The complete guide
to spinning out of academia
III. Patents
IV. Tech Transfer
VI. Incorporation
VII. The License
VIII. Negotiations
IX. Closing Thoughts
In academic labs right now, researchers are inventing solutions to the climate crisis, disease, malnutrition, connectivity, and a host of other important problems. Commercializing this science is one of the fastest paths to solving the world's biggest problems. Unfortunately, spinning technology out of a university into a startup is a murky process, limiting the number of impactful companies created. At Fifty Years, we’ve backed dozens of university spinouts and have advised hundreds of scientist founders. The main element of spinning out – tech transfer of IP from academia – has been improving, but it’s far from solved. We’ve seen negotiations with tech transfer offices (TTOs) take so long that founders burned out, teams lost momentum, and investors walked away. We’ve seen TTOs introduce terms that hurt the startup’s ability to raise capital. We’ve seen so many exotic term combinations that we wondered: how are founders supposed to make sense of it all? So we made this Spinout Paybook to help aspiring scientist founders make sense of the process.
Starting any company is hard. Early stage startups live and die by the energy of their founders and their speed of execution. We hope this playbook can save your energy, help you move faster, and avoid some mistakes along the way.
This Playbook is meant to be a living document. There are some solid efforts by TTOs, VCs, and major law firms to make university licensing more transparent and rational as well as to improve the negotiation process for everyone involved. However, there still isn’t a single great resource for prospective founders with advice that maps to their experience. We hope this Playbook becomes that resource. Please send comments and improvements to spinout@50y.com.
We focus on U.S. spinouts; some advice might not translate directly to international companies. If you’ve already spun out, please fill out our anonymous survey and help us level the playing field in founder <> TTO negotiations.
You're in academia, working on the frontier of science. You've made an exciting discovery that could positively impact millions. Should you spin out? Even if the invention is incredibly exciting, it doesn’t mean it will translate well to a product. Even if you have a revolutionary product in mind, there might be no market for it. Or maybe the product and the market are there, but that’s just not a company you’d be excited to run for the next decade.
Before you decide:
Ask yourself: Will this technology advance faster in academia? This may be true if you need specialized equipment or resources already (or only!) available in an academic lab. Or you might have excellent grant support that's enabling you to make efficient progress.
Define the product and describe the market for it. Name the need (problem), the persona of a customer, how the product would be sold and delivered, and a rough idea of the customer’s willingness to pay for the product. These will evolve over time, but it’s a useful exercise. Too often, we see spinout founders trying to sell inventions instead of products – or “products” with no real markets. Ask what problem your technology solves for what potential customers. Never assume something “obviously has a market.” Price matters, distribution matters, perception of value matters, ease of use/administration matters, manufacturability matters, serviceability matters, competition matters, and regulatory environment matters.
Can your invention lead to an at least 10x improvement in how things are done today? Whether you’re going B2B or B2C, there’s some existing solution people are using today instead of your product. Don’t underestimate switching costs: actual $ cost of switching, perceived risk of using anything new (and working with a startup), as well as habitual or emotional attachment to how things are currently done. Because of switching costs, incremental improvements are often not enough. Small advantages are also likely to be competed away by the time you’re ready to sell – i.e. someone will deliver marginal improvements in some other way before you do. It’s important that your potential solution be significantly better than what currently exists: at least 10x cheaper, faster, easier, better, or ideally all of these.
Talk to potential customers. You have assumptions about what your customers want, why they want it, and what they’re willing to pay for it. Your assumptions are likely wrong! Test your hypotheses by talking to real potential customers. Even products as obvious and needed as a “cure for cancer” need to be marketed – to pharma companies, to doctors, or directly to consumers. When you’re talking to customers, do they mention the problem before you do? They should! For your product, try to understand: How painful is the problem you’re trying to solve? How will your customers learn about your solution? Why would they choose it over an alternative? How much are they willing to pay? While you’re at it, it’s helpful to constantly iterate on the narrative of your solution/company to see what “clicks” with specific partners.
Study companies and founders in your space. Why and how did they succeed? Why and how did they fail? What did they learn from their customers? What was most surprising to them? If you can, talk to successful and unsuccessful founders and operators in your space directly – many learnings are not publicly shared nor accurately represented in the media. You might learn that your approach was tried and proved unfeasible before – you might shortcut many years by not repeating the same mistakes. Finally, later when you spin out, you’ll appreciate having mentors to talk to!
Is the market big? Solving big problems can create big financial opportunities – but it’s not a rule. You can get pigeonholed to a small market, as many medical device companies do. Market reports are okay for market-sizing but a first-principles analysis is better – e.g. “There are ~500 enterprise buyers for this in the world; if each spent $500K annually with us, we’d make $250M per year in revenue.” This can be very back-of-the-napkin. Does the size of the opportunity motivate you? Would you need to get VC funding to spin out? There are plenty of ways to fund your startup: early partnership revenue, grants, angel investors, family offices, etc. VC funding generally only makes sense when you see a path to $100M or more in annual revenue.
Who are your co-pilots? If startups are made of people, spinouts are made of people and a sprinkle of IP – the team is still the most important ingredient. Do you want to start the company with other folks who did the research? Do you have any friends you’d consider as co-founders? Do you still need to recruit co-founders? This is the single most important decision you’ll make. Some tips on this later!
Are all co-founders ready to leave academia? You might think you can do many things at once. In 99% of cases, you will be hurting your startup. You can work from your academic lab in the beginning, benefitting from the free space and other resources – but only do this if it’s the fastest way to move your startup forward. In the vast majority of cases, making a clean break and dedicating 100% of your time and effort to the startup is essential for success.
What’s the impact your company could have? If all goes well, how will the world be different? The most successful founders – and their teams – are energized daily by the problems they want to solve and the vision for a better world they can co-create. Start to flush this out.
Are you excited to work on this for the next 10 years? You should be; it’s a long journey!
An alternative to spinning out:
Another company could potentially commercialize your invention by buying or licensing the technology from the TTO. If you’re one of the inventors, the TTO would likely be splitting licensing revenue with you. While this sounds plausible as a route to commercializing, most technology never gets licensed. TTOs have limited resources to shop the IP around, and most IP is just not valuable independently of a fast-moving team packaging it into a marketable product. Moreover, large companies often license tech defensibly – with the intent to just sit on it without ever using it (drug IP is a notable exception here). If you care about the broader impact your invention could have, you’ll want to commercialize it yourself.
