The process-of-experimentation element is one of the four parts of the qualified-research test under Section 41 of the Internal Revenue Code. It requires that substantially all of the activities constitute elements of a process of experimentation that relates to a qualified purpose. This page explains the element and addresses a common misunderstanding: that informal trial and error is enough. Like the other elements, it must be satisfied together with them. For the overall framework, see our page on the four-part test.
Process of Experimentation in Plain English
Under the Treasury Regulations (§1.41-4), a process of experimentation is a process designed to evaluate one or more alternatives to achieve a result where the capability, the method of achieving that result, or the appropriate design of that result is uncertain as of the beginning of the taxpayer's research activities. The regulations describe the process as involving several steps and require that it be an evaluative process. In plain terms, the element looks for a structured process of evaluating alternatives to resolve a technical uncertainty — not merely ad hoc troubleshooting.
Identifying the Uncertainty
The regulations describe the first component of a process of experimentation as the identification of uncertainty concerning the development or improvement of a business component. This connects to the elimination of uncertainty element: the process begins by identifying the technical question about capability, method, or appropriate design that the research is intended to resolve.
Identifying Alternatives
The regulations next describe the identification of one or more alternatives intended to eliminate that uncertainty. A process of experimentation involves considering alternatives — different possible designs, methods, materials, configurations, or approaches — that could resolve the uncertainty. The regulations state that a process of experimentation must be an evaluative process and generally should be capable of evaluating more than one alternative. This is one reason that informal tinkering with a single approach generally does not, by itself, satisfy the element.
Evaluating Alternatives
The regulations then describe the identification and conduct of a process of evaluating the alternatives. The evaluation is the substance of the experimentation: testing, comparing, or otherwise assessing the alternatives against the result the taxpayer is trying to achieve. The evaluation must fundamentally rely on principles of the physical or biological sciences, engineering, or computer science — connecting this element to the technological in nature element.
Modeling and Simulation
The regulations expressly identify modeling and simulation as examples of evaluative processes that can constitute a process of experimentation. Where a taxpayer models or simulates alternatives to evaluate them against a result — for example, simulating the performance of alternative designs — that may be the kind of evaluative process the element addresses, provided the other requirements are met. Modeling and simulation are not required; they are examples of how the evaluation can be conducted.
Systematic Trial and Error
The regulations also identify a "systematic trial and error methodology" as an example. The key word is "systematic." The element is not satisfied by random, ad hoc, or unstructured trial and error. A systematic methodology — one that identifies the uncertainty, identifies alternatives, and evaluates them in a structured way — may constitute a process of experimentation. The distinction between systematic trial and error and informal tinkering is central to this element.
What Does Not Automatically Constitute Experimentation
Several things do not, by themselves, constitute a process of experimentation:
- informal troubleshooting or ad hoc debugging;
- ordinary iteration without an identified uncertainty or alternatives;
- routine testing or quality control that follows established procedures;
- production activities, even if problems are encountered and solved on the line;
- work that evaluates only a single predetermined approach without considering alternatives.
The element looks for an evaluative process capable of evaluating alternatives, fundamentally relying on the hard sciences, engineering, or computer science.
Failed Tests and Iteration
Failed tests and iteration can be consistent with a process of experimentation. The regulations do not require the taxpayer to succeed, and experimentation often involves tests that do not produce the desired result. What matters is whether the failures are part of an evaluative process of alternatives — for example, where a failed test of one alternative informs the evaluation of another. Iteration that is part of a structured evaluative process can support the element; iteration that is merely repeated informal attempts generally does not.
Relationship to the Business Component
The process of experimentation must relate to a qualified purpose — the development or improvement of a business component. This connects to the permitted purpose element. The experimentation must be directed at resolving a technical uncertainty about a business component, not at a general inquiry. The substantially-all requirement is applied at the level of the business component: substantially all of the activities for that business component must constitute elements of a process of experimentation.
The "Substantially All" Formulation
Section 41 and the regulations require that substantially all of the activities constitute elements of a process of experimentation. The cited authority uses the phrase "substantially all" without stating a fixed percentage. The question is whether, on the whole, the activities for the business component constitute an evaluative process of alternatives rather than incidental experimentation mixed with non-qualifying work. Because the authority does not state a specific percentage, this page does not state one; the determination is facts-and-circumstances.
Documentation That May Help
Records describing the uncertainty identified, the alternatives considered, the evaluative process used, and the results of the evaluation can help support this element. Experiment logs, test plans, records of alternatives considered, modeling or simulation inputs and outputs, and records of how results informed next steps tend to be useful. For more, see our page on R&D tax credit documentation.
Example Fact Patterns
The following are general examples that may warrant review; none automatically qualifies, and each depends on whether all four elements are satisfied:
- A team that identifies a design uncertainty, models several alternative designs, simulates their performance, and selects among them based on the results.
- A team that identifies a method uncertainty, conducts systematic tests of alternative approaches, and evaluates the results against a target.
- A team that identifies an appropriate-design uncertainty and evaluates alternative configurations through structured testing.
In each case, the question is whether substantially all of the activities constitute an evaluative process of alternatives fundamentally relying on the hard sciences, engineering, or computer science.
Key Takeaway
The process-of-experimentation element requires that substantially all of the activities constitute an evaluative process designed to evaluate one or more alternatives to resolve a technical uncertainty about a business component. Informal tinkering, ad hoc troubleshooting, or ordinary iteration is not automatically enough. The element is one of four that must be satisfied together, and the analysis turns on the specific facts. For the broader context, see our pages on qualified research and what the R&D tax credit is. Because these determinations are fact-specific, professional review is appropriate before claiming the credit.