Most business owners assume R&D tax credits are only for successful projects. The logic feels obvious: if you spent money developing something that never made it to market, why would HMRC give you money back?
But that assumption is wrong. And it costs UK businesses millions in unclaimed relief every year.
HMRC's R&D rules are built around one question: did you attempt to resolve a technological uncertainty? Not did you succeed. Not did it sell. Not did it generate profit. The commercial outcome is irrelevant to the technical test.
This article explains exactly how to claim R&D tax credits when your project failed commercially, what HMRC looks for, and where most claims go wrong.
What HMRC Actually Requires for an R&D Claim
The legal definition of R&D for tax purposes comes from the BEIS (Department for Business, Energy and Industrial Strategy) guidelines, which HMRC follows. The core test has three parts:
- You sought an advance in science or technology
- You faced a technological uncertainty that could not be resolved by a competent professional in the field
- You attempted to resolve that uncertainty through a defined project
Notice what is not in that list. Commercial success. Market viability. Revenue generation. Profitability. None of it.
As ICAEW qualified accountants, we regularly see clients who spent £50,000, £100,000 or more on R&D projects that failed technically or commercially. They assume the money is lost for relief purposes. In most cases, it is not.
Why Failed Projects Still Qualify
R&D tax credits exist to encourage innovation. Innovation is inherently risky. If HMRC only rewarded successful projects, the relief would only go to businesses that already knew the answer before they started. That defeats the purpose.
The key distinction is between technological failure and commercial failure.
Technological failure means you tried to achieve an advance but could not. For example, you attempted to develop a new composite material that could withstand 800 degrees Celsius, but after 18 months of testing, every formulation failed above 650 degrees. That is a technological failure. It still qualifies for relief because you attempted the advance and faced genuine uncertainty.
Commercial failure means the technology worked but the product did not sell. You developed a new software algorithm that processed data 40% faster than existing solutions. It worked. But the market did not adopt it. Competitors launched cheaper alternatives. You shut the product down. That project still qualifies for R&D relief because the technological advance existed regardless of commercial outcomes.
Both scenarios qualify. The only scenario that does not is where no genuine technological uncertainty existed in the first place.
Real Example: A Bristol Engineering Firm
Take a real case we worked on. An engineering company in Bristol spent £87,000 developing a new type of hydraulic valve for offshore wind turbines. The valve worked in lab conditions but failed catastrophically in saltwater field trials after 14 months. The project was abandoned.
The directors assumed they could not claim R&D relief because the project failed. They were wrong.
The company had attempted to solve a genuine technological uncertainty: how to maintain seal integrity under continuous saltwater exposure at depths exceeding 40 metres. No competent engineer in the field knew the answer at the start. The company designed prototypes, ran simulations, built test rigs and conducted field trials. All of that activity qualifies as R&D.
The claim returned £16,530 in corporation tax relief. The company had written the project off as a loss. That money went back into the business for a different project.
What HMRC Scrutinises in Failed Project Claims
Failed project claims do attract more HMRC enquiries. Not because they are invalid, but because HMRC wants to confirm the project genuinely involved technological uncertainty and was not just routine development that did not work out.
HMRC will typically look at:
- The technological baseline. What was the state of knowledge before the project started? What specifically could existing technology not do?
- The uncertainties identified. What specific technical problems did you not know how to solve at the outset? Name them explicitly.
- The activities undertaken. What did you actually do? Design work, prototyping, testing, analysis, iteration. HMRC wants the detail, not a one-line description.
- The records created. Technical reports, test results, design documents, emails between engineers discussing problems. The more evidence you have, the stronger the claim.
A common mistake we see is businesses describing the failure without describing the technological attempt. Saying "the product did not work" is not enough. You need to explain what you tried to achieve, why it was uncertain, and how you went about solving it.
The Documentation You Need for a Failed Project Claim
Documentation matters more for failed projects because there is no commercial outcome to point to as evidence the work happened. You need to prove the R&D activity occurred through technical records.
Strong documentation includes:
- Project initiation documents showing the technical objectives and the uncertainties identified
- Design records showing iterations, design choices and why certain approaches were abandoned
- Test results showing what was tested, what failed and what was learned
- Technical meeting notes showing the problems the team discussed and the decisions made
- Time records showing which staff worked on which technical activities
- Supplier invoices for materials, subcontractors or external testing
You do not need a perfect paper trail. HMRC accepts reasonable evidence. But the less documentation you have, the harder it is to prove the project involved genuine R&D rather than routine problem-solving.
If you are reading this after the project has already failed and you did not keep great records, do not give up. You can still compile evidence retrospectively from emails, calendar entries, supplier records and staff recollections. It is harder but still possible.
How the Relief Is Calculated for Failed Projects
The calculation works exactly the same way as for successful projects. There is no special "failed project" rate.