The team, not the invention, is the true foundation of your company. If you worked on the initial research with other folks you think are both brilliant and can move very fast – they are your natural candidates for co-founders. If you worked on your own, you might be tempted to go as a solo founder.
Do I need a co-founder?
No, but it sure helps. If you can find folks who energize you, who you communicate well with, who you’re vision and values-aligned with, and who are ready to share the emotional burden of running a company, that’s an incredible asset. And if they have complementary skills, knowledge, or networks to boot, that’s fantastic. In the spinout database we’re sourcing (contribute here), the most common configuration is 2 or 3 co-founders.
Looking for a co-founder
Co-founder breakup is a common reason startups fail, so you’re screening just as much for technical skills and past evidence of speed in execution as you are for communication skills and how you get along (chemistry). Ideally, your co-founder is someone you’ve known for a long time and possibly worked with – at another company, on a research project, a Burning Man art car, or something similar. Second best is someone recommended by your smart and hard-working friends – someone whom they’ve actually worked with.
Step 1: Define “perfect”
Identify what’s important for the startup to succeed and compare that to what the current co-founding team has. Write down a persona of an ideal candidate.
Key areas to focus on:
Complimentary skills or experience. Who’s world-class at the skills you need? What would the perfect experience look like? What research have they previously done? What have they built previously? Which companies did they work at? This point is mostly about identifying perfect experience but do write down names of companies, schools, and labs these people might have been at – this will make it much easier to reach out to your network and actually find them!
Unfair advantages. Who has free access to the research facility that could de-risk your startup? Who has deep ties to customers and could guarantee meaningful commercial traction? Who has a big following among the kind of people you’ll need to recruit?
Personal qualities. What qualities would you want to see in a co-founder? These are as important as skills and experience. Write them down.
Mission alignment. Who’s obsessed with solving what you want to solve? Who cares about the same problems you do?
Step 2: Prepare to sell
Great people are almost always happily employed and not looking for jobs. Get ready to sell them on the mission of the company, the exciting technology you have, and why you’re the best person to work on this with.
Create and practice your Story of Self. Learn to explain in 3-4 minutes who you are, what your values are, what drives you, and why this company is the most important thing you need to be doing right now. Get feedback from friends, co-workers, old bosses, and family members about how inspiring and authentic your narrative feels. Every time you meet a new candidate, share your story of self and only then talk about the startup.
Vision and mission of your startup. Learn to explain succinctly what you want to do, the mission of the startup, and the vision for the world where you succeed. It may feel weird to share the big vision of your tiny little company, but potential co-founders will resonate with it.
Step 3: Look for potential co-founders in your network
Take the candidate personas you developed in Step 1.
Go through all of your Gmail/LinkedIn contacts and identify possible matches.
Do a “memory palace” exercise where you go through every phase of your life to find people to talk to. Specific circumstances to run the “memory palace” questions for: conferences, seminars, on-campus clubs, and online communities.
Reach out to the most impressive people you know, share your idea and how it relates to your story of self (see Step 2), and ask for recommendations and introductions to people you should talk to. “Who’s the best co-founder for me?” is a hard question. Instead, ask for the best people in field x, in an attribute (e.g. “most grit”), or whom your contact “would love to work with.”
Send forwardable emails to your network. Be short and clear about what you’re asking for. Make it as easy on your contacts as possible to just forward. It will lead to more introductions.
Step 4: Meet, then work together
Meet them in person. Learn about their background and if they’d be a good fit for the startup. The best conversations will leave you feeling extremely energized about the startup – and each other. If that’s the case, you can suggest answering a set of “founder questions” like the ones on this list. The next step is to see what it’s like to work together. You can engage them in building a prototype, writing a grant, or doing some other useful task. What you want to see here is creativity and extreme ownership of moving things forward fast. As you get to know them, consider their:
Competence: Does their professional background suggest they’re amazing at the skills the company needs to succeed? Do you like the way their mind works? Do they incorporate new information fast? Do they have a growth mindset?
Values: Do you understand their values? Do they share yours?
Vulnerability: Do you connect on a personal level? Are they comfortable talking about past failures? Did they candidly share their doubts?
Communication: Do they quickly grok your ideas without much miscommunication? Do they ask great questions? Is coordination easy? Do others seem to understand what they’re saying? Do they take critical feedback well?
Energy: Do they energize you? Do they move fast? Do they accelerate progress? Can you imagine that working with them will be fun?
Focus and commitment: Are they at a point in their lives where they can make the company one of the most important things in their life? A company will inevitably compete for time with family, hobbies, and other life goals. They might need to take a serious salary cut in exchange for high-risk equity. Are they actually able to commit to the startup life?
If things are looking good, move to references. You can interview them using a structured framework (we use the “Who” method) and conduct personal and professional references (we’d recommend 7 or more).
Step 5: Stress-test the relationship
This step might be controversial but it can speed up years of more “natural” getting to know each other. The goal is to determine whether you still get along in adverse situations. If you both enthusiastically agree to do this, you’ll want to find a situation or a challenge 1) where you both need to work extremely hard, 2) that has high physical demands or stresses the body, and 3) have it last longer than 24 hours. You might think “can I really afford the time to do this?” but absolutely nothing will lose you more time in the future than a broken co-founder relationship, so it’s worth it.
Ideas:
Apply for a grant in an extremely short amount of time. Make it so both (or all, if there are more than two) potential co-founders need to work long hours (>16-hour days) for 3-4 days.
Go on an extreme backpacking trip together or plan another intense experience like living together for a few weeks.
An example from a 50Y portfolio: two potential co-founders created an art car and went to Burning Man together. This worked on multiple levels. Both still had full-time jobs, so all the work was done after hours. The work involved intense physical labor and time pressure. It helped them see how their future co-founder behaves when they are tired and uncomfortable and how they approach decision-making and execution.
Step 6: Decide on the equity split
Typically, equity should be split equally between full-time co-founders. If you aren't willing to give your co-founder(s) an equal share, then perhaps you are choosing the wrong co-founder(s). Y Combinator has great advice for splitting equity between co-founders. They highlight some of the most often cited – and generally wrong – reasons for unequal equity splits:
“I came up with the idea for the company.”
