How to Find a University Partner for STTR (When You Don't Know a Single Professor)

Cada has written 100+ grant proposals across 30+ agencies, including NIH, NSF, NASA, and DoD.

Every STTR guide assumes you already have a professor. Most are written by a university's technology transfer office for its own faculty, so they start from the one position you are not in.

Here is how to find a university partner for STTR from a cold start: define the 30% you are outsourcing as a testable aim, build a list of 12 to 20 candidate principal investigators from NIH RePORTER and NSF Award Search, cold-email them with the scope and the funding source named, and budget 8 to 10 weeks for the paperwork.

That is the method. The rest is the detail behind each step, the per-agency rules that change the answer, and the three places founders lose a cycle.

If you have not yet decided between SBIR and STTR, decide that first. This guide picks up at the point where STTR is the answer and the partner is the blocker.

What counts as a university partner for STTR?

An STTR research institution partner must be a single US college or university, a federally funded research and development center, or a domestic nonprofit research institution. It must be domestic and nonprofit, and the 30% cannot be split across several institutions. The partner is the institution, not the professor.

You recruit a principal investigator, but the entity that signs the subaward, sets the indirect rate, and negotiates the IP terms is the institution behind them.

Four things that sound like they qualify and do not:

• A professor consulting personally. Work billed through a personal LLC or consulting agreement is a consultant cost, not a research institution subaward. It does not satisfy the STTR structure.
• A foreign university. The institution must be domestic. An overseas collaboration can exist alongside the STTR, but it cannot be the 30%.
• A for-profit contract research organization. A CRO can do excellent work and can be a subcontractor. It is not a nonprofit research institution.
• A national lab you found on a website. Federally funded research and development centers can qualify, but they partner through their own agreements office, not a friendly scientist.

If you cannot name the office that will countersign the subaward, you do not have a partner yet.

What is the STTR work split? It is 40/30, not 30/60

Founders search for "STTR 30 60 split requirements" constantly. That phrasing is not the rule. The 30 is real, but it is the research institution's floor rather than yours, and 60 appears nowhere in the statute.

Under STTR, the small business must perform at least 40% of the research and development work and the partnering research institution must perform at least 30%. The remaining 30% can go to either party or to a third party. The split is measured against total cost, it is set by statute, and it cannot be waived. Source: SBA SBIR/STTR Policy Directive (May 2023).

The SBIR rule, from the same directive, is different and constantly confused with it:

Two things follow from that table, and both change what you should apply for.

First, STTR locks 70% of your budget to two named parties before you write a word. That is not a drawback, it is the deal: you trade budget flexibility for institutional capability you do not have to build.

Second, if you want a university involved but not the 30% handover, you do not need STTR at all. An SBIR Phase I lets you subcontract up to 33% to that same university with no cooperative agreement, no allocation-of-rights requirement, and no co-applicant. Founders route themselves into STTR when a plain SBIR subcontract would have been faster.

On a $300,000 NIH STTR Phase I, the 30% floor is $90,000 routed to the institution. Lead with that number in any outreach: it converts an abstract collaboration request into a funded scope.

Which agencies actually run STTR?

Only five federal agencies run an STTR program (sbir.gov): the Department of Defense, the Department of Energy, Health and Human Services (which is NIH for most applicants), NASA, and the National Science Foundation. Eleven agencies run SBIR.

If your best-fit agency is not on that list, the partner question is moot. You are looking at an SBIR with a university subcontract, and the paperwork gets easier.

For the five that participate, the rules that change your timeline are not the split, which is statutory and identical everywhere, but the IP agreement timing and PI employment flexibility.

Sources: NIH Grants Policy Statement Section 18.5, the NSF SBIR/STTR Phase I solicitation, and the DoD STTR BAA.

The NSF row is the one that costs people cycles. NSF wants something in the package and does not want a fully executed agreement.

A draft, or a letter from the institution naming itself and committing to provide the agreement on notification of a potential award, satisfies it. Read it as "required now" and you burn four weeks negotiating early. Read it as "not until award" and you omit a required document.

If you are applying to DOE or NASA, confirm the current solicitation rather than this table. Both revise partner documentation between cycles more often than NIH or NSF.

