Your Project Failed. What Does That Mean For Your R&D Tax Credit?
If your company has a pile of projects that didn’t quite make it, you’re not alone. A prototype that cracked under load testing, a software build the team scrapped after six months, or a new formulation that didn’t meet shelf-life expectations can all get pushed to the side because business owners see them as sunk costs. For R&D tax credit purposes, that can be an expensive mistake.
The research credit under IRC §41 rewards the attempt to solve a technical problem, not the final result of those attempts. Nothing in the statute or regulations requires the attempt to result in a success. In fact, failed projects are often some of the clearest examples of qualified research that companies possess, because the failure itself shows the team didn’t know exactly what would work, pointing to uncertainty.
What the Regulations Say
Under IRC §41(d)(1)(A), qualified research must involve expenditures treated as domestic research or experimental expenditures under §174A. Treas. Reg. §1.174-2(a)(1), addresses failure directly:
"The ultimate success, failure, sale, or use of the product is not relevant to a determination of eligibility under section 174."
Treas. Reg. §1.41-4(a)(5)(i) measures uncertainty as of the beginning of the taxpayer’s research activities. If your team didn’t know at the start whether it could reach the result, or how, a failed outcome at the end doesn’t erase that uncertainty.
What a Failure Can and Cannot Prove
A failed project often strengthens a claim. If your team tried three approaches and none of them worked, that’s good evidence the answer was unclear at the outset. This is exactly what the uncertainty element of the four-part test asks about.
Just as failure doesn’t guarantee that your business will not qualify for R&D tax credits, it also doesn’t guarantee that they will. Treas. Reg. §1.41-4(a)(5)(i) states that uncertainty about a business component does not establish that every activity undertaken to develop it is a process of experimentation, so a failed project must still meet each part of the four-part test.
• Permitted Purpose: The activity must be intended to develop or improve the functionality, performance, reliability, or quality of a business component.
• Technological in Nature: The work must rely on engineering, physical or biological science, or computer science principles.
• Technical Uncertainty: The activity must be undertaken to discover information intended to eliminate uncertainty about the capability, method, or appropriate design of the business component.
• Process of Experimentation: Your team must have a process that evaluates alternatives through modeling, simulation, systematic trial and error, or other methods to eliminate uncertainty.
The reason a project failed also isn’t relevant. A product can clear every technical hurdle and still fail because customers didn’t want it, a competitor beat it to market, or the budget ran out. The technical work behind the project may still qualify. If your project failed without ever involving research, i.e. your team followed a process already proven to work and they simply ran out of time or money, there may not have been any technical uncertainty to resolve in the first place.
Note: market research, customer surveys, and other activities excluded under §41(d)(4)(D) remain excluded whether the project succeeded or not.
The Credit is Built Around the Risk of Failure
The rules for contract research reinforce the same idea. When a company performs research for a customer, the funded research exclusion under IRC §41(d)(4)(H) asks who carries the risk if the research fails. Under Treas. Reg. §1.41-4A(d)(1), payments that are contingent on the success of the research are not treated as funded. A company that only gets paid if the work succeeds bears the financial risk of failure, and that risk is part of what lets that company claim the credit, provided the company also maintains substantial rights to the research results.
What a Qualifying Failure Can Look Like
Failed research shows up in nearly every industry. A few examples of failed research activities that may still qualify for the R&D tax credit include:
• Software and SaaS: A team spends months developing and building new data pipeline architecture, load-tests it, and finds that it can’t meet latency targets at scale, leading the team to scrap it and take a different approach. The abandoned architecture may qualify.
• Manufacturing: A shop designs custom tooling for a new part, runs trials, and the tooling fails under production stresses. The design and testing work may qualify despite the shop not being able to use the new tooling.
• Food, beverage, and consumer products: A company formulates a product around a new ingredient, runs stability testing, and the product doesn’t hold up. The formulation may never launch, but the experimentation behind it may qualify.
• Agriculture: A livestock operation tests an experimental diet against a control group and sees no improvement. The diet didn’t work, but the trial behind it may still qualify.
• Biotech and life sciences: A candidate compound fails toxicity screening or shows no efficacy in early studies. That discovery and preclinical work is still research, whether or not the compound advances into further trial phases.
Why Failed Projects Get Ignored
Though there is nothing in the regulations that state failed projects cannot be qualified R&D activities, they tend to be the activities most often overlooked while compiling a study.
• Claims often get built from conversations about what the company produced that year. Projects that were unsuccessful are often left out of those discussions.
• Once a project is shelved, test data, design files, and notes may get archived or deleted, leading to them being overlooked.
• Engineers, scientists, or other professionals working on the research may have moved on to new work, or a whole new company, by the time the R&D tax credit is being discussed.
• Owners simply assume they don’t count or don’t know that the regulations address failure at all.
