Customer funding can exclude otherwise technical work
The research credit excludes research to the extent funded by another person, including a governmental entity. Contract analysis examines whether payment is contingent on successful research and whether the taxpayer retains substantial rights in the research. A developer cannot establish credit eligibility merely by showing that its engineers solved difficult technical problems. Who bears the economic risk and who holds the research rights can change the result.
The funded-research analysis is distinct from whether the activity meets the general qualified-research tests. Both must be satisfied. A fixed-price contract can create risk for the developer, but its label does not resolve every payment or rights issue. A cost-reimbursement arrangement can shift risk in a different way. Review the actual terms, modifications, acceptance procedures, termination provisions, and intellectual-property rights. Government work is not automatically excluded in every circumstance simply because a public customer is involved; the extent of funding and the applicable rules must be evaluated. Likewise, a commercial customer does not remove the funded-research restriction. The practical starting point is the complete executed contract, not the marketing description of the project.
Separate two hypothetical project pools
Assume a developer incurs $300,000 on Project A and $200,000 on Project B. Both involve domestic technical experimentation, but a reviewed contract analysis concludes that Project A is fully funded for research-credit purposes while Project B is not. Assume the developer otherwise identifies $160,000 of qualifying in-house expenses within Project B. The starting qualified-expense pool is $160,000, not the full $500,000 spent on both projects.
If a preliminary sales estimate simply applied a hypothetical effective credit percentage of six percent to total spending, it would suggest $30,000. Applying that illustrative percentage to the supported $160,000 pool would instead suggest $9,600, before base calculations, section 280C coordination, and credit limitations. The six percent rate is only a planning assumption and is not the statutory formula. Actual credit computation depends on the selected method and historical information. This example demonstrates the size of a classification error rather than calculating a return-ready credit. The excluded spending can still have a separate deduction or capitalization treatment; losing research-credit eligibility does not itself establish that the operating cost is nondeductible.
Read acceptance and intellectual-property clauses together
Create a contract matrix showing the customer, project, payment milestones, acceptance conditions, refund obligations, rights retained, and rights transferred. Link amendments and statements of work to the original agreement. A change order may shift risk or rights midway through a project, so the tax conclusion may need to distinguish phases instead of assigning one label to all annual spending.
Ask project managers how the contract operates in practice. Determine whether the customer pays for hours regardless of technical success, whether failed work must be redone without additional payment, and what happens if the project ends early. Preserve contemporaneous correspondence about acceptance and disputed invoices. For research rights, identify whether the developer can use the resulting technical information in its business and whether payments are required to exercise those rights. The legal analysis should address substantial rights under the governing rules rather than equating possession of source files with ownership. Keep the contract assessment beside the expense schedule, because payroll records alone cannot support the funding conclusion. Reconcile excluded projects to the ledger so reviewers can see that the company screened the entire research population.
Change the bid model before the next contract
The developer's expected credit should be based on the rights and risks it will actually accept. If management prices a project assuming a research credit and later grants the customer terms that defeat eligibility, the bid margin can be overstated. Have finance review the credit assumptions when contract terms materially change, alongside delivery costs and liability exposure.
Do not rewrite a completed contract's economic history to match a desired tax outcome. Future agreements can be negotiated deliberately, but the credit analysis for past work must follow the executed terms and facts. Maintain separate schedules for project spending, credit-qualified expenses, and the amount of credit currently usable. A supported credit can still become a carryforward rather than immediate cash. Document unsuccessful research as carefully as successful work where it otherwise qualifies, while applying the funding limitation consistently. The useful business outcome is a credible project-margin forecast and a claim supported by contractual evidence, technical records, and cost records that tell the same story.
Sources & further reading
Consult the original guidance and its effective dates. These references support the educational discussion; their application depends on the specific facts.