The Foundry Manifesto: On Scientists Starting Companies
or, the foundation of Foundry.
There are high-potential founders in our research institutions. They can create and scale world-changing solutions in startups. The potential is in the future of these rare people, not their IP. Those who have the courage to found and build startups should be rewarded. They deserve to be supported from the very beginning with the full force of the best frontier technology founders, operators, and capital in the world.
We face large-scale, and in some cases existential, unsolved problems. They are wicked problems: food scarcity, cardiovascular disease, cancer, climate change, antibiotic resistance, rising rates of chronic and mental illness, falling fertility, building semiconductors to meet our exponentially growing appetite for compute, the cost (avg <$2B) and failure rate (>90%) of drug discovery—the list goes on.
What will it take to surmount these challenges? Terence Tao—Australia’s first Fields Medalist—coined a law of conservation of difficulty:
In order to prove any genuinely non-trivial result, some hard work has to be done somewhere.
Or, to generalise:
Our wicked, complex problems demand solving in a way that is, in some way, wicked and complex.
That is: there is no single person, technology, policy, or company that will solve any one of our wicked challenges—no silver bullet, no easy way out. However…
[1] Research-led companies will cornerstone the solutions to humanity’s greatest problems.
We’ve witnessed incredible feats achieved on the back of research over the past few years. The COVID-19 vaccines were developed within a year, where most vaccines take a decade or more to bring to market. Billions of dollars of funding and the urgency of a global pandemic definitely sped things up, but we also got lucky: the technology to create mRNA vaccines was just ripe enough, after some fifteen years of research and investment, to save the world.
New fundamental technologies will need to be raised to the greatest possible impact with urgency and excellence this decade. Startups are great for this: creative business models + ambitious technical roadmaps make startups true experiments in scale of impact.
It is only through focused investment—both in effort and capital—that we will make these solutions ready.
At the intersection of global-scale challenges and frontier technology, ambitious scientists are uniquely placed to make a change.
[2] Researchers can found and lead great companies.
Researchers are curious, creative, first-principles builders who imagine better worlds and grind to make them real. They are individual marching bands of skills: scientists, technologists, managers, marketers, recruiters, advisors, writers, strategists, advocates, and so much more. Many are hungry to move fast and have great impact by creating and scaling solutions to important problems.
In short: they are early founders.
When researchers found companies, the core technology and mission become embedded in company culture. Technology builds to product, which builds to business and market; in researcher-led companies, the technology is a living, breathing, ever-growing thing. The result is a vivid technical roadmap that continues to progress over decades, as magic compounds across multiple generations of the core technology. At the foundation is that earned insight about how the world is or should be.
Corollary: We believe that researchers, scientists, and engineers do not need extensive business training or to be artificially paired with MBAs to build a great company. If you have the capability to become among the very best in the world at something - as scientist-founders often already are - then with the right coaching, community, and encouragement, you can master the intricacies of customer discovery, sales, and pricing.
[3] Australia and New Zealand’s researchers are particularly well positioned to be founders.
It’s widely accepted that Australia and New Zealand punch above their weight in terms of quantity and quality of scientific output – their contribution stands up to peers as the best in the world. We want to call out three other aspects where we see our researchers shining.
First, they are native to ambiguity, rapid learning, and creative experimentation. A few examples:
In the early 90s, Fiona Wood and Marie Stoner invented spray-on skin—a revolutionary modality for burn treatment. Wood is now collaborating with startup Inventia Life Sciences, Gordon Wallace (University of Wollongong, ARC Centre of Excellence for Electromaterials Science), and Pritinder Kaur (Curtin University) to create a bioprinter for use in surgery: Ligō.
Alan MacDiarmid won the 2000 Nobel Prize in Chemistry for his creation of electrically conductive polymers, which are widely used now in OLEDs and organic polymer solar cells, a next–generation solar technology.
Gerard Milburn, along with US collaborators, designed the initial scheme for quantum computing with linear optics in 2001. Fifteen years later, Aussies Jeremy O’Brien and Terry Rudolph, and their British cofounders Pete Shadbolt and Mark Thompson, used that work to create quantum computing behemoth PsiQuantum.
Elizabeth Blackburn won a 2009 Nobel Prize for her discovery of telomeres and their regenerating enzyme, telomerase—a fundamental piece in our understanding of cells that continues to affect therapeutics development and longevity science.
Second, our researchers are relentlessly resourceful and endlessly determined. Academic research is a difficult path. PhD candidates, who do the bulk of “hands-on” research, are paid ~$29k stipends to work investment-banking hours. They should be paid more. For those who gain a coveted continuing position in a university, there is the annual hyper-competitive wrestle for funding. Australian Research Council grant success rates sit at ~18% (for the Discovery scheme), and National Medical Health Research Council sit just below 15%.