“I started working on the project before my co-founder.”
“I brought on my co-founder after raising thousands of dollars.”
“I invented the core technology behind the company.”
Startups are about execution, not about ideas. Almost all value in a startup is created through years of execution: technical de-risking, building the team, building the products, getting commercial traction, etc. Small contribution differences in year one do not justify massively different founder equity splits in years 2-10+, when most of the value is created. Finally, the team will likely regret the unequal equity in the long term. An uneven equity split that feels justified in year one won’t feel justified in year five. Even if your relationship and your co-founder’s commitment seem very strong today, unequal equity can lead to resentment, especially when you’re both pulling many all-nighters to get something done. You’re both working equally long days, but one of you is vesting more equity with each hour – it’s easy to see how that wouldn’t feel right. It is much harder to rebalance the equity down the road – not to mention there can be serious tax implications when you do – so better to make it equal up front.
What’s a reasonable explanation for an unequal equity split?
Rarely, an unequal equity split makes sense. It’s most common when a co-founder who significantly contributed to the IP is not joining full-time and staying in academia instead (e.g. a PI). They usually get referred to as a “scientific co-founder” or an “academic co-founder”. Academic co-founders should get significantly less equity – total equity for all academic co-founders should be 15% or less. Typically an academic co-founder would get ~5%, with up to 10% could be acceptable for those that will remain very involved part-time. Importantly, this equity should still be on a 4 year or longer vesting schedule with pre-agreed milestones for vesting or time commitments (e.g. 10 hours per week). All other co-founders should join full-time – and only vest equity if they do.
Another rare reason to split equity unevenly is when one founder backed the company financially. For example, they self-funded the startup with $100K (generally not just to pay their own salary) and they are “converting” that investment into some amount of additional common shares at a valuation that an outside investor would have backed the company at. In this case, simply do the math to decide on how much additional equity they should get for their cash investment into the company. For example, say they put $50k of their savings into the company to buy early equipment a few months before the company raised money at a $4m valuation. Here the math is $50k / $4m = 1.25% extra.
Separately, investors want only full-time co-founders to have meaningful equity. Academic or scientific co-founders who maintain full-time roles outside the startup should not get “full-time” levels of equity. That equity should be given to someone else who will contribute daily to a startup’s success or failure. Between 5% to 10% for non-full-time founders is normal with the lower end (5%) being healthiest for the company. More than 10% is troublesome. Everyone, including academic co-founders and advisors, should be on a 4+ year vesting schedule linked to employment or other specific daily/weekly/monthly contributions.
Should I hire an experienced CEO?
If you want to run the company then no, you absolutely shouldn't. You won’t be able to hire someone amazing who also cares as much as you do about the company. The best CEO at this stage is one of the scientists who did the original research and who’s excited to commercialize it. Sometimes, it can be a co-founder you bring on, especially if they contributed to the technology already. You should care less about experience and more about extreme hunger, ability to learn, growth mindset, speed of execution, and solid communication skills. It can be a scary prospect to be a CEO but the good news is that you’ll have the chance to learn a lot by doing! One example: the CEO’s communication skills will need to become great fairly fast as CEOs are constantly pitching: to investors (fundraising), customers (selling), and potential employees (hiring).
Many scientist founders think “but I don’t know how to be a CEO!” Don’t worry. No first-time founder had the skills, experience, knowledge, or network they needed to be a great CEO when they founded their company. This includes the most successful of all time – Jeff Bezos, Steve Jobs, Bill Gates, Mark Zuckerberg – all first-time founder CEOs! They didn’t know how to be great CEOs when they started. They learned along the way and you can too.
Many founders conflate the question of “should we hire an experienced CEO?” with “should we hire an experienced executive team?” All the industry experience, know-how, and intuition the company needs can be brought in at the executive level. The ability to convince industry veterans to join the team (typically post-Series A) will be critical and founders can complement their passion and technical expertise with the experience the company needs. No need to hire a CEO.
What about MBAs?
Some accelerators push PhD founders to immediately team up with MBAs they just met to compensate for a lack of “business skills”. It’s much better to start a company with someone you've known for a long time or worked with before. All-PhD founding teams work perfectly well if at least one of the founders is excited to run the non-scientific part of the operation and grow in that role. Besides, most MBA programs teach best management practices for much later stage companies – which is great and needed but does not translate well to early stage startups.
There will be, however, time to build out your business development (BD) team and it can be a good idea to recruit BD advisors or hires with relevant industry experience early (e.g., 5+ years in BD in the industry selling similar products to the kinds of customers you will sell to). Having industry veterans on the team is a good signal for potential partners and investors alike and can be great for the company.
Overall, don’t start a company with someone just because they have an MBA. Don’t make someone a CEO if you don’t also see them as a true co-founder and are excited to work with them for many years through all ups and downs of your company.
What’s the right time to leave academia?
You should be moving as fast as possible to build your startup. If that means conducting derisking experiments in an already set-up lab at the university, do that. But at some point, you will move faster outside of academia. There are a few reasons:
Focus. Spending time and attention on tasks for the university is time and attention not spent on your startup.
Space and equipment. At some point, it’ll make more sense to set up your own lab with your own facilities for exactly what you need to be moving as fast as possible to de-risk the technology. University labs aren’t usually equipped to go at that speed or give you the space to expand.
Team. You can’t build a team inside an academic lab.
Culture. Most academic labs move at an academic pace. In your early days, move as fast as possible. It’s nearly impossible to build that culture inside a slow-moving, intentionally unfocused environment.
Customers and partners. If you have something worth commercializing, you want to talk to potential customers. There are many risks when working with a small team, and the risks seem even higher if it becomes clear you’re still one foot in academia. Why would someone work with you if YOU are not ready to commit to your startup?
Investors. For similar reasons to customers, most investors will want you to have left academia by the time you talk to them – or at least have a clear plan on when you’re leaving.
In our experience, most founders say they left too late rather than too soon!
The invention might feel like “yours”, but you likely signed an agreement as a graduate student, postdoc, or staff scientist that transfers all rights to any intellectual property developed at the university to the university. If you want to commercialize your invention, you will likely need to work with the university to (1) patent and then (2) license the invention.
What is a patent?
