How to buy research and technology development
In research the deliverable is knowledge, so the contract has to say who owns it. Settle inventions, data and publication before the kickoff meeting.
This category covers applied research and technology maturation: laboratories, small research firms and university-affiliated groups that take a concept from an idea to something a programme can use. Design and stress engineering is a separate guide, and flight test is another. The buyer's problem here is that the deliverable is knowledge, which is harder to specify, easier to dispute and more valuable later than either side expects at signature.
Ownership is decided by the funding, so read it early
Where federal funding is involved, the rules on rights to inventions made by nonprofit organisations and small business firms set the framework, including the standard patent rights clauses, the treatment of contractors who take supplemental funding from other sources for closely related work, and the government's march-in rights.1 On the data side, rights in technical data are defined by clause: unlimited rights, government purpose rights for five years or another negotiated period where funding was mixed, and limited rights that keep data inside the government and bar its use for manufacture without written permission.2 Ask who will own inventions, who will own the data, and what licence each party gets. Ask what happens if the technology is later commercialised, and write that down while everyone is still friendly.
Specify the result, not the effort
Research contracts fail on vagueness more than on science. Define the question being answered, the technology readiness level at entry and the one claimed at exit, and the evidence that will demonstrate it. Define the test article, who builds it, who owns it afterwards, and where it is stored. Define the reporting cadence, the format, and what a negative result looks like, because a lab that can only report success is a lab that will spend your money proving nothing. Ask what has to be true for the work to stop early, and agree the off-ramp before you start.
Know what the small firm's business model is
Many suppliers in this category are funded through the small business innovation programmes, which are federal seed funding for research and development by small firms.3 That is not a criticism, it is a fact with consequences: the firm's roadmap may be shaped by award cycles, its staff may be committed across several awards, and the technology you are funding may already carry rights from earlier work. Ask which prior awards touch the technology you are buying, and what rights attach to them. Federal award data is public and shows what a firm has been contracted to do, for how much and when, which is the cheapest diligence available on a research supplier.4 It shows funding history and scale. It does not show whether the work succeeded, and the final reports it references are often the only place that answer exists.
Questions that sort labs
Ask who personally will do the work and what else they are committed to. Ask for two prior projects at the readiness level you need, and what happened next. Ask what equipment is in house and what is booked at somebody else's facility. Ask what the publication expectation is, and whether you can defer it.
Sources, in the order cited
- 1Electronic Code of Federal Regulations · 37 CFR Part 401, Rights to Inventions Made by Nonprofit Organizations and Small Business Firms
- 2Acquisition.gov · DFARS 252.227-7013, Rights in Technical Data, Noncommercial Items
- 3US Small Business Administration · SBIR and STTR, America's Seed Fund
- 4United States Department of the Treasury · The USAspending API, federal award data