Universities Are Sitting on the Next Great Companies. Here’s How to Launch Them.

You Can’t License Your Way to Commercialization

The Texas Innovation Conference & Awards at TCU didn’t change university tech transfer overnight but that’s not a criticism, it’s almost a compliment. What it did was harder and rarer: it put the people who actually need to be in the same room, in the same room, and let them be provocative about why what universities have been doing for the past forty years mostly isn’t working.

That’s not a small thing; institutional honesty is genuinely scarce. Most conferences about innovation exist to celebrate the 3% that worked while carefully avoiding the 97% still sitting in a filing cabinet somewhere between the patents office and retirement. This one didn’t do that. This one named the problem.

Fort Worth, of all places, hosted a collision of minds from across the state and region, from my Austin to Dallas, with Adriana Ocampo Senior joining us, and Tulsa represented by Darrel Frater. Thanks to Tom Wavering and Texas Christian University for making that convergence happen, because that kind of geographic cross-pollination is exactly the connective tissue most startup ecosystems are still missing. And I was fortunate enough to host a panel featuring Ed Curtis from YTexas, Bobby Ahdieh the Vice President for Professional Schools & Programs at Texas A&M, Chief Operating Officer at Texas A&M-Fort Worth, and Dean of the Texas A&M University School of Law, and Thomas Wavering who is Chief University Strategy and Innovation Officer at TCU. The quality of the conversation in that room was infectious.

Thomas Wavering himself put the stakes plainly in his post-event remarks, noting that the inaugural conference brought together “startups, universities, investors, corporations, and ecosystem leaders from across Texas” as “a powerful demonstration of what collaboration across our state can achieve.” Four hundred attendees, 100 awardees, and sponsors including Esri, UNT Health Fort Worth, the National Academy of Inventors, and Oracle. This was not a meetup, it was a statement. Mark your calendars now for April 12-13, 2027, when it returns to Fort Worth; Year 1 being this productive means Year 2 will be remarkable.

Jasper Cannon captured what made the event structurally important, “Silicon Valley did not become Silicon Valley because it had good weather and smart people. It became Silicon Valley because Stanford decided to be part of the economy rather than just study it. MIT did the same in Boston. The research stayed. The talent stayed. The capital followed.” Cannon continued, noting that “DFW has both. We also have low taxes, lower regulatory burdens, great weather, great food, and a much better cost of living. All of the ingredients necessary to make the metroplex the next great startup hub.” He’s right that the ingredients are there. The question is whether the institutions will choose to be catalysts or continue auditioning for the role.

Pesh Chalise, a Fort Worth transplant from Minnesota who has watched the region’s momentum build for over a decade, cut straight to the core problem I want to explore more, “Across Texas universities, there’s an incredible amount of innovation happening; patents, research, and IP being developed every day. But much of it never makes it into the hands of founders, operators, and investors who can actually bring it to life. This isn’t just a platform or funding issue; there are plenty of both. It’s an access, visibility, and connection problem. Who gets to see what’s being built? How do the right people discover it at the right time? Why is so much high-potential innovation still sitting on the sidelines?”

That’s the right questions.

The Room Finally Said It Out Loud

The session hosted by Victor Fishman of Tech Research Alliance was incredible precisely because of its acknowledgment that inter-university collaboration, serving entrepreneurs, and bringing IP to market aren’t being as best as possible. When the people responsible for the problem look at the room and say, “we’re not doing this right,” it’s either the beginning of change or another round of well-intentioned theater. The private sector voices in that room echoed my own obvious questions, Why aren’t we just doing equity agreements? Why aren’t warrants on the table? Why does this still take two years? These are not exotic ideas, these are tools that every competent venture capitalist, every experienced operator, and every serious startup attorney uses constantly. In a university context, somehow, they’re considered novel.

