Most business owners assume R&D tax credits are only for projects that succeed commercially. That is wrong. HMRC's R&D rules do not require a project to make money, sell a product, or even reach the market. The test is whether you attempted to resolve scientific or technological uncertainty. If you did, the project qualifies regardless of the commercial outcome.
This misconception costs UK businesses millions in unclaimed relief every year. A failed project often involves more technical effort than a straightforward one, because you had to try multiple approaches, test dead ends, and work through problems that turned out to be unsolvable with current technology. All of that counts.
As ICAEW qualified accountants, we regularly see claims rejected because the narrative focuses on the business failure rather than the technical work. This article explains the specific rule, the evidence you need, and the three mistakes that kill otherwise valid claims.
What HMRC Actually Requires for R&D Tax Credits
The legal definition of R&D for tax purposes comes from the BEIS Guidelines (Department for Business, Energy and Industrial Strategy). The guidelines are consistent across both the old SME scheme and the merged RDEC scheme that applies from 1 April 2024.
To qualify, your project must meet all three of these conditions:
- Technological uncertainty existed. You did not know, at the start, whether the project was achievable or how to achieve it. The answer was not readily deducible by a competent professional in your field.
- You sought a technological advance. The work aimed to improve overall knowledge or capability in your field, not just for your business.
- The work was part of a defined project. It had a start, a planned approach, and a conclusion (even if that conclusion was "we cannot solve this").
Nowhere in the guidelines does it say the project must be commercially successful. Nowhere does it say you must have launched a product, generated revenue, or turned a profit. The test is purely technical.
The Failed Project Rule: What the Legislation Says
HMRC's internal manual at CIRD81900 is explicit. It states that "a project which does not succeed in achieving its aims or which is abandoned before completion can still be R&D." The manual goes on to say that the costs incurred up to the point of abandonment or failure are qualifying expenditure, provided they relate to resolving technological uncertainty.
Consider a real example. A Manchester-based engineering consultancy spent £74,000 developing a new composite material for lightweight structural panels. After 14 months of work, the material failed stress testing at a level well below the target. The client abandoned the project. The commercial outcome was zero. No product, no revenue, no IP sale.
But the technical work was substantial. They had tested four different resin formulations, modified curing processes, and developed new testing rigs. Every iteration resolved a specific technical question, even if the final answer was "this approach does not work." The claim was valid. They received £14,060 in repayable tax credit.
Three Mistakes That Kill Valid Failed-Project Claims
Mistake 1: Framing the Narrative Around the Failure
The most common error. A director writes: "We tried to build a new software platform but it didn't work, so we abandoned it." That is a commercial story, not a technical one. HMRC reads that and sees no R&D.
Rewrite the narrative around the technical work. What was the technological uncertainty? What approaches did you try? What did each test reveal? The fact that the project failed commercially is almost irrelevant to the claim. The technical work is what matters.
Mistake 2: Not Documenting the Dead Ends
R&D claims for successful projects often rely on the final product as evidence. Failed projects lack that anchor. You need contemporaneous records of the work: lab notebooks, test results, meeting minutes, emails discussing technical problems, version control logs, prototype photos.
If you have no documentation, the claim is weak. HMRC will accept retrospective narratives if the work happened, but the more evidence you have, the stronger the claim. A failed project with good documentation is safer than a successful project with none.
Mistake 3: Claiming Costs After the Project Was Abandoned
Once you decide to stop the project, costs stop qualifying. You cannot claim winding-down costs that are purely administrative, like writing a closure report or selling off equipment. The cut-off is the point at which you concluded the technological uncertainty could not be resolved (or was not worth resolving). Costs after that date are not R&D.
Be precise about the abandonment date. If you spent three months after the decision paying a contractor to tidy up code, those costs are not allowable. HMRC will disallow them.
What Costs Can You Claim on a Failed Project?
The same costs that apply to successful projects apply to failed ones. The key categories are:
- Staff costs. Salaries, employer NI, and pension contributions for employees directly working on the project. Apportion on time sheets or project records.
- Consumables. Materials, components, and software licences consumed in the R&D. For software, that includes cloud hosting costs for test environments.
- Externally provided workers. Contractors or agency staff working under your supervision. Different rules apply depending on whether they are under your direction or not.
- Subcontracted costs. Work you paid another company to do on your behalf. The rates differ between the SME scheme and RDEC.
- Software licence costs. If the software was used directly for the R&D.
Capital expenditure does not qualify under the revenue-based R&D schemes. You claim capital allowances separately.
For a failed project, the total qualifying costs are often lower than a successful one because you stopped earlier. But the per-pound value of the claim can be higher if you were loss-making, because you may receive a repayable tax credit rather than just a reduction in corporation tax.
How the Relief Works for a Failed Project
The mechanics depend on which scheme you are under.
For accounting periods starting before 1 April 2024 (old SME scheme):
- If you are profit-making: you deduct 186% of qualifying costs from trading profits, reducing corporation tax. If the deduction creates or increases a loss, you surrender the loss for a cash credit of up to 14.5% of the surrendered amount.