How to Document a Project That Didn’t Work
Not every project requires extensive notes and calculations. Even simple notes go a long way when it comes to documenting research, especially failed research.
• Write down what you didn’t know at the start. Even small notes on questions or why the work is being performed in the first place can be helpful.
• If you do reach a successful end, keep notes on what failed. Every failure reinforces evidence of experimentation.
• Note why the project ended. Whether it couldn’t hit a performance target or there was no measurable improvement to a product, this ties the outcome back to technical uncertainty.
• Track who worked on what and the duration of the work, even informally.
• Separate the costs of materials consumed in prototype or test runs from your “everyday” consumables that you would’ve purchased regardless.
• If you’re shelving a project or an employee moves on, archive data, design iterations, or notes instead of simply deleting them.
Frequently Asked Questions
Can I claim the R&D tax credit for a project that failed?
Yes, if the work meets the four-part test. Treas. Reg. §1.174-2(a)(1) states that the ultimate success, failure, sale, or use of a product is not relevant to whether its costs are research expenditures.
Does a failed project automatically qualify for R&D tax credits?
No. While failure does generally show uncertainty in an activity, it must still meet all four parts of the four-part test, not just the uncertainty prong.
Do prototypes destroyed or damaged during testing count?
They can. Treas. Reg. §1.174-2 treats the cost of producing models built to resolve design uncertainty as research expenditures, even when testing leaves them inoperable.
What if the project failed for business reasons, not technical reasons?
The IRS will not look at the reason behind the failure, but whether the technical work qualifies. The scientific work will be evaluated on its own. Activities like market research and consumer surveys will be excluded either way.
Can I go back and claim a failed project from a prior year?
Generally yes. Business can amend returns for open tax years, but refund claims involving the research credit must have the proper documentation of activities performed, individuals who performed them, information they sought to discover, and the related expenses.
How RK Partners Approaches Failed Projects
When we work with clients to build an R&D tax credit study, we ask about all the projects your company worked on, successful or not. Our interviews with your team are designed to surface shelved prototypes, abandoned builds, and unsuccessful trials, and we test each one against the four-part test.
Our team is comprised of tax attorneys, CPAs, and engineers with deep experience in R&D tax credits. We’ll work with your CPA to make sure that we have all the information we need to build a defensible study.
Even if your team performed research that might’ve failed, that’s no reason to assume that you won’t qualify for the credit. The initial conversation is risk-free, and might lead you to identify credits that reward your company’s innovation.
Your Project Failed. What Does That Mean For Your R&D Tax Credit?
If your company has a pile of projects that didn’t quite make it, you’re not alone. A prototype that cracked under load testing, a software build the team scrapped after six months, or a new formulation that didn’t meet shelf-life expectations can all get pushed to the side because business owners see them as sunk costs. For R&D tax credit purposes, that can be an expensive mistake.
The research credit under IRC §41 rewards the attempt to solve a technical problem, not the final result of those attempts. Nothing in the statute or regulations requires the attempt to result in a success. In fact, failed projects are often some of the clearest examples of qualified research that companies possess, because the failure itself shows the team didn’t know exactly what would work, pointing to uncertainty.
What the Regulations Say
Under IRC §41(d)(1)(A), qualified research must involve expenditures treated as domestic research or experimental expenditures under §174A. Treas. Reg. §1.174-2(a)(1), addresses failure directly:
"The ultimate success, failure, sale, or use of the product is not relevant to a determination of eligibility under section 174."
Treas. Reg. §1.41-4(a)(5)(i) measures uncertainty as of the beginning of the taxpayer’s research activities. If your team didn’t know at the start whether it could reach the result, or how, a failed outcome at the end doesn’t erase that uncertainty.
What a Failure Can and Cannot Prove
A failed project often strengthens a claim. If your team tried three approaches and none of them worked, that’s good evidence the answer was unclear at the outset. This is exactly what the uncertainty element of the four-part test asks about.
Just as failure doesn’t guarantee that your business will not qualify for R&D tax credits, it also doesn’t guarantee that they will. Treas. Reg. §1.41-4(a)(5)(i) states that uncertainty about a business component does not establish that every activity undertaken to develop it is a process of experimentation, so a failed project must still meet each part of the four-part test.
• Permitted Purpose: The activity must be intended to develop or improve the functionality, performance, reliability, or quality of a business component.
• Technological in Nature: The work must rely on engineering, physical or biological science, or computer science principles.
• Technical Uncertainty: The activity must be undertaken to discover information intended to eliminate uncertainty about the capability, method, or appropriate design of the business component.
• Process of Experimentation: Your team must have a process that evaluates alternatives through modeling, simulation, systematic trial and error, or other methods to eliminate uncertainty.