Marcus Wolfe, Professor of Entrepreneurship; Andrew Maas, Assistant Vice President for Technology Transfer; Aaron Barker, Partner and Head of Venture Best; Suchi Acharya of AyuVis Research; and Erik Halvorsen of pH Partners were all present for a discussion of what kills deep-tech commercialization. Universities aren’t oriented to launch companies, researchers aren’t incentivized to do so, and very few people within the institutional structure have the marketing orientation or operational startup experience required to build companies from scratch. That is a structural problem, not a personality problem, and structural problems require structural solutions.

The data on this is not ambiguous; a peer-reviewed study published in PMC examining AUTM Licensing Survey data found that “a small minority of institutions producing the vast majority of technological and economic benefits” and that “a clear barrier to effective commercialization of university technology is the widespread lack of access to experienced, motivated, and well-resourced technology transfer offices” (PMC, 2018). The Federal Reserve Bank of Minneapolis has documented that less than 1 percent of active licenses held by U.S. universities yielded $1 million or more in annual revenue, and that “the vast majority of licensed inventions gain little traction in the market.” These numbers have not dramatically improved. And yet the licensing model remains the default. Somewhere, a committee is protecting the committee.

This isn’t a recent concern, either. I’ve written previously that the entire posture toward university IP quietly flipped around the 2000s, and we’re still pretending nothing fundamental changed. The Bayh-Dole Act of 1980, which allowed universities to retain ownership of federally funded research, was built for a world of slow, geographically constrained knowledge diffusion. Knowledge now diffuses in hours. AI iterates published research faster than a licensing committee can schedule its first meeting. The OECD has documented that IP ownership is increasingly decoupled from where value is actually created; yet our universities still behave as though 1985 is calling and it wants its monetization model back.

Stanford didn’t produce Silicon Valley because of its licensing office; it produced Silicon Valley despite it. The culture there was and remains one of spinning talented people out into companies. The IP office was never the main character. What institutions (and more throughout the world) need to appreciate most is that the conversation in Fort Worth was valuable not because it celebrated what’s working, but because it named what isn’t; and the naming only matters if it produces structural change. Every ecosystem city in America has hosted the acknowledgment session, the ecosystems that win are the ones that followed it with actual policy, actual incentive restructuring, and actual mechanism redesign.

University IP Is Extraordinary. University Commercialization Doesn’t Have to Be Broken.

Before getting to the real solutions, it’s worth reiterating here since this is largely, likely, a different audience that who usually reads my work: more conferences, more demo days, more accelerator ribbon-cuttings, and more innovation district announcements will not fix this. We’re uncovering that they never really fixed much; they’ve made administrators feel productive and given journalists something to photograph. As I’ve argued consistently, “you cannot committee your way to entrepreneurship.” The private sector operators in Victor Fishman’s session were asking the right questions precisely because they work in a world where deals get done instead of discussed. The gap between how universities handle IP and how the private sector handles value creation is not philosophical; it is operational and incentive-driven.

Fix the incentives and the operations follow.

Leave them unchanged and every conference produces another round of earnest frustration.

Embry-Riddle Aeronautical University is a useful comparison point here because what they’ve done in Daytona Beach with the MicaPlex Research Park isn’t a demo day program; it’s a structural embedding of companies into the university’s research environment. Companies like Wave Vector Technologies aren’t passive tenants; they’re co-developing IP in real time with faculty and students. The university’s COO Rodney Cruise was clear that the research park is “designed to foster a technology-focused ecosystem, bridging the gap between university researchers and students, and the businesses, entrepreneurs and start-ups that drive innovation.” That’s a structural decision, not an event strategy. And as Phoenix demonstrates in BioScience, the right move is never to copy Silicon Valley but to lead with the sector where you already have genuine strength.

Texas has plenty of genuine strength, your state does too, the question is whether universities will restructure to let that strength compound into companies instead of letting it age gracefully in patent portfolios.

Here is what that restructuring actually looks like.

The Research Is There. The Deals Aren’t.