Third, they are great teachers and mentors. Australia and New Zealand are home to generations of great science.
[4] Yet, ANZ has not yet lived up to the fullness of this potential in frontier-technology startups.
If we believe researchers can build great companies, that these companies will solve consequential problems, and that Australia and New Zealand’s researchers are among the best in the world, then shouldn’t it follow that ANZ researchers will build a consistent number of world-class frontier-technology startups?
This is not the case. In fact, our rate of startup formation out of research institutions fell during the greatest venture funding bull run in history, so that Australian researchers produce companies at <25% the rate of their American or Canadian counterparts, normalised to research funding or publication output.
In 2021, Knowledge Commercialisation Australasia (the peak body for tech transfer in the region) reported in their Survey of Commercialisation Outcomes from Public Research that with nearly 44,000 researchers across Australia and New Zealand, and 1605 invention disclosures, 79 startups/spinouts (69 AU/10 NZ) were founded—an increase from 66 the prior year, but still relatively low.
Volume of startups created out of research institutions is not the target metric. Prolificity is the foundation of impact: jobs created, products in market, revenue generated, and, often as a proxy, capital raised. That is what is missing from our companies that have spun out of research institutions: insofar as valuation is a proxy metric, we have not yet had a founder-led multi-billion-dollar company spin out of an ANZ research institution. We are at the very beginning.
[5] Startup formation should be a pillar of more research groups.
There are grand problems to be solved. Often, applied research is already trying to solve them—but impact is limited, because there’s only so much you can do at the bench and in a large institution of any kind. If you are eager to solve a real problem, but frustrated by the constraints—of pace, capital, and scale—in an institution, then you should consider founding a startup.
A research career is a worthwhile pursuit in itself, and established researchers should not feel they need to start companies on the side. In fact, we think this part-time founding usually leads to a net neutral or negative result. Startups might be seen as an experiment, but they should not be half-assed.
However, we see startup formation as a pillar of the modern applied research group. Established researchers partnering with early-career researchers to create companies builds a talent-centric flywheel:
There are examples of this across the world, and particularly in the US academic system. Bob Langer at MIT has co-founded over 40 companies, including Moderna; George Church at Harvard has spun out around 50. Langer and Church remain as leaders in their research groups, learning from their spinout experiences and better empowering their students and associated early/mid-career researchers to found and lead startups with (relative) ease.
High-performing startup culture in research institutions also pushes a second flywheel: startup successes attract the best commercially minded research talent from overseas, improve talent concentration and generate more high-performing startups, and raise the impact (and prestige) of our research institutions. Ambitious, creative students are attracted to programs where career success—including startup formation—is a viable outcome.
Implicit in this flywheel is the belief that the education is a good one—that the teachers are great, and the lessons sound. Not all startup learnings are equal. We—investors, operators, and founders—have a responsibility to inject the best global frontier-tech information possible to the ecosystem. We are lucky to be able to import the best lessons from more mature markets and use them to leapfrog the failures others have had to experience.
This is not without difficulty! It takes years to build momentum in such a flywheel. Church began spinning his research-startup flywheel in the mid-2000s, only hitting extraordinary productivity in about 2019. Australia and New Zealand’s research groups and institutions are certainly building that momentum, but it is still early days for their ability to produce founder-led startups.
There is also particular difficulty in many ANZ institutions, where conflict of interest policies and their implementations make “Co-founder and Advisor” roles nearly impossible. Academic promotion and hiring pathways are evolving, but still often weighted to publication productivity over commercialisation. We observe that both actually support each other, and should be treated as such—high-impact publications and commercial announcements go hand-in-hand, and startups are a way to accelerate research and impact beyond the auspices of the larger institution.
We would love to invite research institutions to come to the table on revising these policies and the structures by which they are enforced. The outcomes for existing students and researchers, and attracting new talent, could be immensely positive for the institutions, as well as Australia and New Zealand’s economies.
[7] Researchers who have the courage to found and build startups should be rewarded.
We believe that the best startups are led by their founders, and that those founders should be incentivised with ownership.
We are often asked what commercial terms we consider investible for startups that have undergone technology transfers out of research institutions. We will lay them out here, noting that these are only a guide, and represent only our point of view. We also note that there are no rules in startups, and we reserve the right to ignore our own rules if the situation demands it.
We prefer all-equity deals, full assignment or exclusive perpetual licensing of IP, and speed to close. In particular:
A research institution should hold up to ~10% of a founder-led company with no anti-dilution protections or atypical terms at the point of an equity capital raise. We benchmark our expectations to what we’ve seen work in global best-in-class ecosystems, in particular Stanford and MIT.