For profit-making companies, the R&D expenditure reduces your taxable profits. If you spent £50,000 on qualifying R&D costs and your corporation tax rate is 25%, you save £12,500 in tax. The enhanced deduction (up to 186% for pre-April 2024 claims, or the merged RDEC-style scheme from April 2024) applies regardless of commercial outcome.
For loss-making companies, you can surrender the R&D loss for a cash payment. The payable credit is calculated at the relevant rate (typically 14.5% or 10% depending on the scheme and period). A loss-making company that spent £50,000 on R&D could receive a cash payment of several thousand pounds, even though the project failed commercially.
That cash is not a loan. It is not repayable. It is tax relief for genuine innovation that did not work out.
We have worked with a software consultancy in Manchester that spent £63,400 developing a new data visualisation tool. The tool worked technically but the market was not there. The company was loss-making at the time. The R&D claim returned £9,193 in cash. That paid for three months of rent while they pivoted to a different product.
Common Mistakes on Failed Project Claims
We see the same errors repeatedly:
Mistake 1: Not claiming because you think failure disqualifies you. This is the biggest one. If you attempted a genuine technological advance and it did not work, you almost certainly have a valid claim.
Mistake 2: Describing the failure instead of the attempt. Your R&D narrative should focus on what you tried to achieve and how you went about it, not on why the product did not sell.
Mistake 3: Including non-qualifying costs. Failed projects often involve significant marketing research, market testing or customer feedback work. Those are not R&D. Separate your qualifying technical costs from your commercial costs.
Mistake 4: Assuming subcontractor costs are always qualifying. If you subcontracted the R&D work to a third party, the rules on qualifying costs differ depending on the scheme and whether the subcontractor is connected to you. Get this wrong and the claim collapses.
Mistake 5: Not claiming because the project was abandoned early. Even a project that ran for three months and was stopped can qualify, provided genuine R&D activity occurred in that period.
When a Failed Project Does NOT Qualify
There are situations where a failed project will not attract R&D relief. The main ones are:
- No technological uncertainty existed. You were doing routine development using established methods, and it simply did not work. That is not R&D.
- The project was purely commercial. You tried a new marketing strategy, a new pricing model or a new distribution channel. Those are commercial activities, not technological advances.
- The work was in the arts, humanities or social sciences. R&D for tax purposes is limited to science and technology. A failed novel or a failed social media campaign does not qualify.
- The project was routine software development. Building a standard ecommerce site using existing frameworks, even if it failed, is not R&D. The test is whether you had to resolve genuine technical uncertainties beyond what a competent developer could handle.
If you are unsure whether your project meets the threshold, speak to an accountant who specialises in R&D claims. A quick review of your project documentation will usually give a clear answer.
How to Structure Your Claim for a Failed Project
If you have a failed project you want to claim for, here is the practical process:
- Identify the qualifying projects. Go through every failed technical project from the last two accounting periods. R&D claims can be made up to two years after the end of the accounting period the work fell in.
- Gather the technical narrative. For each project, write a clear description of the technological advance you attempted, the uncertainties you faced, and the activities you undertook. Focus on the technical detail.
- Calculate the qualifying costs. Staff costs, consumables, software licences, subcontractor costs, externally provided workers. Use the correct categories and exclude non-qualifying items.
- Prepare the claim. Submit the claim through your corporation tax return (CT600) or by amending a previously filed return. Include the additional information form (R&D AIF) if required for accounting periods starting on or after 1 April 2024.
- Keep the evidence. Store all project documentation, cost records and technical narratives. HMRC can open an enquiry up to 12 months after you submit the claim.
Our R&D tax credits page has more detail on the specific costs that qualify and the rates that apply.
What If HMRC Opens an Enquiry?
Failed project claims are more likely to trigger an HMRC enquiry. That does not mean the claim is invalid. It means HMRC wants to understand the technical basis before agreeing the relief.
If you receive an enquiry letter, do not panic. Your accountant will prepare a technical response with supporting evidence. In most cases, HMRC accepts the claim once they see the detail of the technological attempt.
The key is having the documentation ready. If you claimed without a proper technical narrative, the enquiry becomes harder to defend. That is why we recommend preparing the technical case at the time of the claim, not after HMRC writes to you.
Final Thoughts
R&D tax credits reward the attempt, not the outcome. If your business spent money trying to advance science or technology and it did not work commercially, you still have a valid claim.
The mistake most businesses make is assuming failure means no relief. It does not. The relief exists precisely because innovation is uncertain and failure is part of the process.
If you have a failed project from the last two years that you have not claimed for, review it against the technological uncertainty test. You may be leaving money on the table.
For a full breakdown of the rules and rates, visit our R&D fundamentals page. If you want to discuss a specific project, contact our team for a confidential review.