One of the most obvious and consistently underused tools in the university ecosystem toolkit is the Master Collaboration Agreement (MCA): a pre-negotiated, standing framework between universities that governs how they’ll jointly own, license, and commercialize IP when their researchers work together. Right now, every joint research project between two universities typically requires its own from-scratch negotiation of IP terms, which is time-consuming, expensive, and almost designed to kill momentum before a founder can even enter the conversation.

An MCA pre-establishes who owns what percentage of jointly developed IP, how licensing revenues are split, how startups can access the IP, and what governance structure manages the commercialization process.

The entrepreneur brain in me practically fell asleep just trying to write that.

The Engineering Research Centers program operated by the NSF has documented cases where multi-university IP agreements with central management have dramatically reduced friction by creating “one-stop shopping” for industry members who want to license or partner on research developed across institutional lines. Texas has nine major research universities. If even three of them established a standing MCA framework, the number of joint research outputs that could reach founders in 90 days instead of 18 months would be transformational. The infrastructure for this exists. Georgetown University’s Office of Technology Commercialization already operates a multi-agreement framework distinguishing between research collaboration agreements, sponsored research agreements, and inter-institutional agreements; the architecture is understood.

What’s missing in Texas is the political will across university presidents and general counsel offices to prioritize velocity over institutional sovereignty.

Stop Licensing University IP. Start Building Companies From It.

The most direct structural fix to the tech transfer problem is also the one universities resist most fiercely (which is a reliable tell that it’s the right answer): replace licensing fees and royalty structures with equity positions, and hold those equity positions in a separate 501(c)(3) nonprofit that is purpose-built to manage startup stakes without the conflicts of interest and administrative overhead of a standard university IP office.

The current licensing model requires a founder to pay fees upfront or promise royalties before they’ve validated a market, which is the financial equivalent of charging someone to attempt a marathon before they’ve proved they can run even a mile. It handicaps the very company the university theoretically wants to succeed. As documented in ScienceDirect’s research on university equity positions, equity-based approaches allow universities to share in the upside of success rather than extracting value from the process of attempting it; and “research suggests that equity is better for universities in the long run” even though “most universities still seem to prefer the royalty route.” The reason most universities prefer the royalty route, if we’re being direct, is that royalties feel predictable and controllable to administrators who’ve never operated a startup and have no framework for evaluating equity risk. That is not a principled position; it is bureaucratic risk aversion at the expense of the founders and communities they’re nominally trying to serve.

The 501(c)(3) holding structure resolves the university’s legitimate concern about managing private equity stakes, handling conflicts of interest, and navigating the governance complexity of being on a startup’s cap table. A purpose-built nonprofit entity with a board that includes experienced venture investors, operators, and university representatives can hold these equity positions professionally, make rational decisions about when to sell or dilute, and reinvest proceeds into the next generation of university spinouts. This is not a radical idea; MIT’s economic impact model has long prioritized startup formation over licensing revenue, and the economic evidence from MIT alumni-created companies consistently dwarfs what its licensing office ever generated. Universities don’t need to invent this model; they need to adopt it and adapt it to their sectors.

The Equity Play Universities Keep Avoiding

A warrant agreement, an instrument common in private markets where a company grants the holder the right to purchase equity at a defined price and time, is the most elegant structural compromise available to universities that aren’t ready to fully abandon licensing but recognize that upfront fees are killing early-stage startups before they get traction.

Under a warrant structure, the university doesn’t charge the founder a licensing fee or claim royalties during the startup’s fragile early stage. Instead, the university receives a warrant to purchase equity in the company at a predetermined price, exercisable if and when the company raises a defined amount of capital or reaches a specified revenue milestone. This means the university has real economic upside tied to the startup’s actual success rather than a tax on its attempt to exist. The founder doesn’t have to fund a licensing office while simultaneously trying to find product-market fit; they can move, test, and build. The university doesn’t walk away empty-handed if the company succeeds; it participates in the outcome.

