The R&D Tax Credit 4 Part Test

The research and development (R&D) tax credit is an underutilized tax credit. Many companies that are eligible for this credit aren’t aware that they are eligible. While you may be able to claim the credit fairly easily by consulting with a company that handles the credit for you, eligibility isn’t always super clear.

It’s important to note that your company doesn’t just “qualify” or “not qualify.” Determining eligibility for the R&D tax credit is really about determining what activities your company is doing and if they are eligible to claim. Then, you add up the dollar value of all those activities as the basis for determining how much you eventually receive in R&D credits. Here’s how you actually determine your eligibility:

An Overview of The 4 Part Test

To qualify, a company’s activities must pass the 4 part test. These 4 parts are:

You have to either create from scratch or improve a business component. This means the new thing (either brand new or just improved) has a better function or capability in some way.
Activities have to eliminate “uncertainty” in some way. This sounds a little weird, but what they really mean here is that you can’t make improvements for something that is a known method and call it R&D, because that doesn’t increase our total, overall knowledge of that thing, which is the point of the credit.
The activities done to reduce uncertainty have to have a process of experimentation. Basically if there is uncertainty, you have to try things out to see what works.
This last one is the easiest one: the activities have to be scientific. So they must involve a hard science like biology, chemistry, physics, engineering, or computer science.

Let’s get into the details and examples for each of these parts.

1. A New or Improved Business Component

According to the IRS documentation, a "business component" refers to any product, process, computer software, technique, formula, or invention which is either intended for sale, lease, or license, or used by the taxpayer in their trade or business. Basically, a business component is just something that’s a part of a product that customers pay them for. This part of the test emphasizes either the new creation or enhancement of existing components with the goal of improving the function, performance, reliability, or quality, but importantly it explicitly excludes activities related to style, taste, cosmetic, or seasonal design factors.

To put this into perspective with examples, let's consider a software development company working on a new encryption technology. If this technology enhances the security (function) or speed (performance) of data transmission, it would likely qualify as a new or improved business component. Another example could be a manufacturing firm developing a new metal for automotive parts that improves fuel efficiency (performance) and durability (reliability). Both cases involve either inventing or improving a business component that’s part of their product. Note that this doesn’t necessarily mean these must be completely new products, they can just be parts of a product.

On the other hand, a fashion company designing a new clothing line for the upcoming season, despite potentially involving considerable effort, experimentation, and innovation, would not qualify because the goal there is for cosmetic or aesthetic purposes.

2. Eliminating Uncertainty

This part focuses on the company's efforts to eliminate technological uncertainty about the development or improvement of a business component. According to the IRS, this involves the process of experimentation aimed at discovering information that could eliminate uncertainties regarding the capability, method, or the appropriate design of a new or improved business component.

For instance, a software company developing a new algorithm to enhance data processing speeds is trying to figure out the best way to do this, so they are faced with uncertainty and therefore must experiment to reduce that uncertainty. This scenario passes this part of the test because it is working toward figuring something out that wasn’t known before.

In the above example, we should note two things. First, it doesn’t technically matter if the research works. Because so much of R&D is often figuring out all the ways something doesn’t work before it eventually does, the government still wants to incentivize people to try these things and take risks. So the resources that go into this can still qualify as research activities even if they don’t end up working. Another thing to note here is that this can simply be a new parallel path. In the data processing example, there can be alternative ways to do what the company is trying to achieve, but if they are figuring out a brand new way that still abides by the other parts of the test, that is still a qualifying research activity.

Conversely, a company conducting market research to determine customer preferences for the color of an existing product does not qualify. This activity does not address technological uncertainties related to the development or improvement of a business component but rather focuses on subjective consumer tastes, which are explicitly excluded from qualifying activities.

3. A Process of Experimentation

This requirement is aimed at ensuring that the activities for which the credit is claimed involve a systematic process designed to evaluate one or more alternatives to achieve a desired result.

For example, a company engaging in the development of a new software platform that utilizes machine learning to predict consumer behavior would likely qualify. This is because the development process would involve a systematic process of trial and error, testing various algorithms and data sets to improve prediction accuracy.

In a different scenario, a business could be going through a process of experimentation based on customer feedback, without engaging in a systematic process to resolve technological uncertainties, would not qualify. This activity, while potentially valuable for the business, does not involve the technological experimentation required by the IRS.

Experimentation is an interesting component of the 4 part test because different industries, and different activities within those industries, can have wildly different iteration cycles. There may be research and development costs going into research on cold fusion that goes decades without finding a breakthrough, but they qualify under the other parts of the test and are systematically going through a process of experimentation. In contrast, you may have a software developer trying 10 new ways to write a certain piece of code in a few hours and that would also be a process of experimentation.

It's important to differentiate between activities that genuinely seek to advance technology through a process of experimentation and those that are more routine or administrative in nature. The IRS is looking for evidence of a methodical approach to solving technological problems, not just iterative improvements based on customer preferences or market research, because remember, the overall goal is to invent new things and incentivize innovation.

4. Technological in Nature

This is one part of the test that should be fairly straightforward. You likely know you qualify for this already from a quick look at the rule.

But for more specific details, according to the IRS, for research to be considered "technological in nature," it must be based upon the principles of physical or biological sciences, engineering, or computer science. This means that the research must fundamentally rely on scientific principles to solve a problem or achieve an outcome.

For instance, a biotech firm conducting experiments to produce a new drug is engaging in activities that are technological in nature. Their work relies on biological and chemical sciences to create a new or improved product, fitting squarely within the IRS's requirements.