A patent is a right to prevent others from making, using, selling, offering to sell, or importing what’s covered by the patent. The U.S. government provides this right in exchange for the ability to make the patent public. All patents that have ever been published are searchable. Patents are made up of two sections – one describing the invention using text and diagrams, and then the “claims” section, which defines exactly what the patent covers and doesn’t. Legally, the claims section is the most important part of a patent.
There are three types of patents:
Utility – covers process, machines, manufacturing, or composition of matter
Design – covers the unique visual aspects of a product
Plant – covers any “invented or discovered and asexually reproduced distinct and new variety of plant”
We focus on utility patents as these are most commonly used in early stage spinouts. Utility patents protect for 20 years and the clock starts from the earliest filing date for a non-provisional patent.
Is my invention patentable?
Your invention is patentable if it fits these criteria:
Useful: the invention has a useful purpose.
Novel: there is no existing patent claiming this invention, published documents, or public use of your invention (previously-known work related to your invention is called “prior art”).
Non-obvious: there is a sufficient difference from what has been used or described before. A person who has ordinary skills in the invention’s technical field would not find it obvious to make the change.
Disclosure impacts patentability
If you’d like to file a patent, there are rules around disclosing details about the invention. In the U.S., you have one year after public disclosure (e.g., presenting the invention at a conference or publishing on bioRxiv) to file a patent. “Public disclosure” is defined as sharing enough information that another person, skilled in the art, could recreate the invention. Rarely, even a single tweet could get you in trouble!
If someone disclosed the invention and the university hasn’t applied for a patent one year after the initial disclosure, it’s too late to file a patent. But do your own research – some universities (e.g., Stanford’s Office of Technology Licensing) have policies to not patent inventions that have been disclosed even if the 12 months have not passed.
Is the patent useful?
Not always. Patents do two things: 1) protect you from being sued by contesting parties for using your technology and 2) block other people from using your technology. There are scenarios where you may actually consider strategic public disclosure to make the technology free to use by everyone (including yourself). Why would you do that? For (1): if there’s no patent filed and the technology is publicly disclosed, others can’t sue you for using it. For (2): if others are incapable of using the technology (e.g. additional IP or trade secrets are needed to use it), then there’s no business advantage in having patents. Note that even in the case where there is no patent, the university may still own trade secrets or other intellectual property derived from your work, and therefore you may still need to license the technology from the university.
Some scenarios in which strategic public disclosure – essentially putting the IP in the public domain – might make sense:
The business model relies on broad adoption of a technology and its strategic public disclosure allows for free adoption of said technology;
Quick and immediate prevention of competitors getting infringing patents – often you can share much more in a strategic disclosure, eliminating the ability for others to develop infringing patents.
Some patents aren’t protective due to issues with discoverability. Imagine patenting a new growth media formulation to differentiate stem cells into T-cells. If the formulation doesn’t leave any trace in the final T-cell product, it’s impossible to know if another company is infringing on the patent without going to their lab and chemically analyzing their media. By patenting, you have now revealed the formulation (all patents are public) and gave your competitors deeper knowledge of your technology. Instead, you could keep this information a trade secret instead – though the university would likely still own it and might have claims to it.
When there’s no business advantage or a patent isn’t enforceable, save yourself some time and skip patenting.
What if the technology is not fully developed yet?
If you’re still iterating on your invention, you can file a provisional patent to protect your work. You then have 12 months to convert the provisional to a non-provisional patent. As the technology gets fleshed out, you can add details to your provisional patent over that 12-month period. Why would you want to file a provisional at all? Because the date you file the provisional determines the priority date for the patent (i.e. who patented this first?), so sooner is better.
What if I made the invention outside of my graduate or postdoctoral work?
This varies state-by-state and depends on other factors. For example, CA, DE, IL, KS, MN, NC, and WA have laws that assign IP to employees (incl. academic employees) when they’re working on their own time using only their own resources (more here). In those states, generally speaking, if you never used university resources to develop the technology (including things like university wifi) then the university has no rights over your IP.
You want to refer directly to the invention assignment agreement you signed. Here’s Caltech’s, which specifically calls out that, “when [you] make inventions, write computer software or other written work entirely on [your] own time, without using Caltech’s equipment or facilities […] and without using Caltech funds, [you] retain ownership of those inventions, software, and written work.” If you are sure the university doesn’t have an ownership interest in your IP, you can contact your university tech transfer office (TTO) to get a formal letter stating they have no claim. It’s worth doing that to have proof in case future investors, acquirers, or other entities want to know you’re legally free and clear to sell/license your technology.
What if I improved on the technology after I left the university?
Once you’ve left and if you used no university resources, the university has no claim over that additional IP (assuming there’s no agreement in place that assigns improvement rights – see below). You can file and wholly own those improvements or keep them as trade secrets.
Do all universities own the IP created by students, postdocs, and PIs?
No. Some universities, like the University of Waterloo, are famously founder-friendly. Their "creator-owned" IP rights policy grants ownership of all IP to the inventor, not the university. This means that anyone is free to patent the inventions they made. However, Waterloo can be electively involved in tech transfer via their Commercialization Office (WatCo). Their website lists clear & simple terms: the inventors will continue to own the IP, but they can assign the right to WatCo to commercialize on their behalf in exchange for 25% of profits. There are no other fees or terms involved, and WatCo pays all the patent filing fees and processing costs.
If you’ve decided a patent is useful, and your invention requires the university to patent it, you’ll now need to work with the university tech transfer office, or a TTO (sometimes called other names, e.g. the Office of Technology Licensing). TTO’s responsibility is to file, license, prosecute, and maintain patents for the university. You’ll be working with them to 1) patent your invention, and 2) license it back to your startup. This process definitely requires careful preparation.
Start by talking to your PI (assuming you’re not the PI!)
Before approaching the TTO, you should talk with the professor running your lab (PI). TTOs often defer to PIs on decisions on whether to patent something and whom to license it to. Their buy-in and support are important, but there are limits to generalizing advice here. A lot will depend on the specific PI, the relationship you have with them, the way they operate their lab, their relationship with the university and the TTO, etc. You definitely want to talk to other founders from your lab who tried spinning out and learn from their recent experiences. As general advice, you should frame the first conversation with your PI by saying you want to commercialize, sell them on the vision of the company, and why you’re the right person to do it. Depending on what you want, ask them to become an advisor or scientific co-founder, and describe what it would mean in terms of responsibilities, equity, and vesting. Unless your PI is leaving their academic position, like Carl Hansen at AbCellera, they should have significantly less equity than full-time founders (more on that later).