How to find a university partner for STTR in four steps

Step 1: Write the 30% before you look for anyone

The most common sourcing failure is emailing professors before you know what you want them to do.

Write one paragraph: the single technical sub-question you cannot answer in-house, the method that answers it, the deliverable, and the dollar figure attached. Roughly 30% of the total.

An illustrative version, for a fictional company developing a wearable sensor:

We need an independent validation of our sensor's drift characteristics across a 12-week wear period in 30 subjects, run in a lab with existing IRB coverage for wearable device studies. Deliverable is a validation dataset and a methods write-up suitable for the Phase I final report. Roughly $90,000 of a $300,000 NIH STTR Phase I.

That paragraph is your entire sourcing asset. It makes a cold email answerable in 30 seconds instead of a 40-minute discovery call the professor has no reason to take.

If you cannot write it, the problem is not that you lack a partner. Your Phase I scope is not decomposed yet, and any partner you sign will be doing work you invent later to justify the split. Reviewers see that in the proposal.

Step 2: Build the candidate list from funding databases, not Google

Google gives you the most visible person in a field. You want the most available one who has already done the method.

Two free databases do almost all of this work.

NIH RePORTER (reporter.nih.gov) indexes every NIH-funded project with abstracts, PI names, institutions, and publications.

Search on your method keywords, not your market. "Wearable sensor drift validation" finds people who do the work. "Digital health" finds 40,000 projects and no one useful.

NSF Award Search does the same for NSF-funded work and is the better source for engineering, materials, computing, and physical sciences.

Three filters that matter more than prestige:

1. Active funding in the last 3 years. An expired grant means the lab may have moved on, lost the postdoc who ran the assay, or dismantled the rig.
2. Any R41 or R42 history, the NIH STTR Phase I and Phase II activity codes. A PI who has done an STTR before means the sponsored-programs office has a template, has had the allocation-of-rights conversation, and will not treat you as a novel case. Worth more than any other single factor, and routinely cuts three weeks off the timeline.
3. Mid-career, not the department chair. Associate professors with active labs answer email, and they are building their own funding portfolio, so a new sponsor relationship is worth something. The famous name has 400 unread messages and delegates anything cold.

Target 12 to 20 names. You are running a funnel, not a search for the one right person.

Step 3: Send the cold email

Here is the template in full. There is nothing proprietary about it.

Subject: STTR partnership -- [specific method] validation, ~$90K subaward

Dr. [Name],

I read [specific paper, with the actual finding in five words]. We are building [one sentence on the technology] and we need exactly the measurement your group does.

We are preparing an NIH STTR Phase I (R41) for the [month] deadline. The structure routes at least 30% of the budget to a research institution partner, which here would be roughly $90,000 of a $300,000 award.

The scope: [the paragraph from Step 1].

Would you have 20 minutes in the next two weeks to tell me whether this is a fit? If it is not, I would value a pointer to someone it is a fit for.

[Name], [Title], [Company]

Four things make it work, and all four are load-bearing.

A named, specific paper. Not "your work on sensors." The actual finding. This is the only signal separating you from the automated outreach in their inbox.

The scope, not the vision. They are evaluating whether they can do a defined thing, not whether they believe in your company.

The money, already identified. Most academic collaboration requests arrive with no funding attached, making them unpaid favors. Naming the mechanism, deadline, and dollar figure moves the request into a category their sponsored-programs office already handles.

A bounded ask, plus an exit. Twenty minutes, and an explicit offer of a referral instead. The exit raises reply rates: it gives a busy person a way to be helpful in 30 seconds.

Send 12 to 20 and expect two to five replies, of which one or two become real conversations. If you send three emails and conclude STTR is impossible, you did not run the process.

Step 4: Run the 20-minute call

Two of these five questions are about the science. Three are about the institution, and those determine whether you can actually submit.

1. "Has your group done an SBIR or STTR subaward before?" If yes, most of your timeline risk just evaporated. If no, add three to four weeks and expect to explain the mechanism to their sponsored-programs office yourself.
2. "Who handles the subaward budget, and what is their turnaround?" You want the name of the office and a number of days. If the PI does not know, that is information: they have not done this recently.
3. "What is your institution's F&A rate for this kind of work?" This determines how much research your 30% actually buys. Section below.
4. "Who does the day-to-day work, and what percent effort would you personally commit?" A named postdoc at 50% effort beats the PI claiming 10%. The answer you cannot work with is no name and no number.
5. "Does your technology transfer office have a standard allocation-of-rights template?" Ask for it on the call. Reading their default position is worth more than any advice about negotiating it.

