Adapted, Duplicated, or Routine? The Exclusions Engineering Firms Should Know Before Claiming the R&D Credit

TL;DR

• The federal R&D credit is built around a four-part definition of qualified research, and Section 41(d)(4) lists activities that are carved out even when real engineering skill is involved.
• For engineering firms, the exclusions that come up most are adaptation to a customer’s need, duplication of an existing design, routine testing and inspection, efficiency or management studies, and work done after a design is ready for use.
• Knowing where the line sits does not shrink a claim. It helps a firm describe the work that does qualify more precisely, and that can make the whole claim easier to support.

What the credit is, in a sentence

The federal research credit under Internal Revenue Code Section 41 can reward businesses for spending on qualified research, which the statute defines as technological work intended to discover information useful in developing a new or improved business component, where substantially all of the activities are elements of a process of experimentation. You can read the statute at Cornell’s Legal Information Institute and the implementing regulation at eCFR, 26 CFR 1.41-4.

That definition is the “yes” side. This post is about the “no” side: the list in Section 41(d)(4), which says qualified research “shall not include” certain activities. Engineering firms tend to meet this list more often than most businesses, because so much of the day-to-day work is close to the line.

Exclusion one: adapting an existing design to a customer’s need

The statute excludes research related to adapting an existing business component to a particular customer’s requirement or need. In engineering terms, think of taking a proven structural, mechanical, or electrical design and tailoring it to a new site, a new client specification, or a new set of dimensions.

Two details are worth knowing. First, the regulation (1.41-4(c)(3)) says that a project being intended for one specific customer does not, by itself, trigger the exclusion. A one-off design for a single client can still qualify if the work involves real technical uncertainty and experimentation. Second, the exclusion is about adapting something that already exists. If the team could not know at the outset whether any design would meet the performance requirement, and had to evaluate alternatives to find out, that is a different fact pattern than adjusting known values.

Exclusion two: duplicating what already exists

Research related to reproducing an existing business component is excluded when it comes from a physical examination of the component, or from plans, blueprints, detailed specifications, or publicly available information. For a firm that works from standards, reference designs, or a client’s legacy drawings, this is a useful gut check.

The regulation (1.41-4(c)(4)) also notes that examining an existing component while developing your own does not trigger the exclusion on its own. Studying a predecessor design is not the problem. Reproducing it is.

Exclusion three: surveys, studies, and routine activities

Section 41(d)(4)(D) and 1.41-4(c)(5) list activities that are excluded even when they take skill and time: efficiency surveys, management functions or techniques, market research, routine data collection, and routine or ordinary quality-control testing or inspection.

For an engineering firm, the practical sorting question is often whether a test was run to find out something the team did not know, or to confirm that a result met a known standard. Testing that resolves a real design uncertainty is part of how a process of experimentation can look. Routine inspection against a checklist generally is not. Internal workflow changes, such as reorganizing how a department staffs projects, fall on the management-technique side of the line.

Exclusion four: work after the design is ready for use

Research conducted after the beginning of commercial production of the business component is excluded. The regulation (1.41-4(c)(2)) says production begins when the component is ready for commercial sale or use, or meets the taxpayer’s basic functional and economic requirements. It lists examples of activities treated as happening after that point, including preproduction planning, tooling-up, trial production runs, troubleshooting production equipment or process faults, accumulating production-process data, and debugging flaws.

For a firm whose deliverable is a design, a plan set, or a system specification, the useful question is when that deliverable reached the point where it met its basic requirements. Effort before that point may be part of the analysis. Routine revisions, field fixes, and punch-list items afterward are a different story, depending on the specific activities and facts.

Other exclusions worth a quick look

Several other items on the Section 41(d)(4) list matter at the margins for engineering firms. Research on software developed mainly for the firm’s own internal use faces a higher bar. Work performed outside the United States is excluded. Research to the extent another person or governmental entity funds it is excluded, which makes client contract terms important. Finally, Section 41(d)(3) says research relating to style, taste, cosmetic, or seasonal design factors does not count, which is relevant for work with an aesthetic component.

Reading the exclusions the right way

It is tempting to read this list as a reason to be pessimistic. In practice it works more like a map. The regulation also makes clear (1.41-4(a)(3)(ii)) that research does not need to go beyond the common knowledge of skilled professionals in the field, and that the taxpayer does not have to succeed in developing a new or improved component. A project can fall short of its goal and still contain qualifying work, and a project can succeed and contain a lot of work that does not qualify.

That is why a careful claim usually separates a project into its parts. The uncertain, experimental portion is described on its own terms, and the routine, adapted, or post-release portions are set aside. Claims built that way tend to be easier to explain and easier to support.

What this means for your firm

If you are an owner or CFO, a few questions can help you see where your projects land:

1. Which of our projects started from an existing design, and which started from a question we could not answer on day one?
2. Where did our testing resolve something unknown, and where did it confirm something expected?
3. When did each deliverable reach the point where it met its core requirements?
4. Which client contracts might change who is treated as funding the work?

You do not need perfect answers to start. A specialist can walk through these with you, and Strata offers a $0 initial assessment to size the opportunity. For firms that move forward, a study typically takes about 4 to 8 weeks depending on the specific facts and how quickly information is available, and most of the work sits with the provider rather than your team. You receive the finished package and pass it to your CPA, who files the return.

Interested in a quick read on where your projects might land? Book a 15-minute review at stratataxgroup.com/contact.

Sources

• 26 U.S.C. 41, Credit for increasing research activities (Cornell LII)
• 26 CFR 1.41-4, Qualified research for expenditures paid or incurred in taxable years ending on or after December 31, 2003 (eCFR)
• About Form 6765, Credit for Increasing Research Activities (IRS.gov)

This post is for informational purposes only and does not constitute tax or legal advice. Consult a qualified tax professional regarding your specific circumstances.