Extantia Capital’s Yair Reem analyzed of tech transfer negotiation, noting that “universities can stand to gain more revenue from this [equity route] than from upfront payments or royalties” and that warrants are a legitimate mechanism for structuring anti-dilution protections with defined triggers. The legal infrastructure for this already exists in standard venture agreements. University general counsel offices need to get comfortable with it, which is an investment in understanding modern startup finance, not a legal moonshot. Aaron Barker’s presence at the TCU session as a venture partner is exactly the kind of expertise universities need inside the building.

Researchers Invent It. Nobody Builds It. Policy Is Part of Why.

One of the underappreciated barriers to university commercialization isn’t greed or bureaucracy alone; it is genuine institutional risk aversion grounded in real legal exposure.

• University administrators worry about conflicts of interest when faculty take equity in spinouts
• Researchers worry about their academic tenure and publication freedom being compromised by commercial entanglements
• Tech transfer offices worry about Bayh-Dole compliance, export control regulations, and the liability that comes with placing a university’s institutional reputation behind a startup that might fail spectacularly

These fears aren’t entirely irrational, but they are currently being managed by avoiding commercialization rather than by building policy cover that enables it safely. The solution is for state governments to create explicit regulatory safe harbors for university equity participation in startups, define clear conflict-of-interest disclosure frameworks that don’t require faculty to choose between entrepreneurship and academia, and establish liability shields for university tech transfer offices that act in good faith to commercialize research. Senator Cornyn’s office has engaged on defense-related tech transfer policy; there is a legislative pathway here that doesn’t require a federal act of Congress.

The model already exists in part. The SBIR/STTR programs provide federal grant frameworks that create defined commercialization pathways without IP transfer risks; the issue is that most university researchers don’t have the entrepreneurial fluency to navigate them independently. Pairing state-level regulatory cover with SBIR-aligned commercialization frameworks, and training research staff in how to use both, would materially change the math for faculty who currently see commercialization as career risk rather than career opportunity. The Kauffman Foundation has repeatedly documented that high-growth firms emerge from execution and timing rather than invention alone; giving researchers the policy cover to act on timing is the direct intervention.

The IP Is on the Shelf. The Solutions Aren’t.

Pesh Chalise nailed the diagnosis at the TCU event: this is an access, visibility, and connection problem.

Founders and investors can’t build on university IP they don’t know exists while researchers can’t find partners for their IP if it’s locked in a PDF somewhere on a university’s internal network, cross-referenced with a patent number that tells you nothing about market application. The solution here is both technically straightforward and institutionally difficult, which is why it hasn’t been done: a shared, searchable, public-facing platform that indexes the available IP across universities, attaches plain-language market application descriptions to each patent, identifies the researcher and their contact information, and indicates the commercialization terms available.

This isn’t a novel concept. Organizations like Inpart operate online matchmaking platforms operate online globally that connect university research to industry by “simplifying the initial connection between teams in academia and industry based on the alignment of research interests and priorities.” What states need are state-specific versions of this, co-owned by the participating universities, integrated with the state’s startup support infrastructure, and actively populated by tech transfer offices that are incentivized to see their IP matched with founders rather than to accumulate it as a portfolio metric. Chalise has already indicated he’s working to build something in this space; that effort deserves institutional support from the universities whose research would populate it. A platform without IP is useless, and IP without a platform stays on the shelf.

The mentor side of this equation matters equally. The room at the Innovation Conference contained extraordinary expertise; Ed Curtis’ network across YTexas, the legal and policy depth Bobby Ahdieh brings from Texas A&M Law, people like me and Darrel Frater, the entrepreneurship faculty and practitioners spread across TCU, UT, A&M, and the Dallas-Fort Worth ecosystem. Currently, a founder trying to navigate deep-tech commercialization has to already know who to call. The platform needs to also index mentors, advisors, and domain experts alongside the IP so that a founder discovering a piece of research immediately gets a connected path to the humans who can help them act on it.

