NSF's America's Seed Fund funds research, not product development. Learn the research-vs-development distinction and why technical risk is what wins NSF SBIR/STTR awards.

The most common reason engineers and founders lose an NSF SBIR/STTR proposal isn’t a weak product or a small market. It’s that they proposed development when NSF funds research.

This is a subtle but decisive distinction, and the National Science Foundation explains it directly in its applicant guide “Research 101 for Engineers” (NSF, America’s Seed Fund). The guide exists because so many technically brilliant applicants describe building a product when NSF is asking them to answer a research question. Understanding the difference is the single highest-leverage thing you can do before applying.

This post unpacks that distinction, what “technical risk” means to NSF, and how to frame an SBIR/STTR proposal so reviewers see research, not engineering.

Source: This post is based on and cites the National Science Foundation’s applicant resource Research 101 for Engineers (NSF SBIR/STTR Program — America’s Seed Fund). All interpretation is our own; consult the original document and current NSF SBIR/STTR solicitations before applying.

TL;DR: Quick Answers

What does NSF SBIR/STTR fund? Research with commercial potential, work that reduces a genuine technical risk, not product development or market validation.
What’s the research vs. development distinction? Research answers an open question whose outcome is genuinely uncertain (“will this work?”). Development applies known principles to build something.
What is “technical risk”? The risk that the core innovation might not work for fundamental scientific or technical reasons. NSF funds projects that confront and reduce this risk.
How is technical risk different from market risk? Market risk is “will customers buy it?” NSF Phase I centers on technical risk (“can it be made to work at all?”), with commercialization assessed alongside.

Why This Distinction Decides Awards

NSF SBIR/STTR, branded America’s Seed Fund, provides non-dilutive funding to small businesses and startups for high-risk, high-reward R&D with commercial potential. It is part of the broader federal SBIR/STTR system and sits alongside NSF’s other research funding.

As NSF’s Research 101 for Engineers makes clear, the program funds research, the pursuit of new knowledge under genuine uncertainty, not the engineering work of building a product from known principles. Reviewers are explicitly evaluating whether a proposal tackles a real technical unknown and offers a credible path to resolving it.

When an engineer writes “we will build a device that does X,” they’ve described development. When they write “it is currently unknown whether X can be achieved because of [specific technical barrier]; we will investigate whether [approach] can overcome it,” they’ve described research. The first loses; the second is fundable.

Research vs. Development: The Core Framework

Drawing on NSF’s framing in Research 101 for Engineers:

Research advances scientific or technical knowledge. Its defining feature is uncertainty of outcome. Before the work, you genuinely don’t know whether the approach will succeed, and answering that question produces new knowledge. Research confronts technical risk.

Development applies existing, established knowledge to design, build, or refine a product or process. The principles are known; the work is execution and engineering. Development carries execution risk, but not fundamental technical risk.

A useful test NSF effectively poses: Could the project fail for a fundamental technical reason that no one currently knows the answer to? If yes, it’s research. If the only risk is “it will take effort and money to build,” it’s development, and NSF SBIR/STTR is the wrong fit.

Engineers are especially prone to this trap because their training rewards building working systems from known principles, exactly the opposite of what a research funder wants to see. Research 101 for Engineers exists precisely to retrain that instinct for the SBIR context.

What “Technical Risk” Means to NSF

Technical risk is the heart of an NSF SBIR/STTR proposal. It is the risk that the central innovation might not work because of a scientific or technical barrier whose resolution is genuinely unknown.

Strong proposals:

Name the specific technical risk. Not “it’s hard to build,” but “it is unknown whether [material/method/algorithm] can achieve [measurable threshold] under [conditions].”
Explain why it’s an open question. What about current science or engineering leaves the answer genuinely uncertain?
Propose a credible path to reduce it. A research plan whose results would meaningfully resolve the uncertainty.
Define success quantitatively. What measurable result would demonstrate technical feasibility?

This mirrors the broader logic of evaluation plans and SMART goals: name the unknown, define how you’ll measure resolution, and tie the work to it.

Technical Risk vs. Market Risk

A second distinction Research 101 for Engineers reinforces: NSF separates technical risk from market risk.

Technical risk: Can the innovation be made to work at all? (NSF Phase I’s central concern.)
Market risk: Will customers buy it, at scale, profitably? (Important to NSF’s commercialization assessment, but not the core of the research case.)

NSF SBIR/STTR is not venture capital, and it is not a product-development grant. It funds research that reduces technical risk and has plausible commercial potential. A proposal that’s all market opportunity with no genuine technical unknown will fail the research test; a proposal with deep technical risk but no commercial path will fail the commercialization test. You need both, but the research/technical-risk case is what distinguishes SBIR from ordinary startup funding.

How NSF SBIR/STTR Is Structured

For context (verify current details on the NSF SBIR/STTR site, as solicitations evolve):

Project Pitch: NSF uses a short pitch step to screen for fit before inviting a full proposal, an efficient way to confirm your project is research with technical risk and commercial potential before investing in a full application.
Phase I: Establishes technical feasibility, proof of concept that reduces the central technical risk.
Phase II: Funds further R&D building on Phase I results, moving toward commercialization.
Phase IIB / supplements: Additional support tied to matching investment, advancing commercialization.

Throughout, the question NSF keeps asking is the one from Research 101 for Engineers: is this genuine research that reduces technical risk, or is it development dressed up in research language?

How to Reframe a “Development” Idea as Research

If your instinct is to describe building a product, reframe before you write:

Find the unknown. What part of your innovation has no settled answer in current science or engineering? That’s your research core.
State it as a question. “Can [approach] achieve [threshold] given [barrier]?”
Explain the risk honestly. Why might it fail? Reviewers trust applicants who name real risk, the same principle as a risk and mitigation section.
Design research, not a build. Experiments, models, prototypes-as-tests, whose results answer the question, rather than deliverables that assume success.
Keep commercialization in view. Show the market potential, but lead the technical narrative with the research, see how to read a grant RFP for matching each funder’s actual evaluation criteria.

How This Fits the Broader Funding Picture

NSF SBIR/STTR is one doorway in the federal research funding and SBA-coordinated SBIR/STTR landscape. Other agencies (NIH, DOE, DOD, NASA, USDA) run their own SBIR/STTR programs with different emphases, see our federal vs. foundation vs. state grants overview and grants for research labs. The research-vs-development discipline NSF teaches applies, with variations, across all of them.

How Grantboost Helps

NSF SBIR/STTR proposals demand a sharp research framing, and that framing is irreducibly the applicant’s expertise. What Grantboost helps with is everything around it: surfacing relevant SBIR/STTR and complementary opportunities through continuous discovery, drafting the organizational and commercialization sections in your venture’s voice (see training AI on your past proposals), and keeping the multi-phase timeline organized in one pipeline.

Try Grantboost free and spend your time on the research case, not the paperwork.

NSF SBIR/STTR: Research vs. Development and Why Technical Risk Wins Funding