A common question is the difference between software maintenance and qualified research for the R&D tax credit. The short answer is that routine software maintenance — applying updates, patches, and monitoring using known methods — generally is not qualified research. Maintenance that involves a technical uncertainty and a process of experimentation may warrant review. This page explains the framework in general terms. It is educational and is not individualized advice. For the foundational framework, see our page on qualified research.
Routine Software Maintenance
Routine software maintenance generally is not qualified research. The characteristics of routine maintenance include:
- Known method — the maintenance procedure is established.
- Known update — the update or patch is known.
- No technical uncertainty — there is no question about the capability, method, or design of the software.
Where all three are present, the work is applying established knowledge, not conducting research. The four-part test is generally not satisfied.
Maintenance That May Warrant Review
Software maintenance may warrant review when it involves a technical uncertainty and systematic testing of alternatives:
- Technical uncertainty — a question about whether a maintenance change can improve the software to a new or better performance.
- Alternatives — alternative approaches are evaluated.
- Systematic testing — the alternatives are tested through a structured process.
Where the maintenance involves evaluating alternatives to resolve a technical uncertainty, the work may warrant review as qualified research.
The Distinction Is in the Uncertainty and Experimentation
The distinction between routine maintenance and qualified research is in the uncertainty and experimentation:
- Routine maintenance — known method, known update, no technical uncertainty. Not qualified research.
- Maintenance as research — technical uncertainty, alternatives evaluated, systematic testing. May warrant review.
The same physical activity — updating or patching software — can be maintenance or research depending on whether there is a genuine technical uncertainty and a process of experimentation.
Hypothetical Example
Consider a company that performs routine updates and patches on its software using established methods. This is routine maintenance, not research.
By contrast, if the company evaluates alternative approaches to resolve a technical uncertainty about whether a new maintenance approach can improve software reliability, and tests the alternatives through a structured process, the work may warrant review as qualified research.
These examples are illustrative only and do not state whether any particular activity qualifies.
Documentation That May Help
Records that can help support the maintenance vs. research analysis include records showing whether the work was routine maintenance or research, records of the technical uncertainty (if any), and records of the process of experimentation (if any). For more, see our page on R&D tax credit documentation.
Key Takeaway
Routine software maintenance — applying updates, patches, and monitoring using known methods — generally is not qualified research. Maintenance that involves a technical uncertainty and a process of experimentation may warrant review. Because the distinction is fact-specific, professional review is appropriate before claiming the credit.