We favour all-equity deals. In the case that royalties are discussed, they should strictly apply to revenues derived from the IP.
The intellectual property should be exclusively licensed or, if possible, assigned to the startup. We are open to IP clawback clauses and milestone-based payments to ensure that licensed IP does not sit dormant.
The process should take on the order of months, and shorter if possible.
The tech-transfer process should be painless for both founders and research commercialisation offices. It should prioritise the long-term success of the company. We would support the establishment of a world-class standard IP license structure for any founder-led startup out of an Australian or New Zealand research institution—in particular, up to valuations and deal sizes that reflect the state of global startup funding.
Early investors should also not dominate the cap table. As a ballpark, we hope that founders and employees continue to own ≥~65% of the company following an initial capital raise. We understand that acceptable capital models have traditionally varied across verticals. However, we target this founder ownership as standard; it aligns with Index Ventures’ industry-standard benchmarking.
Institutions holding larger proportions of startups inadvertently hamstring those companies. It becomes far more difficult for startups to raise the high-quality capital they need to grow and succeed, and discourages the founders who will undertake the arduous business-building journey.
Institutions and their associated startups are both worse off for this, as consequential ideas that require substantial capital over many years are discouraged from pursuing private funding.
[8] Researcher-founders should be encouraged to pursue their greatest impact
There are some projects that take a few days of development before first reaching customers. Some take a few years of development before truly going to market. There are others that are ten-or-more-year builds.
The knee-jerk reaction is to pull back from the ambition when it is challenged. Building the future is a difficult thing, after all, and the reaction in our part of the world is more often than not, “Be more realistic - that’s going to take ages/cost too much/has technical risk.”
Impactful frontier-technology solutions will be more capital intensive than, and face a different set of risks to, their enterprise software and consumer counterparts. However, unretired technical risk and capital intensity is not something to fear. The real question we need to ask is whether a successful outcome is grand enough to compensate for the technical risk and capital intensity.
Reality will follow ambition. We invite researcher-founders to an ambition that may be taboo:
Ignore money and time for a moment. What is the most ambitious version of what you are building? What does the world look like in the case you are successful? What would you need to see to believe that was possible? How can you measure progress along the path to make sure that your hypotheses hold true?
[9] Investors must be as ambitious as the founders they exist to support.
Researcher-founders deserve the best capital in the world: ambitious, founder-friendly, and abundant enough to build something great.
If founders and research institutions can calibrate to lofty ambitions and creative business paths, then we are prepared to take the truly early risks. This means that Blackbird will invest before the technology or product is fully developed, at or before technology readiness level (TRL) 4. We invested in almost all of our existing frontier and medical technology portfolio companies at the idea or proof of concept stage.
We are always open for a conversation, no matter your stage. We will be sounding boards for your underbaked ideas. We will push you when it doesn’t make sense, and encourage you when it does. We will connect you to people who can help. We will provide grounded, realistic, ambitious advice based on the actual state of capital.
That is not a burden; it is an exceptional opportunity. It’s what it means to invest and believe right at the very beginning.
[10] Private funding must not replace government-funded research.
Blue-sky research holds great intrinsic and instrumental long-term value. It should not be sacrificed in favour of applied research or the pursuit of “commercialisable” IP creation.
For example: the Australian Government has supported some of the world’s best fundamental quantum science research since the 1980s, starting in condensed matter physics and quantum optics. No one could have predicted the downstream results. Australians have pioneered two world-leading quantum computing modalities, and lead the world’s best quantum computing companies. An Aussie wrote the seminal text on quantum computation and information. We are world leaders in quantum sensing research and applications.
Universities do not exist for the primary purpose of startup formation. However, we believe startups are a pillar of modern higher education and applied research. Startups are an opportunity to extend the impact of our research, empower our best researchers to accelerate development, and educate our tertiary students in the realities of building a rapidly scaling, highly impactful company.
Startups also create an opportunity for our economies to build the future industries that we need, capture a meaningful share of the commercial value that our great research produces, and guarantee our first access to the best technologies of the coming generation.
Conclusions
There are high-potential founders in our research institutions. They can create and scale world-changing solutions in startups. The potential is in the future of these rare people, not their IP. They deserve to be supported from the very beginning with the full force of the best frontier technology founders, operators, and capital in the world.
We are thrilled by the Australian Government’s commitment to critical technologies across the National Reconstruction Fund, Australia’s Economic Accelerator, and the Defence Advanced Strategic Capabilities Accelerator. We are also incredibly excited by the movement by research institutions to embrace global-standard innovation practices.
There are a great many consequential problems to be solved. This is our offer and our challenge: if we offer scientist-founders our very best, right at the very beginning, what bright future will they create?