Making the Question Visible. A copper-colored airfoil meets delicate lines of flow in an imagined wind tunnel. Original editorial painting about testing an idea against the conditions it must survive.
Match technical uncertainty to public research needs and a practical route to use
By Arbab Naseebullah Kasi, Chief Executive Officer - CEO, Feel Worldwide Foundation Inc..
A founder can spend weeks searching for funding before asking the question that would make the search useful: what uncertainty would this money allow us to resolve?
The answer may be more demanding than finishing a prototype. Perhaps a new coating works in a clean laboratory but loses its properties after repeated use. Perhaps a sensor can detect a signal, but nobody knows whether it can distinguish that signal from the background under realistic conditions. Perhaps an experimental platform could make a new kind of research possible, if its measurements can be trusted.
Those are research questions. They identify something important that remains unknown and suggest how evidence could change a decision.
For a technology venture seeking public research funding, this is a productive place to begin. Establish the problem, explain why existing approaches leave it unresolved, and show how a disciplined investigation could create something useful. Then look for a program whose purpose matches that work.
Separate research support from general startup finance
The U.S. Small Business Administration explicitly says it does not provide grants simply to start or expand a business. Its grants guidance, checked on November 9, 2026, distinguishes that general need from federal support for scientific research and development through the Small Business Innovation Research and Small Business Technology Transfer programs, commonly called SBIR and STTR.
A shop needing inventory, a consultancy needing a website, or an established product needing a marketing campaign may have a sound business case. Those needs do not become competitive R&D merely because the business uses technology. The applicant must identify the research itself and explain why the proposed work belongs within the specific funding opportunity.
Research support also brings obligations. An award supports the approved work under its terms. It should not be treated as unrestricted cash for everything a young company wants to do. Build a separate view of the venture's operating needs, the proposed research costs and the resources required after the research ends.
That separation protects the business as well as the funder. It prevents a founder from constructing a survival plan around an award that is uncertain, restricted and potentially slower than the company's next bill.
Check the program and the particular opening
Funding advice needs a date. As of November 9, 2026, the SBIR and STTR programs have been reauthorized through September 30, 2031. Public Law 119-83, approved April 13, 2026, extended the authorities following the previous September 30, 2025 expiration. An older notice describing the lapse should not be presented as the current nationwide position.
Authorization, however, does not establish that a particular agency is accepting a particular proposal today.
For example, NSF solicitation 26-510, posted May 22, 2026, was marked active when checked. Its November 4, 2026 deadline had passed. The next listed full-proposal deadline was March 4, 2027, at 5 p.m. in the submitting organization's local time. Phase I applicants need an official Project Pitch invitation before submitting a full proposal; the invitation's validity and submission limits also matter.
By contrast, NASA's Phase I page reported that its 2026 Appendix A SBIR and Appendix B SBIR and STTR solicitations closed on May 21, 2026. That statement concerns those appendices, not every possible NASA opportunity. NASA directs applicants to its Broad Agency Announcement information hub for other releases.
Before committing proposal time, record the opportunity number, current status, amendments, deadline, eligible phase, submission route and any required preliminary step. Recheck them before submission. A search result, an old webinar or a promising agency homepage is not enough to establish that a call is available to your company.
Read the agency need before describing your invention
Begin with the problem the agency is trying to address. Some opportunities identify tightly defined technical needs. Others support a broader research and commercialization mission. Read the actual language rather than forcing every program into the same template.
NSF's current program overview explains that its SBIR/STTR programs support high-risk technologies across many fields and do not procure goods and services from applicant companies. That distinction matters: the research funder is not necessarily the eventual customer.
A useful fit statement connects three things. It identifies the stated public research purpose, the unresolved technical barrier, and the proposed advance that would help address it. Each connection should be specific enough that another person can challenge it.
If your statement could describe a hundred unrelated startups, keep working. If it depends on changing the meaning of a solicitation, look elsewhere. A responsible decision not to apply can save more time than an impressive but poorly matched proposal.
