If HMRC ever challenges your R&D claim, the outcome won't depend on how innovative your projects felt from the inside; it will depend on what you can prove. The First-tier Tribunal's decision in Beer Express Ltd v HMRC shows how that plays out in practice.
The case involved a drinks wholesaler with four projects, including an AI-enabled forecasting system, and it failed on evidence rather than on any technical detail. Here's what happened, and what you can learn for your own claim.
What happened in Beer Express v HMRC?
An R&D tax adviser reached out to Beer Express Ltd, a wholesaler of alcoholic drinks, suggesting they could qualify for R&D tax credits. With the adviser’s help, they prepared a R&D tax relief claim for two accounting periods, ended 30 June 2020 and 30 June 2021. The claims covered four projects: an AI-enabled stock forecasting system, a new pale lager, a move from kegs to bottles with rebranding and new logistics software, and a business-to-business web platform.
HMRC rejected both claims in full. Unfortunately, the adviser was no longer contactable at the time of the rejection, so Beer Express was required to defend their claim alone.
The company appealed the rejection, and the Tribunal dismissed the appeal in May 2026, deciding that the company hadn't proved that they were carrying out R&D. The burden of proof sits with the claimant, and that's the point every other claimant should take from this case.
Why did the Tribunal reject the claim?
The Tribunal applied the government's guidelines on what counts as R&D for tax purposes, the same core definition that applies to claims today. It found three gaps in the evidence:
- No competent professional explained whether existing technology was enough for the company's needs.
- There was no convincing evidence that the activities went beyond standard development.
- There was no detailed explanation of the uncertainties, or why a professional in the field couldn't have resolved them readily.
None of these is about how much work the company did. They're about whether the company could show, with evidence, why the work was R&D. If you'd like a refresher on the underlying tests, our guides to what counts as an advance and scientific or technological uncertainty cover them.
Lesson 1: The person who speaks to the project matters
The evidence came from the company's founder, who oversaw strategy and reviewed the claims but had limited technical knowledge of the projects. He couldn't recall specific technical difficulties, and he couldn't say whether the software went beyond what competitors offered. The former shareholder who led the software work was not available to give further evidence, and another person named in the expenditure reports had no identified role or expertise.
The guidelines don't mandate particular qualifications, but the Tribunal said a competent professional needs to show “appropriate qualifications, experience and up-to-date knowledge” of the relevant principles. It also expected evidence from someone who was involved at the time and could be questioned about the work. The software lead was identified as having relevant experience, but the Tribunal said there was no evidence from him about his skills, how they related to the project, or what he had actually done.
The accounting periods ended in 2020 and 2021, and the decision came in 2026. By the time of the hearing, key people had moved on, and the company couldn't call them.
That's why we encourage every claimant to name the right person for each project and write a short summary of their experience. Our guide to what a competent professional is explains how to choose them. It’s also critical to get a full description of the work done and evidence that can prove your uncertainties and advances at the time of preparing the claim, in case you can’t rely on the testimony of your competent professionals later down the line.
Lesson 2: Having confidence in your adviser
The reports prepared by the adviser described the projects and the uncertainties. The founder confirmed they were accurate, but the Tribunal described them as “no more than bald assertions unsupported by evidence from a competent professional with contemporaneous involvement in the projects”. Since the advisers who prepared them weren't there to explain, the claim was on unsteady ground.
When these claims were prepared, the R&D tax credit scheme was less regulated than it is, allowing some advisers to reach out to companies and prepare claims based on very little information. In this case, the company’s director wasn’t sure if the adviser even spoke with the development lead.
Trusting your adviser to make a genuine, robust claim is just as important as making the claim itself. If you choose to use an adviser for your claim, you need to be certain that they aren’t just making a claim so they can cash in their fee.
The main thing to check with your adviser is what support they offer if your claim is queried or rejected. Tax Cloud offers total enquiry support, at no extra charge, with every claim we submit. Though we don’t expect HMRC to challenge your eligibility, a random check into your claim could result in unnecessary hours spent defending your claim; we take care of this for you.
Lesson 3: New to you isn't the same as new to the field
The Tribunal found no convincing evidence that developing the lager went beyond a brewer trialling a different flavour. It also found nothing showing how moving from kegs to bottles went beyond products already in existence. On the forecasting system, there was no clear evidence of how using AI to adapt software to the company's needs went beyond the existing baseline of knowledge.
This reflects a principle in HMRC's guidelines: “the routine analysis, copying or adaptation of an existing process, material, device, product or service will not advance overall knowledge or capability, even though it may be completely new to the company.” Trial and error can be part of genuine R&D, but it needs a documented baseline of what was already possible, and a clear account of what wasn't. Many companies confuse challenges with uncertainty; HMRC is very clear that your difficulties must lie in whether you can scientifically or technologically carry out your goals, not just hard for your business to carry out.
If you use off-the-shelf tools in your projects, our guide to claiming R&D relief when using off-the-shelf tools is a good next read.
What should you do differently?
Here's a checklist for each project in your claim:
- Name the person with hands-on involvement. Write two or three lines on their experience and how it relates to the project.
- Record the baseline at the start. Note what existing technology could do when the project began, and where it fell short.
- Describe the uncertainties in specific terms. Explain what you didn't know how to do, and why a competent professional couldn't have worked it out from existing knowledge.
- Keep dated records of the work. Test results, failed attempts and design changes all count, especially when first-hand testimony can’t be used.
- Build the report with your technical team. Make sure someone can explain every statement in it.
Key takeaways
- The burden of proof is yours. You need to show on the balance of probabilities that each project met the definition of R&D.
- Evidence has to come from someone who was there. An oversight role isn't enough, and neither is a report entirely written by someone who wasn't involved.
- Reports need documents behind them. The Tribunal called an unsupported report “bald assertions”.
- Plan for the long term. Claims can be challenged years later, so capture accounts from key staff while the work is fresh.
Good evidence doesn't have to be elaborate. A short summary from the right person, a written baseline and dated test results carry more weight than pages of description. If you'd like help building the evidence behind your claim, get in touch and we'll walk you through it.