The PI might be excited to hear the news or they might have a very different plan in mind for commercializing the technology (like bringing in a seasoned CEO or licensing to a big company). If they’re unfamiliar with startups where the inventors also build the company, you should have them talk to another PI, investor, or founder who is familiar with this model. Give examples of very successful companies started this way, like Ginkgo Bioworks, Apeel Sciences, Dyno Therapeutics, or Commonwealth Fusion.
Notify your university they should file a patent
TTOs get notified of inventions via Invention Disclosure Forms (IDF). You can find the specific form you need to use on your university TTO’s website. TTOs use IDFs to decide whether to patent an invention and whom to work with to do so.
Key data to gather for an IDF:
Description of the invention
The novelty of the invention
The problem the technology solves and how it solves it
Advantages over current technologies
The PI whose lab it came out of
Conception date
Disclosure date (if any)
Full list of inventors
Once you file the initial IDF, the TTO will reach out for an initial meeting and ask for more information. The following questions are common, so be prepared to talk about these in a convincing way:
What is the invention?
What’s the novelty of the invention?
Can the technology be used to develop more than one potential product? Is it a platform technology?
What’s the unmet need/problem that the invention could address?
What are the advantages of the invention over competing commercially deployed technologies/products?
List all people who contributed to the invention's conception and/or development.
Do you have a summary of contributions from each person involved?
Have you gathered/submitted all drafts, manuscripts, presentations, and evidence related to the invention?
Have you avoided public disclosure before submitting an invention disclosure to your TTO? If not, share details of the disclosure (date, platform, etc.)
Do you have a plan for this IP? => This is when you tell the TTO that you are considering spinning out, so they can license the IP to you. Sharing early helps you as TTOs are incentivized to find a licensee and they prefer licensing to university-affiliated startups.
What are the next steps? What’s required to generate a proof-of-concept (POC) or gather experimental data? How long will it take? How much will it cost? Will it require any more university resources? Are additional patents needed?
What’s your current thinking about the team to take this to market?
Have you received interest from potential investors or industry partners?
Can the university decline to patent my invention?
Yes. They might think:
It’s not patentable. Some software, methods of doing business, and algorithms are just not patentable. There’s not much you can do here.
It’s not worthwhile to patent – or not worthwhile to patent as broadly. A cost-conscious TTO might decide that generating revenue from this particular patent is unlikely. If this is the case, a call from a PI asking the TTO to file a patent, a stronger proof of commercial interest, or a better explanation of the business you want to build on top of the IP could change their minds. In the rare case that they still don’t want to file, you can ask the TTO to transfer the IP rights to you so that you can file for the patent personally. Whether they’ll agree depends on who funded the invention (who might want to file themselves) but if the university and research funders are not interested, in rare cases TTOs will agree to let the inventor patent the IP personally.
If a TTO declines to patent an invention, you can ask that the IP be transferred to you so that you may pursue patenting it yourself. The TTO will need to satisfy any obligations over the IP arising from the funding source before transferring. Once the TTO has transferred the IP to you, you can file a patent on the work.
TTOs will often start marketing your invention to other potentially interested parties after the provisional patent is filed. They do this to ensure “fair and open access” and to test commercial interest before the filing of a non-provisional application. They might contact specific people at target companies or “shop it around” through public databases, like Stanford’s TechFinder. This can be nerve-wracking (“oh no, someone can take my baby”) but the vast majority of patents (from 70% to 95%) are not licensed this way. Most good TTOs in the U.S. also strongly favor licensing to their own spinouts. So time to negotiate!
Negotiations with the TTO
First, find out who can lead the negotiation. Many universities have conflict of interest policies that don’t allow students or employees to negotiate with them while still at the university. If you’re not sure, contact your university’s conflict of interest (COI) office to confirm if you can negotiate directly. If you can’t, you’ll either need to drop out (so you’re unaffiliated) or find someone to negotiate the financial terms of the license. This could be a co-founder who is unaffiliated with the university or a third party willing to negotiate on your behalf (lawyers, advisors, or investors). Sometimes you can incorporate while still at the university and set yourself up as a consultant under terms that the university is okay with (e.g. this is possible with Harvard).
There’s a perception that law firms can be misaligned because they care about a good future relationship with the TTO (and they do!). But great law firms also care about their reputation with startups. Moreover, licensing agreements are not where law firms make most of their revenues working with startups. They want the company not just to survive, but raise a lot of money and do well so the company can drive much more business for them in the future. Great lawyers can be extremely helpful but also costly, especially if billing hourly. The extra layer of communication (startup ⇔ lawyer ⇔ TTO) can slow things down and add unnecessary adversity but ultimately will give you extra peace of mind.
Sometimes lawyers will be hesitant to “play hardball” with TTOs in order to preserve their relationship with the TTO. In those cases, founders should push their lawyers to play hardball.
If you hire a lawyer, look for:
impeccable referrals from founders they’ve worked with
experience with your TTO
a database of comparable recent licensing agreements (it helps in negotiations)
a fixed fee and, ideally, a deferred payment
someone who wants to build a relationship with you long-term so they are incentivized for the company’s long term success
Before signing up with a law firm, ask them whether the university is already their client – and, if yes – how significant. Law firms might be more lenient with TTOs if they do a significant amount of business with the university.
You’ll also want to confirm with the COI what roles you can play in the startup (at least on paper) and what other actions you need to take before spinning out. For example, as a graduate student at Stanford University, you’re allowed to start a company but aren’t officially allowed to take an executive title until you graduate. But as a postdoc at Stanford, you’re not allowed to be involved at all without approval from a few deans and department heads. Other universities have their own rules.
Prepare a business plan for your TTO
Writing a detailed business plan is generally not very useful, but TTOs often require one.
In the business plan, make the markets and applications as broad as possible. This will be important in the negotiations over fields of use (more on that later).
TTO business plan summarizes the technology, initial and future products, market, team, fundraising strategy, and timelines. Many TTOs have their own templates and requirements. For example, Stanford’s Office of Technology Licensing office lists these main components for a business plan:
Company name
Mission statement (a guiding vision for the company)
Current market situation – How big is the market? What are its critical problems and shortcomings? How is the landscape changing? Who is the competition? Is it a consolidated or fragmented industry?