What must be true before you name an STTR partner?

Naming an institution you have not secured is the most common quiet failure. All six must be true:

1. A named PI has said yes in writing. Email is fine. Verbal on a call is not, because people forget.
2. The sponsored-programs office knows the proposal exists. If the PI has not filed internal notice, the institution has not agreed to anything. The failure mode is a signature request landing on a stranger's desk four days before the deadline.
3. You have a budget from the institution, not an estimate from the PI. PIs routinely quote direct costs and forget F&A, which understates the true subaward cost by roughly 20% to 35% depending on the institution's rate.
4. The 30% is real work you are not otherwise doing. If the institution's scope duplicates your own aims, reviewers read it as a compliance maneuver rather than a collaboration.
5. You know your agency's IP agreement timing. See the table above. Getting this wrong costs either a required document or four weeks.
6. You have read the institution's standard allocation-of-rights template. Not negotiated it. Read it. You need to know whether their default position is survivable before committing the proposal to them.

How long does a university subaward actually take?

Plan on 8 to 10 weeks from first cold email to a signature-ready package. Four weeks is the compressed floor and only works if the PI has done an STTR before.

That last row is the one that ends applications. Universities impose internal submission deadlines ahead of the sponsor's.

The University of Alabama at Birmingham requires complete materials five full business days before the sponsor deadline. Utah State requires four.

Those are published policies, not courtesies. A small business asking for an exception has no standing to get one.

If the NIH deadline is September 5, the university's real deadline is around August 29. Work backward from that date, not the agency's.

On the allocation-of-rights range, I should be straight about the uncertainty. Publicly reported timelines run from two weeks to several months depending on term complexity, institutional variance is enormous, and no surveyed dataset exists. Treat 2 to 6 weeks as a planning assumption for standard terms, more if an exclusive license is negotiated alongside.

What is an STTR allocation of rights agreement?

An STTR allocation of rights agreement is the written contract between the small business and the research institution that does two things: it allocates ownership of intellectual property developed under the award, and it allocates the right to carry out follow-on research, development, and commercialization.

The second matters more to investors. Ownership of a Phase I result is worth less than the right to commercialize it.

Four terms decide whether the structure is survivable.

Background IP carve-out. Your pre-existing technology stays yours. Uncontroversial in principle, occasionally not in practice.

Foreground IP ownership. Inventions made under the award. Institutions generally retain ownership of what their employees invent, under the Bayh-Dole framework. Normal, and not by itself a problem.

Option to an exclusive license in your field of use. The term that decides everything. If the institution owns the foreground IP and you have no option to license it exclusively in your commercial field, you have funded research you cannot build a company on.

A standard pattern grants the small business an option period followed by a defined window to conclude the commercial license: often 90 days from the point the option is exercised, extendable by agreement.

Fee and royalty structure. Where the actual negotiation happens, and usually the slowest term.

Public model agreements exist and reading them costs nothing. The Department of Defense publishes an STTR Model Agreement for the Allocation of Rights, NIH publishes a model STTR agreement, and many sponsored-programs offices post their own.

To be direct about the boundary: you can run the sourcing yourself, and this guide gives you what you need.

Do not sign license terms without counsel. Sourcing is a process problem; licensing is a legal one, and a wrong field-of-use term shows up two years later in diligence.

How much does the university subaward actually cost?

Thirty percent of the work is not 30% of your research, because the institution's indirect rate comes out of the same envelope.

Work the numbers on a $300,000 NIH STTR Phase I with a $90,000 subaward.

If the institution applies an off-campus F&A rate near 26%, common in published university rate schedules when work happens outside university facilities:

• Subaward direct costs: about $71,400
• Institutional F&A: about $18,600
• Total: $90,000

If the same institution applies a 55% on-campus rate:

• Subaward direct costs: about $58,100
• Institutional F&A: about $31,900
• Total: $90,000

Same $90,000. About $13,300 less actual research.

