainly does. The key is being able to draw that line clearly in your records and defend it.
What if you're building on open-source or third-party platforms?
Builds on existing platforms can qualify, but only for the parts that involve genuine technical uncertainty. This matters most for fintech companies working with AI, which is consistently flagged as one of the sector's biggest growth areas. Adapting an existing machine learning model to a new dataset isn't R&D, but training a novel model on proprietary financial data to solve a problem with no established solution might be.
This is one of the most common sticking points in Revenue audits of fintech claims. You need to show where the technological uncertainty lay, not just what you built. If your technical report describes outputs rather than the investigative process and the uncertainties you were trying to resolve, you could be at risk of an audit.
Can you claim if your R&D didn't succeed?
Yes, and this is an important point. Revenue is interested in the process, not the outcome. A fintech that spent 18 months building an embedded finance product, hit real technical barriers, investigated solutions systematically, and ultimately shelved the project because the business model didn't stack up has still done qualifying R&D. Revenue is assessing whether you carried out the work in a systematic way and whether real uncertainty was involved.
Failed projects can and do qualify. They can also clearly demonstrate technological uncertainty, if you didn’t pursue your project because of a technological roadblock. Don't leave them out of the claim simply because they didn't land.
What Revenue expects when you claim
Revenue states that qualifying R&D activities must seek to achieve scientific or technological advancement through the resolution of scientific or technological uncertainty.
What that means in practice: your supporting documentation needs to do more than list what you built. It should explain what you were trying to achieve, what the technological uncertainty was, what approaches you investigated, and what you found. That's the standard a well-prepared technical report should meet, and it's also what Revenue will look for if they raise an audit on your claim.
With 25 fintech VC deals in Ireland in 2024 and the sector projected to grow by 26% by 2028, Revenue's scrutiny of fintech R&D claims has increased. Understanding the evidence standards Revenue applies before you file is a much better use of your time than dealing with queries afterwards.
One more thing worth flagging if you're claiming for the first time: since 2024, first-time claimants are required to send Revenue a pre-filing notification before submitting a claim. It's a step you can't miss and one that catches more companies than you'd expect.
Key takeaways
- Technological uncertainty is the central test. Complexity alone doesn't qualify; you need to show that a competent professional in your field couldn't have known, at the outset, how to solve the problem.
- Fintech's strongest qualifying areas include novel AI models, regtech built for emerging frameworks, and embedded finance infrastructure where the technical solutions aren't established.
- Costs must be directly linked to qualifying projects. Broad cost categories without a clear connection to specific R&D activity are what Revenue scrutinises first.
- Failed projects can still qualify. The test is the investigative process, not the commercial outcome.
- Documentation is the claim. The R&D tax credit is part of what makes Ireland an attractive place to build a fintech company. Using it well means claiming accurately and being ready to defend what you've submitted.
The companies that find R&D claims straightforward aren't doing more paperwork. They've built a habit of recording what they're already doing. If you'd like help working out whether your R&D programme qualifies, get in touch and we'll walk you through it.
