In recent years, the world of R&D tax relief has entered a new era of scrutiny. While the relief itself has been in place for over two decades, the number of cases reaching tribunal has increased sharply.

Simon Briton’s webinar “R&D Tax Tribunal Cases – What Accountants Need to Know” shed light on what’s driving these disputes — and, more importantly, what themes consistently emerge from tribunal rulings. The message for accountants is clear: understanding these patterns is essential to building claims that can stand up to HMRC’s increasingly robust challenges.

A Shift Toward Litigation

For much of the scheme’s history, R&D tax disputes were quietly resolved through negotiation or informal correspondence with HMRC. But that landscape has changed.

Tribunal hearings are now testing not only the letter of the law but also the practical boundaries of what HMRC will accept as “qualifying R&D.” While first-tier tribunal outcomes don’t create binding precedent, they provide a powerful signal of how future cases may be interpreted — and how accountants should frame their claims to withstand review.

1. True Uncertainty Must Be Technological

A recurring theme is the tribunal’s insistence that qualifying uncertainty must be scientific or technological, not commercial.
It’s not enough that a company has done something new in its marketplace or developed a product that’s “first in the industry.”

The tribunal consistently distinguishes between commercial novelty and technological advance. Claims fail where innovation is defined by business context rather than by genuine scientific or technical problem-solving.

Simon notes that successful cases strip away marketing language and focus solely on the technological problem and the method of overcoming it. If the work wouldn’t stretch the capabilities of a competent professional in the field, it won’t pass muster.

2. Real-Time Evidence Matters

Another dominant theme is the importance of contemporaneous documentation.
Tribunals give little weight to retrospective explanations written as part of tax compliance. What convinces them are design notes, test logs, or technical records produced during the project — tangible proof that the company was engaged in research and development, not simply reporting it after the fact.

Documents should demonstrate:

• The problem being tackled;
• Why it was technically difficult;
• How the team sought to overcome that challenge; and
• Why the outcome represents an advance in capability or knowledge.

In short, the evidence should tell the story of experimentation, not just its results.

3. Incremental Isn’t Always Innovative

Tribunals frequently draw a line between incremental development and genuine R&D.
Small, iterative improvements are common in most industries, but they only qualify when they involve genuine technical uncertainty. If a competent professional could have achieved the result through routine methods, the tribunal is likely to side with HMRC.

The benchmark is not scale of change but the nature of the challenge. Even modest projects can qualify if they overcome a gap in existing scientific or technical understanding.

4. Be Precise About People

One of HMRC’s biggest areas of contention — echoed by tribunals — is who counts as R&D staff.
Claims that include generic percentages of all employees’ time have consistently been struck down. The tribunal expects a clear link between the qualifying work and the people performing it.

That means:

• Identifying only those who directly tackled scientific or technological uncertainties;
• Excluding administrative or commercial testing roles; and
• Providing evidence of what each contributor actually did.

There’s nuance here: overemphasising staff expertise can backfire by implying the task wasn’t truly uncertain. Briton’s advice — be specific, not inflated.

5. Software Development: The Hotspot

No area of R&D has seen more disputes than software.
HMRC remains cautious, often sceptical, about software claims — particularly when they involve known frameworks or standard coding practices.

Tribunals have rejected claims where businesses simply adapted existing technologies for a new commercial use. Instead, the qualifying activity must relate to solving technical problems such as system scalability, algorithm efficiency, or data architecture improvements.

Work focused on design, UX, or customer behaviour doesn’t qualify — those are commercial challenges, not technological ones.

6. Cost Allocation: Precision Over Percentages

The tribunal record is equally clear on expenditure. Blanket allocations — “20% of all materials” or “half of the developer’s salary” — do not meet evidential standards.

Each cost must be tied to a specific qualifying activity.
Rather than estimate averages, claimants should record when and for how long staff worked on R&D tasks. Even short, concentrated efforts are preferable to imprecise annual percentages.

Simon's takeaway: accuracy outweighs ambition. A small, well-supported claim is far safer than an inflated one built on guesswork.

Preparing Every Claim as Tribunal-Ready

Perhaps the most practical insight from Simon's analysis is philosophical rather than procedural:

“Every R&D submission to HMRC should be prepared as if it might one day be read out in a tribunal.”

That mindset changes everything — from the language used in the claim report to the depth of evidence maintained throughout the project.

It also encourages alignment between accountants, clients, and technical specialists from day one. By embedding compliance into the R&D process rather than retrofitting it afterwards, firms reduce risk and improve claim credibility.

Conclusion: Tribunals Are Teaching Us How to Get It Right

While first-tier decisions don’t set precedent, the growing body of tribunal rulings serves as a practical guidebook for advisers.
They reinforce that R&D tax relief is not a reward for being innovative in business terms — it’s an incentive for overcoming genuine scientific or technical barriers.

