By Jamie Overberg, featured in Accounting Today

When it comes to research and development tax credits, certain costs related to wages, supplies and contract research are eligible. Often, though, taxpayers – and more specifically, the engineers doing the work – overlook various processes and routine work that can be tapped for extra tax savings.

R&D credits aren’t limited to what’s never been done before, such groundbreaking inventions as the lightbulb and the first cell phone. There is no IRS requirement that technology be revolutionary or meet a first-to-market threshold of innovation to be eligible for credit, only that products and processes be improved.

Yet, each year, corporations and pass-throughs across industries fail to claim R&D tax credits because they aren’t aware that many day-to-day activities qualify for this dollar-for-dollar reduction in income tax liability. Here are some of the most overlooked qualifiers we’ve seen in recent years.

Cloud computing network costs

Cloud computing server, platform and SaaS software application innovation costs may be qualified research expenses (QREs) eligible for federal and state R&D credits. It is now cheaper and easier to innovate by leasing cloud computing time than purchasing servers and outfitting data centers. Pooled computing power is driving innovation and may be eligible when the cloud is operated by someone other than the taxpayer and located off the taxpayer’s premises, and the taxpayer is not the primary user of the computer. Cloud-hosted development platforms and beta testing of pre-released software programs are included; operating platforms are not. Although an easy grab, these network costs are often left out of R&D calculations.

System Migration

As companies migrate their platform or legacy systems, such as during the merging of two companies, the development efforts related to the move may qualify. Migrations, historically considered not qualified, are rarely just “lifts and shifts.” Typically cloud based, the strategy of moving workloads and tasks from one system to another almost always involves design, process improvements, technical uncertainty and points of failure, the baseline criteria for federal research credits.

Replacing an obsolete part or product, or a piece in a larger assembly

This commonly occurs when a supplier stops making a piece that is part of a manufacturing process or product. If a supplier for a filter piece used in medical equipment stops selling the piece, and a replacement component or filter cannot be found, a taxpayer might change their product to accommodate a new filter on the market, requiring design changes related to the size, functionality, geometry and other elements.

When the obsolete part is so old, new technology may be better than a retrofit. Testing must be done to replace a component piece with new technology, such as an air fork in a previously oil-based strut that is no longer available. Taxpayers who can provide proof of concept, demonstrate how the innovation made the product better, and document the failure points may be able to recoup expenses related to design and testing. Expense eligibility can extend beyond engineers to the wages associated with a marketing associate who provided initial design input on the enhancements or features that should be created. However, the taxpayer must clearly demonstrate the individual’s contributions to the process and that they had the technical expertise needed to participate in the design.

Automation that improves efficiencies

Process automation tools, such as automated shelving, robotic arms and labeling systems, are improving workflows and manufacturing processes. When adding robots to workflows, there is uncertainty in where to place the robot, whether the robot meets quality specifications, and what the robot will do, accomplish or complete. While taxpayers cannot include the cost of the robot, the time invested in designing the type of robot (static or dynamic), how and where that robot is used, and the trial runs used in testing the product or process may be qualified expenses.

Machine learning and artificial intelligence

Any product or process using machine learning and artificial intelligence to learn how to do something better, faster and more efficiently qualifies for research credits. Machine learning may help a company gain the requisite understanding of where to place a robotic arm for the highest and best use.

These are just a few of the more exotic qualified expenses you may be overlooking in your manufacturing products and processes. Activities do not have to fit the typical, textbook definition of research.

Jamie Overberg, a partner at TaxOps Minimization, has been executing and managing all aspects of R&D credits for over 20 years at innovative companies nationwide. Ms. Overberg links nexus to qualified expenses by business component or project accounting, delivering the highest standard in audit-ready cost capture. She can be reached at

Let’s talk tax

Jamie Overberg can be reached at or 720.227.0421.

More Tax News

Follow Us