The company’s solutions – Which products or methods will be developed? How long will it take? What are its applications? What are the company’s unique advantages and are those advantages sustainable? How will the current market change due to the company’s products, methods, etc.?
Patent/IP landscape
Marketing and sales strategy – Pricing, Product, Placement. How will the target market know about the product? Which sales distribution channels will be used?
Five to ten-year strategic/financial plan:
– Financial projections – When will the company break even?
– Key milestones required to meet financial projections
– Key metrics to be measured and tracked
– Key assumptions and how they change based on a competitor’s responseFunding requirements
Management team – members with resumes and roles
Timeline and key milestones
Risk factors and mitigation measures
You can see an example TTO business plan here.
You’ll generally need to incorporate before you can license any IP. This is because the TTO needs an entity to negotiate with. The time to incorporate is when (1) you know you have an idea you’re excited to work on, (2) you have a clear plan to spinout, (3) the founders and the PI agreed on the equity split, and (4) the founders are ready to jump full-time or have a very concrete timeline to do so. If your startup depends on licensing, you’ll also want to have (5) a verbal agreement that the TTO’s licensing officer intends to work with you to negotiate a license or a license option (more on that later).
A common misconception amongst first-time founders is that incorporation is complex. In reality, once everyone agrees on the equity division and roles, it should only take an afternoon. The best online tools to incorporate are Clerky and Stripe Atlas. They handle all the paperwork with the Secretary of State of Delaware, which is the state where you want to incorporate even if you’re physically based somewhere else. You don’t need a lawyer to start a Delaware “C Corp” but – for personal tax reasons – do remember to file your 83b election with the IRS within 30 days.
Still, some founders choose to incorporate with the law firm they’ve picked to lead the negotiations with the TTO. That’s a fine alternative but will be more expensive.
Before you enter into negotiations with the TTO, consider:
Not licensing at all. Consider how difficult it would be to recreate the technology using different methods inside the startup and compare it to the cost of licensing (equity, royalties, other fees), the time you’ll need to spend on the negotiations, and how draining it can be. Maybe you can save time and just dive into work? Startups often overestimate how useful their patents will be and often regret wasting time on them later. Never underestimate the importance of your energy and your startup's speed of execution.
Getting an option to license instead of a license. With an option to license, the university agrees not to license the technology for 6-24 months at which time you can enter into proper licensing negotiations. This is useful when you expect you might pivot or if you’re not sure you will need a full license. It also saves you time and money and doesn’t require lawyers to be involved.
Whether you need an exclusive license. Non-exclusive licenses cost less and are typically negotiated faster. Exclusivity is useful if you expect that competitors also want to license the technology and that it would hurt your ability to build a valuable business if they did.
Which patents you will need. If you know that you’re going to need multiple patents, bundle them. Negotiating for three patents together can take just as much work as negotiating one!
Once you decide to license, the TTO will send you a term sheet with major items to be negotiated. You can also send them your own term sheet to speed things up. Here’s a template term sheet we sourced from Stanford which you can copy and edit.
How long will this process take? We hope that in the future it can take weeks or days! For now, from the data we’ve seen, 2 or 3 months are not uncommon with top schools (Stanford, MIT, Harvard, Northwestern) but the same TTOs can take up to 12 months. We also saw total outliers where tech transfer took 18-24 months! It’s important that you don’t let the slowness of the process (1) drain your energy and (2) slow your startup. By all means, keep working on your startup while you’re negotiating your licensing agreement. It’s also important to stay aggressively on top of the process to maximize the chance your negotiation falls into the 2-3 month bucket and not the 18 month bucket. You shouldn’t let more than 2 days go by without asking your lawyers for an update. If you have a direct relationship with the TTO, try to get as much facetime as possible and let them know why it’s important that things move quickly. Err on the side of being annoying!
Main terms and concepts:
Fields of use: Restrict the use of a patent to particular purposes. Your goal should be to get the TTO to grant all fields of use for any application described in your business plan.
Exclusive license: No entity other than your company is allowed to exercise the license in the predefined fields of use. This is the preferred license if you believe other companies could realistically use the technology. If you aren’t sure if you need exclusivity, you can sometimes start with a non-exclusive license with an option to make it exclusive later on.
Non-exclusive license: Grants you the right to use your IP in the field of use, but the university remains free to grant any number of other licensees the same rights to make, use, or sell the technology in the same field of use. This can be OK if the IP is only very valuable with some other IP or trade secret that only you possess.
Royalties: Payments of a fraction of your revenue or profits coming either from the use of the licensed technology or the entire company. Of all terms, universities care most about royalties and success-based milestones payments as they represent the vast majority of their revenues. This is demonstrated in the below chart from this report, which shows revenue by year from licensing royalties vs equity:
In some situations, TTOs will trade these cash payments for equity but that’s relatively rare. Normal ranges for royalties are 0-5%. Ask for clauses to terminate the royalty after a few years or after paying out a cumulative $ amount. Royalties should be tied directly to the use of the licensed technology, and not to overall company revenue.
If your royalties can’t be 0%, this is what you can ask for:
Cumulative payment ceiling (e.g. royalties go away once you pay $5M)
Timeline for royalty phaseout (e.g. royalties go away after 4 years)
Royalty stacking, which allows you to pay less to one licensor if you need additional licenses from other entities (typically 50% of costs can be discounted or “offset” if you’re using that to pay a second licensor)
For “combination products” – where your technology is an improvement to an existing product that isn’t part of the licensing agreement – get the royalty to be applied only on the “enhanced value” (e.g. some tech makes an MRI machine sell for 10% more => the royalty only applies to the 10% increase in revenue)
Also, if you can, confirm that there are no clauses related to acquiring entities that would apply royalties to the acquiring entity’s entire revenue – these could prevent an acquisition you’re excited about from taking place. Example scenario: BigCo is considering acquiring NewCo. NewCo has a clause in its licensing agreement in which royalties take up 3% of revenue. There’s also a clause that states that, in the event of an acquisition, the 3% royalty is applied to all revenue of the acquiring entity. BigCo decides not to acquire NewCo because instead of having to pay 3% royalties only on NewCo’s revenue, they’d have to pay 3% royalties on everything BigCo makes. These clauses are added by universities to force the BigCo to re-negotiate a royalty agreement with the university at the time of acquisition. But because they add time and introduce uncertainty they can simply kill the acquisition deal instead, and can make raising future rounds more difficult.