Similarly, an engineering company designing a new type of bridge that can withstand earthquakes better than existing designs would qualify. Their work involves applying principles of physics and engineering to solve a specific problem, which is the essence of being technological in nature.

Not to overstate the obvious here, but conversely, activities that do not qualify under this criterion include those that are not based on the hard sciences.

For example, a company conducting market research to understand consumer preferences for a new app feature is not engaging in research that is technological in nature. This activity relies on social sciences rather than the physical or biological sciences, engineering, or computer science. Similarly, a business developing a new marketing strategy does not qualify, as this process does not involve the systematic application of scientific principles to resolve a technological uncertainty.

What are some examples of qualified research activities?

We've seen a few examples in different scenarios when walking through the 4 parts above, but what are good examples of activities that typically pass the 4 part test?

Lab research
Architectural design work
Quality assurance and testing
Software development
Formula development for new foods
Manufacturing and machine improvements

While these are good examples of activities that probably qualify, every situation is different, and it's important to keep in mind that this is an activities based credit. So not every instance of these activities will necessarily qualify, and this is something you would work with a tax professional to determine.

How do you claim the credit on those activities?

Going through the process of actually claiming the credit is a different story, but when we say “software development” qualifies for the R&D credit, what does that actually mean?

It means that your company is paying for resources that go into the general category of “software development.” Those expenses are reviewed, either by software automatically or through an interview process, and the dollar amounts are counted up which serves as a basis for determining your eventual credit. For example, what are the things your company is paying for to actually do “software development”?

It could be things like these:

Wages paid to engineers, QA, and testing teams
Cloud services
Chips or servers purchased to run or train AI models
Equipment purchased for your employees that specifically go toward working on R&D

An important thing to note is that the R&D credit is an activities based credit, and similar to the 4 part test eligibility, these are not all or nothing.

So if you pay an engineer and they work 100% on qualifying R&D activities, you can claim 100% of their wages.

But if they only work on R&D activities 50% of the time, you can only claim half of their wages for the credit. This applies to all of the expenses mentioned above.

Do payments to contractors qualify for the R&D credit?

You have to be very careful with contractors and expenses related to contractors when claiming the R&D credit. In general, expenses paid to contractors (either a company or an individual freelancer) are technically eligible for the R&D credit, but there are a few things you should watch out for:

Where are the contractors doing the work located? The only work that qualifies for the credit is work that happens within the United States. The same is true for each of the individual states that offer the credit.
Who owns the rights to the end product? If you own substantial rights to the end product, you can likely claim the credit, but this also likely means that any other party cannot claim those same activities.
Who is taking the risk? Say the R&D you are paying for doesn’t end up working out. If you are the one that is taking the risk and losing money, then you are the one that can claim the credit. If you will be paid for your R&D work either way, then you might not be able to claim the credit. This is important for any industry, but often becomes an issue for architecture firms specifically.

According to Jonathan Tucker from KBKG in a 2021 interview,

“So, you gotta look into the actual contractual arrangement you have with that particular customer or client to understand two concepts in the R&D world.

One is, do you retain substantial rights to that development even though a customer is requesting it, so you’ve gotta retain substantial rights, and who bears the economic risk to that development?

The whole idea around the rights and risks is to basically only allow one company or one person to be able to claim the credit on the exact same activity. So to understand whose research it is.

Substantial rights does not mean exclusive rights. Substantial rights just means you have the right to use the technology or whatever you’ve developed in your trade or business, you use it in your trade or business. And you don’t have to pay for the right to use it, right so you don’t have to pay a royalty back to that client or whatever it is, to use it. So that’s substantial rights.

The risk of loss has to be yours as well, meaning that if it fails or you aren’t able to produce whatever it is that you are trying to develop, that you bear the risk of loss there, that you’re not gonna get compensated regardless of the outcome. Because if you are gonna get compensated regardless of the outcome, then it’s the other person’s R&D because it’s their risk.”

What documentation do you need to pass the 4 part test?

Although there is no exact list of documentation to prove your eligibility, the IRS does stipulate that you must have books and records of your activities to be able to prove they happened and that they qualify. These documents can be as general as meeting minutes, budgets, or progress reports and as specific as the actual data used in testing or blueprints used in engineering or architecture design.

What industries qualify for the R&D tax credit?

It’s important to know that any one industry is not qualified or disqualified. The credit is an activities based credit as mentioned above, which means that what passes or does not pass the 4 part test is based on

You may immediately think of scientists in white lab coats when thinking about the R&D credit, and many of those industries, like biotech, pharmaceuticals, or chemical engineering definitely produce a lot of qualifying activities.

However, a lot of companies are doing qualified research activities without knowing it in other industries. For example, things like manufacturing, software development, and food science also tend to have qualifying research activities. All of the above criteria always apply, but they may apply to things we might not normally think of.

Some industries may be more of the opposite. They seem to be innovative, but actually don’t tend to qualify for as many research activities. For example, financial service companies are often right at the front of adopting more efficient ways to do things, but creating new financial models or instruments don’t actually involve a process of experimentation and likely aren’t using hard sciences.

Similarly, in the retail industry, they frequently direct resources towards new marketing strategies, customer segmentation, sales optimization, and other heavily data driven processes that uncover new information or ideas. While these can be innovative or help a business, they do not necessarily involve the development or improvement of products, processes, software, or formulas in a way that qualifies for the R&D tax credit.

If you have further questions on what exactly qualifies, try asking our Open Advisor AI, and submit your information to get matched with the best R&D tax partner for your company.