- If you are loss-making: you surrender the enhanced R&D loss for a cash payment. The payable credit is capped at the lower of £20,000 plus 3x your total PAYE/NI liability, or the actual surrender value.
For accounting periods starting on or after 1 April 2024 (merged scheme):
- All companies use the RDEC mechanism. You get a 20% above-the-line credit on qualifying costs (15% for the old RDEC, 20% from 1 April 2024). The credit is taxable as income.
- If your company is loss-making, the credit can be surrendered for a cash payment, subject to the PAYE cap.
- If your company is R&D-intensive (more than 30% of total costs on R&D) and loss-making, the Enhanced R&D Intensive Scheme (ERIS) applies, giving a higher payable credit rate.
A failed project often leaves the company with no future income from that work, so the loss-making route is common. The payable credit can be a significant cash injection at exactly the point when the business needs it most.
When a Failed Project Does Not Qualify
Not every failed project qualifies. The failure must be technical, not commercial. If you built a working product that simply did not sell, that is a commercial failure, not a technological one. The R&D tax credit system does not subsidise poor marketing or pricing.
Equally, if the project was routine development that did not attempt a technological advance, failure does not retroactively turn it into R&D. The test is the same: was there technological uncertainty at the start? If the answer is no, the project never qualified, regardless of outcome.
A third scenario: you attempted a technological advance but abandoned the project before any meaningful technical work. If you spent £2,000 on initial research and then stopped, the claim is technically valid but likely too small to be worth the compliance effort. HMRC expects genuine projects with substantive work.
How to Structure a Failed-Project Claim
If you are preparing a claim for a failed project, structure the technical narrative around these four points:
- The technological uncertainty. What specific question could not be answered by existing knowledge? Be precise. "We did not know whether Material X could withstand 200°C without degrading" is better than "We tried to make a better material."
- The approaches attempted. List each technical approach, what you expected, and what actually happened. Show the dead ends. This is the core of the claim.
- The resolution. How did the project conclude? Did you prove the approach was impossible? Did you find a partial solution that was not commercially viable? The conclusion is still a technical outcome.
- The costs. Map the qualifying costs to the period of active technical work. Exclude anything after the abandonment decision.
Worked example: A Bristol-based biotech startup spent £92,800 developing a new diagnostic assay. After 18 months, the assay showed 72% sensitivity, well below the 95% threshold needed for regulatory approval. The company abandoned the project. The technical work included testing 12 different antibody combinations, optimising buffer conditions, and developing a novel signal amplification method. The claim was for staff costs (£63,400), consumables (£21,200), and external testing (£8,200). Total qualifying costs: £92,800. Under the old SME scheme, the enhanced deduction of 186% gave a loss surrender value of approximately £25,000 in cash. The claim was accepted.
Do You Need an Accountant for a Failed-Project Claim?
You can submit a claim yourself. HMRC does not require an accountant to sign off R&D claims. But the failure narrative makes claims harder to write well. The temptation is to describe the business story rather than the technical one, and that gets the claim rejected.
An accountant experienced in R&D claims will know how to extract the technical narrative from your project records, identify all qualifying costs, and structure the claim to survive HMRC scrutiny. They will also handle the computational side, ensuring the correct rates and caps apply.
If the claim is for a significant amount (say, over £20,000 in relief), the cost of professional advice is usually justified by the reduced risk of rejection. A rejected claim can trigger a full HMRC enquiry into your company's tax affairs, which is time-consuming regardless of the outcome.
Our R&D tax credits page explains the process in more detail. If you have a failed project and are unsure whether it qualifies, a conversation with an ICAEW qualified accountant is a low-cost way to find out before you invest time in the paperwork.
Common Questions About Failed-Project R&D Claims
Can I claim if the project was abandoned before completion?
Yes. HMRC's guidance at CIRD81900 confirms that abandoned projects qualify for relief. The key is that you incurred qualifying costs while attempting to resolve technological uncertainty. The abandonment date simply marks the end of the qualifying period.
Does the project need to have been written down as a formal R&D project?
No. Many small businesses do not formally document projects. HMRC accepts retrospective claims provided you can demonstrate the work happened and it sought a technological advance. Contemporaneous records strengthen the claim significantly.
What if the failure was because we ran out of money, not because the technology was impossible?
This is a grey area. If you stopped because of funding constraints before resolving the technological uncertainty, the costs up to that point still qualify. The work was still R&D even if you did not complete it. HMRC will look at whether the work was genuinely R&D at the time it was performed, not at the reason for stopping.
Can I claim for a project that failed but produced useful data or IP?
Yes, absolutely. A failed project that generates patentable data, trade secrets, or technical knowledge has still produced a technological advance. The advance is the knowledge gained, not the product. This is one of the strongest categories of failed-project claims.
For more guidance on R&D claims generally, see our R&D tax credits blog section. If you are considering a claim for a failed project, contact our team for a preliminary assessment.