Arguments to lower royalties:
It’ll take lots of $$$ to commercialize. University IP is incomplete. You’ll have to raise millions of dollars to bring something to market. Funding can be >$500M for companies in the chemical synthesis or fermentation space. For a biotech drug, this can be ~$1B to get a drug through clinical trials. Agreeing on high royalties now doesn't make sense given the massive investment still required to commercialize. Be specific here and talk about your particular situation.
Royalties make it harder to raise money. Royalties lower profits, making the potential return lower for investors. High royalties make the startup less attractive to funders, which might lower the chance the company generates large revenues over time by depriving the company of the capital it needs to grow.
Competition will lower margins. Competitors will almost certainly enter the market and chip away from your assumed profit margins. High royalties on topline revenues can eat away all your profits if your margins are low and risk the sustainability of the business.
A note about royalties: Royalty rates are highly dependent on industry but the median royalty rate for university licenses is around 3%. This is skewed because the most active fields for licensing are (1) Pharmaceuticals, (2) Biotechnology, and (3) Healthcare Products and Supplies, which tend to have high margins and therefore higher acceptable royalties. For biotech, royalty rates also depend on whether you are building a discovery platform (typically 1-2%) or a single-therapeutic asset company (typically 2-4%). Finally, in the case of truly transformative technologies, like CRISPR, royalties can reach up to “high single digits”. Editas, for example, which licensed CRISPR technologies from the Broad Institute, paid high single digit royalties. In addition to the higher royalties, the Broad Institute negotiated for 4.2% of total equity and $14.8M total milestone payments per therapeutic product.
Equity: TTOs will usually demand ownership in the form of common shares in the company. This is true in >90% of licenses we’ve seen in the U.S. The typical range is 1-5%. Above 10% is bad and can make raising from investors more difficult. Many VCs view universities as “dead weight” on a cap table, since their contributions were in the past and are not forward looking (i.e. they don’t help the company raise future financing, hire talent, provide strategic advice, etc.). VCs prefer as much equity as possible to be held by people who will actively help the company succeed (primarily founders and employees, and then secondarily engaged investors and advisors).
Pro-rata rights on equity: TTOs may ask for the ability to maintain their % ownership through additional cash investments. Any pro-rata rights should only apply to priced rounds and the university should be required to confirm their participation within two weeks or so.
Anti-dilution: TTOs may ask for provisions which grant them additional equity for free in order to maintain a percentage of ownership. This is commonly to protect TTOs against founders signing a licensing agreement and immediately diluting the university by issuing more equity to only the founders. Once there are outside investors on the cap table, the risk of founders doing this goes down greatly, hence the TTO asking for anti-dilution until the first round of financing. Founders should either negotiate away anti-dilution or limit the provision to the TTO maintaining a specific % ownership up to the company raising at a reasonable valuation (e.g., the spinout issues additional shares to the TTO to maintain a 3% ownership up until the time the startup raises at a $6M or greater post-money valuation). It’s very important that the TTO doesn’t have anti-dilution protections until much higher valuations, because it makes each round of funding extra-dilutive.
Diligence clauses: These allow the university to modify or terminate the agreement if you don’t meet specific goals. Make these clauses extremely achievable (>95% confidence), set all timelines to be 3-5x what you think they’ll be, and add grace (extension) periods where you can remedy missed timelines without added fees, penalties, or other contractual “triggers”. If possible, add clauses that allow missed diligence clauses to be auto-extended in exchange for a small fee. For example, you can include a 12-month extension for $5,000 that you can use up to three times as long as you are making commercially reasonable efforts to advance the technology (remember to define what constitutes this!).
Milestone payments: Cash payments to the university triggered when particular development or fundraising milestones are met. Common milestones/triggers include: $X million dollars in funding, IND approval by the FDA, first use in humans, beginning Phase III clinical trials, filing a New Drug Application with the FDA to allow marketing of the product. These should be tied to a major funding event to eliminate cash flow risks. Keep milestone payments to a minimum and never pay more than 2% of new $ raised – that’s cash that should be spent on building your company!
Upfront payments: Lump-sum cash payments made to the university at the time of licensing or sale. They are typical for larger companies buying a patent (also called “full assignment”). Spinouts should generally preserve their cash and not pay these. However, it (rarely) does make sense to buy the patent upfront when the equity you’d sell to investors is less than the equity you’d give to the university in order to license. For example, you might be offered a licensing deal for either 5% equity (plus royalties or other terms) or a full assignment for an upfront payment of $500K. If your company can easily raise additional capital at a $10M or greater post-money valuation, you should probably just buy the patent. Why? Raising $500K at a $10M valuation = 5% dilution (which you’d be giving to the university anyhow) but you avoided the royalties. If you can raise that $500K on a valuation higher than $10M, you’d minimize dilution. Only pursue this route if you’re certain the company can raise what it needs and also the extra money to buy the patent.
Maintenance or diligence fee: Admin fees paid annually to keep the license. They should be negligible (<$5K/yr or less).
Reimbursement: Cover the university’s legal costs of filing and, sometimes, prosecuting the patent. Typically these costs are deferred until your spinout gets funding and many universities don’t charge these at all. Anywhere between $10K and $25K is fine. Push back if the fees are higher than that.
Sublicensing: Describes the share of revenue owed to the university if you sublicense the patent to another company. Ideally, keep these at <10%. TTOs are wary of companies who immediately sub-license patents – as that might mean no new tech development – and they might try to push these fees much higher. If you’re very sure you will not be sub-licensing, then it’s OK to agree to higher fees.
Less common terms:
Improvement rights: Describe the university’s rights to use or own improvements to the technology (including patents filed by the startup) for a specified time period after the license. Generally try to avoid these but it’s potentially fine for a university to be granted a non-exclusive license for educational purposes only for any improvements for 1 year.
Patent prosecution: Determines who is responsible for prosecuting the patent. We believe spinouts should stay in control here. This gives startups more control over settlement negotiations, which is typically how most legal actions end, and helps ensure the settlement is in the best interests of the startup, not the university. The downside to controlling prosecution is that you also have to pay for it.