For accountants, that means sharpening definitions, evidencing claims in real time, and resisting the temptation to generalise.
Tribunals aren’t trying to narrow the scheme — they’re clarifying what true R&D looks like. Those who listen are far better positioned to defend their clients, and their claims, when HMRC comes calling.

FAQ:

1. Why are R&D tax tribunals getting more attention lately?

The number of R&D cases reaching tribunal has risen sharply in the past few years. HMRC’s reviews are tougher, and many disputed claims are no longer resolved through back-and-forth negotiation.
Although first-tier tribunal rulings don’t set precedent, they reveal how judges interpret the legislation — providing essential insight for advisers who want to build resilient, defensible claims.

2. What kind of uncertainty qualifies as “R&D”?

The tribunals have drawn a hard line: only scientific or technological uncertainty qualifies.
It’s not enough that a business has done something innovative for its market or customers. The question is whether the project involved overcoming a technical challenge that couldn’t be solved with existing knowledge or methods.
If the difficulty lies in the business model or commercial risk, it won’t meet the R&D threshold.

3. How can I demonstrate genuine technological uncertainty?

You need to show that the team was working to resolve a problem with no obvious or established solution.
Tribunals want to see that the business had to “go off-piste” — tackling something outside its usual operations or engineering practices. If a competent professional could have solved it using readily available techniques, the claim is unlikely to hold up.

4. What kind of evidence should be included with an R&D claim?

Tribunals consistently favour real-time, contemporaneous documentation.
Keep design notes, experiment logs, and test results from during the project — not just a polished report produced later.

A strong record should answer:

• What was the problem?
• Why was it technically hard to solve?
• What approaches were tested or rejected?
• What made the outcome an advance in knowledge or capability?

In short, tell the story of the research process, not just the result.

5. Are small improvements eligible for R&D relief?

They can be — but only if they resolve a genuine scientific or technical unknown.
Incremental tweaks or performance improvements that fall within “business as usual” won’t qualify. Tribunals use the “readily ascertainable” test: if the solution could be reached by normal professional competence, it’s not R&D.

6. Who should be counted in the R&D team for the claim?

Only include staff who actively contributed to resolving the uncertainty.
Generic percentages of everyone’s time — such as “10% of all staff hours” — have been rejected by tribunals.
You’ll need to show:

• Who did the technical work.
• How their contribution related to the problem.
• Why their time or cost is directly linked to the qualifying activity.

Support, admin, and commercial testing roles should be excluded.

7. What’s the biggest challenge with software R&D claims?

Software development remains one of the most disputed areas.
Tribunals and HMRC are clear: reusing existing code, frameworks, or third-party solutions does not qualify. The R&D must involve solving technical issues such as efficiency, scalability, architecture, or data handling — not just deploying known tools in a new market.

User interface (UI) or user experience (UX) work, while important commercially, doesn’t count as R&D.

8. How should R&D expenditure be allocated?

Tribunals expect a direct and traceable connection between costs and qualifying work.
Rough estimates like “20% of materials” or “half of one employee’s time” are not acceptable without evidence.
Track activity periods carefully, note when qualifying work took place, and link costs to those timeframes.
Even short, well-documented bursts of qualifying activity are more defensible than broad averages.

9. What’s the right way to prepare for a potential dispute?

Treat every claim as if it could end up before a tribunal.
That means:

• Writing reports that focus on technical substance, not marketing language.
• Keeping supporting evidence throughout the project.
• Making sure every cost and role is justified with a clear link to the qualifying activity.

If HMRC does open an enquiry, having this groundwork already in place makes a significant difference.

10. What’s the difference between First-Tier and Upper Tribunal decisions?

The First-Tier Tribunal (FTT) provides useful guidance but doesn’t create binding precedent.
The Upper Tribunal (UT), on the other hand, sets legal precedent for future cases.
While HMRC sometimes downplays FTT decisions, advisers can still learn a great deal from their reasoning when shaping claim strategy and risk assessment.

11. What are the most common reasons claims are rejected?

According to Simon's analysis, the main pitfalls are:

• Framing uncertainty as commercial rather than technical.
• Weak or retrospective evidence.
• Including staff or costs with no clear link to qualifying R&D.
• Presenting incremental development as innovation.
• Submitting overly broad or generic claims.

Avoiding these traps is the single best way to reduce audit and tribunal risk.

12. What’s the key takeaway for accountants?

The overarching message is preparation.
As Simon Briton advises, every R&D submission should be built with tribunal-level scrutiny in mind.
That means prioritising evidence, accuracy, and transparency — not just optimism.
When accountants help clients build technically grounded claims supported by real proof, the process becomes smoother, and the risk of costly disputes drops dramatically.

The contents of this article are meant as a guide only and are not a substitute for professional advice. The author/s accept no responsibility for any action taken, or refrained from, as a result of the material contained in this document. Specific advice should be obtained before acting or refraining from acting, in connection with the matters dealt with in this article. The information at the time of publishing was accurate and could be subject to final changes.