Good Research Doesn’t Need a Licensing Office. It Needs a Founder.

Professors are inventors, not founders. This is not an insult; it’s a functional distinction that the university tech transfer world needs to internalize and stop pretending away with entrepreneurship programs. A researcher who spent ten years developing a breakthrough in materials science has skills that are genuinely extraordinary and commercially valuable; the ability to build a sales pipeline, manage investor relations, recruit a go-to-market team, and navigate a seed round is an entirely different capability set that most of them don’t have and shouldn’t be expected to develop on the side of their research careers.

The spinout studio model addresses this directly; instead of asking researchers to become founders, universities partner with experienced operators and venture studios to co-create companies around research insights. The IP is licensed or contributed to the new entity; the university takes an equity stake; the operator team handles company building; the researcher serves as technical advisor or co-founder with a role aligned to their actual skill set. Imperial College London and ETH Zurich have both moved toward this hybrid model, prioritizing repeatable company creation over patent licensing. The key structural insight is that the operator brings the commercial capability the researcher lacks, and the researcher provides the technical depth the operator couldn’t replicate.

The combination is what builds a defensible company.

This means university tech transfer offices actively recruiting and maintaining relationships with venture studios, serial entrepreneurs, and operators who have deep-tech commercialization experience, and building standing agreements that allow those operators to step in quickly when promising research is identified. It also means the university needs to be willing to take a smaller equity position in exchange for faster company formation; the tradeoff of owning 5% of something that actually launches versus 30% of something that never leaves the lab is not a difficult one to evaluate if you’re genuinely trying to create economic impact.

EIR Programs Without Teeth Are Just Business Cards

Most universities with “Entrepreneur-in-Residence” programs have a business card for the role and a parking spot for the participant. The EIR sits in a glass office somewhere near the student union, fields appointments from seniors who want to start a pizza delivery app, and has no formal mandate, no budget authority, no IP access, and no clear pathway to help a researcher commercialize a discovery from the lab down the hall. This is not an EIR program; it is an EIR costume.

A properly resourced EIR program operates with a defined commercialization mandate: identify the most commercially promising research in the university’s pipeline, connect with the researchers, evaluate market fit, and actively drive company formation either as a founding operator or by recruiting one.

The EIR needs access to the university’s patent disclosures and licensing pipeline before they become public, a budget to cover initial company formation costs like legal fees and market research, and a compensation structure that includes meaningful equity in the spinouts they help launch rather than a flat honorarium. The University of Minnesota’s Tech Comm office, ranked second among public universities for startup formation according to AUTM’s 2023 survey, is launching a record 26 startups per fiscal year; that doesn’t happen without dedicated human beings who are incentivized and empowered to make it happen, not just present to observe it.

Most cities have the talent to staff this well; the Dallas Fort Worth ecosystem specifically has, as Jasper Cannon noted, a building set of “connective tissue, the places where researchers meet operators, operators meet founders, and founders meet capital before they have to get on a plane to find it.” EIR programs that are actually resourced are one of the primary mechanisms for building that connective tissue into the university itself rather than hoping it forms organically in the surrounding ecosystem.

If You Won’t Move the IP, Someone Else Should Get To

Here is a policy that would change commercialization overnight (and that university administrators will hate in direct proportion to how necessary it is), if a university licenses IP to a startup or identifies research as commercially viable but fails to commercialize it within a defined window, say 24 to 36 months, the rights automatically revert to the founders, enter a shared commons, or become available for non-exclusive licensing at nominal cost.