Speak with the designated program contact when the rules allow and a real question remains. Prepare a focused question about fit or requirements. A helpful conversation can clarify an opportunity, but it cannot replace a formal eligibility determination or promise a favorable review.
An illustrative venture turns a problem into a test
Consider a hypothetical company developing an optical instrument for researchers studying short-lived chemical reactions. This company, its interviews, targets and test plan are invented for illustration. They are not an actual applicant, an award recipient or a Feel Worldwide Foundation Inc. project.
Its early idea is to build a smaller instrument. That is a product description. The more useful problem emerges when prospective users explain that their current setup cannot reliably separate two rapidly changing signals in the same experiment. The missing capability limits the questions they can investigate.
The company identifies a plausible funding connection in NSF 26-511, the scientific instrumentation pilot posted May 22, 2026. The solicitation seeks enabling instruments and experimental platforms that advance scientific discovery. It was marked active on November 9, 2026, with March 4, 2027 as its next listed full-proposal deadline. This is a possible fit to investigate, not an agency endorsement of the fictional project.
The company frames its research question: can a new optical arrangement distinguish the two signals at a defined sampling rate while staying within an acceptable measurement error under changing laboratory temperatures?
The proposed advance is the measurement capability. Its smaller footprint may help adoption, but convenience alone does not establish the research contribution. The proposal must explain why the new arrangement could work, what existing methods cannot yet deliver and what remains technically uncertain.
For a real application, the team would support that account with literature, relevant prior work and appropriately described preliminary evidence. A patent search and a scientific literature review answer different questions. Neither, by itself, proves commercial freedom to operate or that the experiment will succeed.
Make milestones answer a decision
A milestone should tell the team what it has learned and what happens next. Completing a meeting, buying equipment or assembling a device may be necessary work, but those activities do not demonstrate that the central uncertainty has been resolved.
In the hypothetical instrument project, the team proposes three stages. First, establish a reference measurement and a repeatable test procedure. Second, test the new arrangement against that reference across prespecified operating conditions. Third, have a qualified independent laboratory repeat the key comparison using the agreed protocol.
Suppose the illustrative target is measurement error no greater than 5% across a defined signal range in at least 27 of 30 prespecified test runs. The proposal must explain the reference, the error calculation, how runs are selected and whether repeated observations are genuinely independent. Thirty convenient measurements from one stable setup would not automatically justify conclusions about many laboratories.
Those invented numbers are not an NSF threshold or a recommended design for this field. A real team would justify the sample size, uncertainty analysis and acceptance criteria with relevant technical expertise and user requirements. The value of the example is its structure: a condition, a measurement, a threshold and a decision.
Agree in advance what a shortfall means. An instrument that meets the target only within a narrow temperature range may justify a revised application, further research or a stop decision. Reporting that boundary honestly can be more valuable than celebrating an average that hides failure in the conditions users actually face.
Each milestone should connect to an owner, time allowance, necessary equipment and budget. Include calibration, access to facilities, reproducibility and foreseeable delays. A credible plan shows how the team will learn from an unfavorable result without rewriting the original target after seeing the data.
Confirm who can apply and who will do the work
Eligibility should be checked before drafting the full narrative. NSF's current eligibility guidance calls for a qualifying U.S. small business with fewer than 500 employees including affiliates, qualifying majority U.S. citizen or permanent-resident ownership and control, and U.S.-performed R&D. It excludes majority ownership by venture capital operating companies, hedge funds or private equity firms. These are NSF-specific applicant instructions, not a substitute for the governing rules or advice on a complicated ownership structure.
For NSF projects, the principal investigator must primarily work for the company during the award and have the legal right to work in the United States. A valid full proposal also needs active SAM, Research.gov and SBA Company Registry registrations. STTR requires an eligible nonprofit research institution partner and, under NSF's budget rules, at least 40% allocated to the company and at least 30% to that institution. SBIR Phase I generally allocates at least two-thirds to the company.