“Percentage of proceeds from any liquidity event”: If they can’t hold startup equity directly, the university might ask for a percentage of any liquidity event (merger, acquisition, or IPO). This clause essentially translates to “university gets non-dilutable equity.” Set it to a maximum of one-seventh (15%) of reasonable entry-level equity you’d be willing to give the university to factor in the dilution. The 15% multiplier is based on this data (1, 2) showing that the founder ownership typically goes from 100% at inception to 15% at IPO. Based on that, aim at 15% * 5% (typical entry equity stake for university) = 0.75% as a healthy maximum.
Consent clauses: These give blocking rights to the university for events like a merger or an acquisition. Avoid them! If you can’t, ask for automatic approvals within no later than 10 business days from receiving the request, and for consent “not to be unreasonably withheld.”
Venture Deals has an entire chapter on negotiations. To quote from it: “there are only three things that matter when negotiating financing: achieving a good and fair result, not killing your personal relationship getting there, and understanding the deal that you are striking.” The same applies to negotiating with your TTO. While preparing to negotiate, it’s helpful to realize that tech transfer is a “single game” for founders but a “repeated game” for everyone else. This can be helpful to you in some ways (TTOs, VCs, PIs will want to be friendly because they care about their reputation among founders) but can also have negative ramifications (a PI or a law firm might care more about a continuous good relationship with the TTO than a great relationship with a single startup). These dynamics are impossible to avoid but helpful to keep in mind.
What terms are negotiable? Everything in the term sheet is negotiable.
How to get fair licensing terms? Study benchmarks, talk to other founders who have recently negotiated with the same TTO, talk to advisors, and include your investors if you have any. If a TTO is asking for an equity range outside of what seems fair to you, one of the strongest arguments is referencing another licensing agreement’s terms which were accepted by that TTO.
Where to get comps of licensing data from? We’ve been crowdsourcing licensing data from founders and scraping SEC filing data. We plan to publish it. If you’ve spun out, you can still contribute.
In the meantime, you can find comp data from these sources:
Licensing Executive Society surveys, KTMine, TransACT, Spinout.fyi
Note: Licensing Executive Society doesn't provide raw data and it costs $200 to access their analysis. KTMine and TransACT require ~$3K paid subscriptions. Spinout.fyi is free (thanks Nathan!).
SEC filings for public companies, including S-1s (we collected some here)
Ask your legal counsel about their database of licensing terms
What impacts the final terms you’ll get?
Ranges and comparables don’t cover the full nuance of the licensing logic applied by the TTO as tech licenses are very context-dependent. A few examples of what impacts the terms you’ll be offered:
Applicability, i.e. market potential of the technology. The smaller the market and the harder it is to commercialize, the better the licensing terms for the startup.
Novelty and breadth of the IP. The narrower the IP, the better licensing terms. Widely applicable breakthrough technologies are much more exciting to TTOs and generally harder to license on great terms, because they know other parties are also likely willing to pay.
Competing licensee candidates. If fewer parties want to license a given IP, the terms will be more friendly.
How much money the university has already spent to enable the creation of this IP. The more they already spent, the bigger return they’re looking for.
Who are the inventors? If a famous PI was involved in the creation of the technology, the terms might be more lenient.
Incentives of your TTO officers. At the end of the day, your TTO has to feel they are also getting a fair deal. Take time to understand what your TTO wants and your odds of meeting in a reasonable place will be higher.
Importantly, when looking at comps, all terms need to be looked at in aggregate. TTOs are sophisticated negotiators and they often make calculated trade-offs between terms given the licensee's particular situation. For example, a licensing deal with a particularly low royalty rate might have an above-average equity allocation, or vice versa.
Global economic trends, access to VC capital, inflation, and other factors will impact the terms your TTO is willing to accept. As of Q4’22, we’re in a pretty dynamic environment both in terms of early stage funding markets but – more importantly for the TTOs – in a high inflation environment. Because of that, TTOs might prioritize for upfront payments even if they were prioritizing for equity a year ago. When law firms prepare for a negotiation with the TTO, they look for more recent benchmarks in their internal databases. Specifically, they look through most recent agreements in a similar category for a similar technology with the same TTO to see what terms were offered. Lean on this.
Renegotiating the licensing agreement post-signing: This is technically possible but will be hard to accomplish. As with any agreement, you’ll need leverage. “Our investors think royalties are too high” is usually not enough. But saying that “an investor who is about to give us a term sheet for a major round said we really need to fix the royalties first” could do the job. If negotiations are taking too long, dig in to understand why. Sometimes there’s a bureaucratic misunderstanding that is easily resolved like who needs to consent to the agreement.
Other leverage points to speed things up or improve terms:
VC offer or a conditional promise of an offer for $X “if the licensing terms are reasonable”. This doesn’t have to be a “term sheet” for an equity round, as pre-seed and seed financing commonly happens using convertible notes or SAFEs. An email you can bring to the TTO with a firm offer in writing might be enough. A very effective argument might look something like “The VC firm Science Capital is willing to give us $1m next week but only if you cap the equity you’re taking at 5% and have the anti-dilution end once we hit a $15m valuation.”
PI emailing the TTO that the negotiations are going poorly, or that the TTO is being unfair or uncooperative. Especially if your PI is a “star” at the university, this is calling in the big guns and is one of the most effective levers at your disposal.
At Fifty Years, we believe translating science from the benchtop to the benefit of all is one of the greatest levers for creating positive change. Most technologies won’t translate without the scientists who made the discoveries and invented the tech taking it upon themselves to spin out, so we hope you take the leap! At Fifty Years we love helping great scientists become great entrepreneurs, so if you do want to start a startup, please reach out.
This is meant to be a living playbook. Please send feedback and improvements to spinout@50y.com.
Thanks to Ash Trotman-Grant, Serena Zhang, Josie Kishi, Gleb Kuznetsov, Tess van Stekelenburg, Uri Lopatin, Daniel Hussey, Mohan Iyer, Parikshit Sharma, Stephen Chapman, Chase Moyle, Jason Fontana, Devika “Dee” Thapar, Ale Maiano, Milad Alucozai, and Fifty 50 members Anastasia Ershova, Eeshit Vaishnav, Jack Silberstein, James Banal, and Jarod Rutledge for insightful feedback and help crafting this playbook.