This is why question 3 is not small talk. Ask the F&A rate before choosing between two equally capable labs.

One more mechanic that surprises people. Under federal cost principles (2 CFR 200), indirect costs are calculated on modified total direct costs, and only the first $25,000 of each subaward sits in that base.

Your $90,000 university subaward therefore contributes $25,000 to your own F&A base, not $90,000. At a 40% company rate, that is $10,000 of indirect recovery on a $90,000 pass-through.

One correction, since it circulates widely: there is no 15% NIH indirect cost cap in effect. NIH's February 2025 attempt was permanently enjoined, the First Circuit upheld that injunction on January 5, 2026, and the Justice Department let the Supreme Court deadline lapse that April. Budget against your negotiated rate and confirm it with your target institute.

Three mistakes that cost founders a cycle

Sourcing the partner after the proposal is written

You write the full Phase I, then look for someone to hand 30% to. The 30% becomes whatever is easiest to detach rather than what genuinely needs institutional capability.

Reviewers read this accurately: the collaboration section describes a subcontract, not a partnership, and the aims do not depend on the institution. Fix: Step 1 comes before drafting, always.

Recruiting the famous name who hands it to a postdoc

A well-known PI agrees, commits 5% effort, and the work goes to a second-year postdoc you have never spoken to.

This is not automatically bad. It is bad when you did not know.

Fix: question 4 on the call, then ask to meet whoever does the work. A committed postdoc at 50% effort beats a distracted full professor at 5% every time.

Signing an IP structure that kills the Series A

The institution owns the foreground IP, you have a non-exclusive license, and there is no field-of-use option. Every term looked reasonable in isolation. Two years later, diligence asks who owns the core technology and the answer is complicated.

Fix: item 6 in the readiness checklist. Read the institution's standard template before you commit the proposal to them, and get counsel on the license terms specifically.

Still not sure how to find a university partner for STTR?

The method above is complete. Nothing is held back, and a founder with 8 to 10 weeks can run all four steps without help.

The hardest part is not the outreach. It is knowing which institutions have the capability you need, whether STTR is even your best-fit mechanism, and what the subaward should look like before you start a conversation you cannot take back.

Cada runs a free 20-minute STTR partner-fit call: we identify candidate institutions and mechanisms for your technology and sketch the subaward structure before you send the first email. If a plain SBIR subcontract serves you better, we will tell you that instead.

Bring your Phase I aims in whatever state they are in, a rough total budget, and your target deadline. You leave with a candidate institution list and a decision on SBIR versus STTR. No pitch, no obligation.

If you have a deadline in the next 90 days, have that call this week rather than next month. The university's internal deadline arrives a week before the agency's.

Sources

• SBIR.gov -- SBIR/STTR Policy Directive, the 40/30 STTR work-split requirement, and the list of participating agencies
• NIH Grants Policy Statement, Section 18.5 -- SBIR/STTR eligibility, research institution definition, and PI employment rules
• NIH SEED -- NIH small business program basics and the NIH model STTR agreement
• NIH RePORTER -- searchable index of NIH-funded projects, PIs, institutions, and R41/R42 award history
• NSF Award Search -- searchable index of NSF-funded awards and investigators
• NSF SBIR/STTR (America's Seed Fund) -- Phase I solicitation, Allocation of Rights document timing, and PI work-authorization standard
• DoD SBIR/STTR Innovation Portal (DSIP) -- joint submission requirement and the DoD STTR Model Agreement for the Allocation of Rights
• 2 CFR 200, Uniform Guidance -- modified total direct cost base and the $25,000 per-subaward cap on indirect recovery
• University of Alabama at Birmingham and Utah State University published internal proposal submission deadline policies -- the five-day and four-day internal deadlines
• Sourcing funnel numbers, timeline stages, and call questions are drawn from Cada's STTR partner-sourcing work across NIH and NSF engagements (internal Cada data, 100+ proposals across 30+ agencies)

STTR partner documentation requirements are set per solicitation and change between cycles, and DOE and NASA revise theirs more often than NIH or NSF. Verify against the solicitation you are applying under before you plan a timeline. Nothing here is legal advice, and the allocation-of-rights license terms in particular require counsel rather than a template. All company examples are fictional and used for illustration only.