Do not assume a university collaborator makes the university the applicant. The SBIR application guidance states that nonprofits cannot directly receive an SBIR or STTR award, although eligible research institutions can participate in the specified partner role. A nonprofit interested in research collaboration must establish its actual eligibility for that role rather than infer it from its charitable status.
Check ownership, affiliates, personnel commitments, research-security disclosures, institutional agreements and intellectual-property arrangements with appropriate advisers and the agency. Forming a U.S. corporation does not by itself resolve those requirements. Nor does technical brilliance compensate for a structure that fails the applicable rules.
Write for the criteria reviewers actually use
The NSF review criteria linked from its current solicitation cover Intellectual Merit, Broader Impacts and Commercial Potential. They examine originality, a reasoned plan with a way to assess success, team qualifications and available resources. The commercial questions include the market opportunity, a durable innovation-based advantage, the business model, the team's commercial capability and how NSF support could improve the venture's prospects. Phase II review also asks whether Phase I established a sound foundation.
Use those criteria to test the proposal, not as decorative headings added after it is written. For the hypothetical instrument, explain the new knowledge the experiments could generate, the research capability that successful development could enable and the market in which the company might sustain that capability.
Keep evidence proportional to the claim. A laboratory demonstration supports a technical observation under stated conditions. An interview supports an account of a potential user's needs. A letter expressing interest supports that interest, within its wording. None automatically proves purchase intent, a successful procurement process or broad demand.
Other agencies and phases can use different criteria, priorities and selection procedures. Read their current requirements directly. A proposal that scores well against your own checklist has no guaranteed standing in an actual competition.
Commercialization belongs in the research conversation
Ask who would use the result, who would pay, who would approve the purchase and what would make adoption practical. In a laboratory, those may be different people. The scientist may value measurement performance while a purchasing office needs supplier documentation and a facility manager needs dependable servicing.
The hypothetical instrument company should learn what performance would justify changing equipment, how long validation could take and what training would be needed. It should investigate manufacturing consistency, component supply, maintenance, calibration, data compatibility and relevant regulatory obligations. These questions help determine which technical tests matter first.
A large market estimate cannot replace that understanding. Build from a defined initial customer group and explain the assumptions behind pricing and costs. Separate conversations, evaluations, nonbinding expressions of interest and signed commitments. Preserve the distinction in every financial projection.
Also identify what the award will leave unfinished. A successful feasibility study may still require engineering, further validation, manufacturing investment and a route to customers. The SBIR phase overview describes Phase III as commercialization using non-SBIR/STTR funding. Do not treat a later award, investor commitment or public purchase as automatic.
A realistic financing plan has room for the possibility that the proposal is declined or the research does not support the original idea. Public funding can support difficult learning. It cannot remove uncertainty from the business.
Leave the reader with a research case
Before submitting, ask someone outside the writing team to explain the proposal back to you. Can they identify the problem, the agency fit, the technical uncertainty, the decisive experiment, the qualified team and the route toward use? Can they tell which statements are established, which are preliminary and which are hypotheses?
If those answers are unclear, improve the reasoning before adding more persuasive language. Preserve versioned protocols, source records and honest descriptions of prior results. Check the final submission against the current instructions, and allow time for registrations and administrative review.
Research funding is competitive, and neither eligibility nor a well-prepared proposal guarantees an award. The stronger ambition is to develop a case worth examining: a problem that matters, a question that can be tested and a credible plan for turning the result into something people can use.
Feel Worldwide Foundation Inc. welcomes partnership inquiries about practical entrepreneurship education, responsible research collaboration and sharing relevant expertise. Any collaboration would require a mutually agreed scope and applicable eligibility checks; it would not constitute a funding commitment or an assurance of grant success.
You can also donate to Feel Worldwide Foundation Inc. to support its mission. A donation does not confer research funding, applicant eligibility or influence over any agency's award decisions.