This addresses what I’ve called the “IP hostage” problem: promising research that gets locked into a licensing negotiation or sits in a queue waiting for the tech transfer office to prioritize it, while the market window closes and the founders either give up or find a workaround. Time-bound reversion creates urgency for the institution without punishing the founder. If the university moves efficiently, it retains its commercial position. If it doesn’t, it loses that position to someone who will actually build something with the research. The NIH has experimented with non-exclusive licensing frameworks for certain biomedical technologies with demonstrably better diffusion outcomes; the Federal Reserve’s analysis of tech transfer makes clear that without market pull and commercial urgency, most licensed inventions stall permanently. Reversion policy replaces passive stalling with structural incentive.

Measure Startups, Not Agreements

None of the structural changes above will fully take hold if researchers and staff face no upside from commercialization and no consequences from inaction. Currently, most research university tenure and promotion systems reward publications, citations, and grant revenue; commercialization activity is a secondary consideration at best and, in some departments, actively viewed as a distraction from “real” academic work.

A researcher who successfully spins out a company, creates jobs, and generates economic value for their region is doing something extraordinary for society; their institution’s tenure committee may or may not care.

Restructuring incentives means tenure review frameworks that formally credit commercialization activity alongside publication record, not as a substitute but as a parallel pathway to recognition. It means royalty and equity participation terms for researchers that are competitive with what a co-founder would receive in the private sector, not the current structure where the inventor often sees 20 to 30 percent of net revenues after the tech transfer office takes its costs, which has consistently been shown, amounts to very little given how rarely licenses generate significant revenue. And it means staff at tech transfer offices being evaluated on startup formation rates and early-stage capital raised by spinouts, not just on licensing agreements executed or patent applications filed.

The difference between a metric and an outcome is the difference between measuring activity and measuring impact; and as I’ve argued in the context of startup ecosystem metrics broadly, cities and institutions that measure activity while expecting outcomes are always surprised when the outcomes don’t arrive.

A Node, Not a Spectator. Now Do the Work.

What happened in Texas is what Jasper Cannon described precisely, “a university choosing to be a node rather than a spectator.” That choice has to be made deliberately and repeatedly. The team at the TCU University Strategy and Innovation Office program is, as Cannon put it, “expanding the university’s presence and it couldn’t come at a better time.”

States are not starting from zero; multiple anchor research universities, massive and growing technology and defense sectors, many with low regulatory burdens (some have to fix that), where cost of living is in good shape, entrepreneurs take risks, and when cross-regional energy is present, that events like this one in Texas demonstrate, we have regions and sectors, not cities, wherein startups thrive. The Kansas City comparison is instructive: even Heartland cities are figuring out that university-anchored ecosystems require deliberate infrastructure, not just talent and research. The difference between cities that develop into regional hubs and cities that perpetually “emerge” is structural; it’s whether the institutions in the ecosystem have built the mechanisms for knowledge to flow to founders efficiently, or whether they’re still protecting it behind licensing tables.

Phoenix, in not copying Silicon Valley, is teaching a lesson that every city should absorb; that the winning move is to identify your actual sector strengths and build commercialization infrastructure around those, not to build a generic “innovation district” and hope founders arrive.

In Texas, Fort Worth has defense and aerospace. Austin has GovTech and civic innovation. Houston has energy and biomedical. Dallas has finance, logistics, and professional services. These are not generic innovation assets; they are sectors where university research and startup formation can compound toward genuine global significance if the institutions stop taxing the process and start fueling it.

The private sector participants in that Fort Worth room were asking exactly the right questions, and the institutional participants were acknowledging the honest answers. What comes next determines whether April 2026, for Texas, was the beginning of something real or another data point in the long history of conferences about fixing tech transfer that didn’t fix tech transfer.

The conversation was extraordinary. The work we can do now is what makes it matter.

---

Where is your university or regional ecosystem on the commercialization curve; still defending the licensing model, or starting to build the mechanisms that actually move research into companies? The people who need to act on this aren’t in San Francisco. They’re in rooms like the one in Fort Worth.

Explore more in how university culture and commercialization connect, why the current IP model is a tax on progress, and what university commercialization should actually look like.