The reason a project failed also isn’t relevant. A product can clear every technical hurdle and still fail because customers didn’t want it, a competitor beat it to market, or the budget ran out. The technical work behind the project may still qualify. If your project failed without ever involving research, i.e. your team followed a process already proven to work and they simply ran out of time or money, there may not have been any technical uncertainty to resolve in the first place.
Note: market research, customer surveys, and other activities excluded under §41(d)(4)(D) remain excluded whether the project succeeded or not.
The Credit is Built Around the Risk of Failure
The rules for contract research reinforce the same idea. When a company performs research for a customer, the funded research exclusion under IRC §41(d)(4)(H) asks who carries the risk if the research fails. Under Treas. Reg. §1.41-4A(d)(1), payments that are contingent on the success of the research are not treated as funded. A company that only gets paid if the work succeeds bears the financial risk of failure, and that risk is part of what lets that company claim the credit, provided the company also maintains substantial rights to the research results.
What a Qualifying Failure Can Look Like
Failed research shows up in nearly every industry. A few examples of failed research activities that may still qualify for the R&D tax credit include:
• Software and SaaS: A team spends months developing and building new data pipeline architecture, load-tests it, and finds that it can’t meet latency targets at scale, leading the team to scrap it and take a different approach. The abandoned architecture may qualify.
• Manufacturing: A shop designs custom tooling for a new part, runs trials, and the tooling fails under production stresses. The design and testing work may qualify despite the shop not being able to use the new tooling.
• Food, beverage, and consumer products: A company formulates a product around a new ingredient, runs stability testing, and the product doesn’t hold up. The formulation may never launch, but the experimentation behind it may qualify.
• Agriculture: A livestock operation tests an experimental diet against a control group and sees no improvement. The diet didn’t work, but the trial behind it may still qualify.
• Biotech and life sciences: A candidate compound fails toxicity screening or shows no efficacy in early studies. That discovery and preclinical work is still research, whether or not the compound advances into further trial phases.
Why Failed Projects Get Ignored
Though there is nothing in the regulations that state failed projects cannot be qualified R&D activities, they tend to be the activities most often overlooked while compiling a study.
• Claims often get built from conversations about what the company produced that year. Projects that were unsuccessful are often left out of those discussions.
• Once a project is shelved, test data, design files, and notes may get archived or deleted, leading to them being overlooked.
• Engineers, scientists, or other professionals working on the research may have moved on to new work, or a whole new company, by the time the R&D tax credit is being discussed.
• Owners simply assume they don’t count or don’t know that the regulations address failure at all.
How to Document a Project That Didn’t Work
Not every project requires extensive notes and calculations. Even simple notes go a long way when it comes to documenting research, especially failed research.
• Write down what you didn’t know at the start. Even small notes on questions or why the work is being performed in the first place can be helpful.
• If you do reach a successful end, keep notes on what failed. Every failure reinforces evidence of experimentation.
• Note why the project ended. Whether it couldn’t hit a performance target or there was no measurable improvement to a product, this ties the outcome back to technical uncertainty.
• Track who worked on what and the duration of the work, even informally.
• Separate the costs of materials consumed in prototype or test runs from your “everyday” consumables that you would’ve purchased regardless.
• If you’re shelving a project or an employee moves on, archive data, design iterations, or notes instead of simply deleting them.
Frequently Asked Questions
Can I claim the R&D tax credit for a project that failed?
Yes, if the work meets the four-part test. Treas. Reg. §1.174-2(a)(1) states that the ultimate success, failure, sale, or use of a product is not relevant to whether its costs are research expenditures.
Does a failed project automatically qualify for R&D tax credits?
No. While failure does generally show uncertainty in an activity, it must still meet all four parts of the four-part test, not just the uncertainty prong.
Do prototypes destroyed or damaged during testing count?
They can. Treas. Reg. §1.174-2 treats the cost of producing models built to resolve design uncertainty as research expenditures, even when testing leaves them inoperable.
What if the project failed for business reasons, not technical reasons?
The IRS will not look at the reason behind the failure, but whether the technical work qualifies. The scientific work will be evaluated on its own. Activities like market research and consumer surveys will be excluded either way.
Can I go back and claim a failed project from a prior year?
Generally yes. Business can amend returns for open tax years, but refund claims involving the research credit must have the proper documentation of activities performed, individuals who performed them, information they sought to discover, and the related expenses.
How RK Partners Approaches Failed Projects
When we work with clients to build an R&D tax credit study, we ask about all the projects your company worked on, successful or not. Our interviews with your team are designed to surface shelved prototypes, abandoned builds, and unsuccessful trials, and we test each one against the four-part test.
Our team is comprised of tax attorneys, CPAs, and engineers with deep experience in R&D tax credits. We’ll work with your CPA to make sure that we have all the information we need to build a defensible study.
Even if your team performed research that might’ve failed, that’s no reason to assume that you won’t qualify for the credit. The initial conversation is risk-free, and might lead you to identify credits that reward your company’s innovation.
