Stanford sails to StartX

There is a great tale about how British naval captain Horatio Nelson chased his French peer for months across the ocean only to fail to capture his adversary at the end.
This failure led to calls for his command but when his superior read the ship logs of his day-to-day approach to overhauling the French ship he recommended Nelson for a promotion.
This degree of thoughtfulness – trying to put himself in the person’s shoes when they were making their actions – rather than just judging from the final result is perhaps an appropriate approach to looking at US-based Stanford University’s historic decision to provide “unlimited” support from its balance sheet for a university venturing fund.
Outside of China, Stanford is the epicentre of Silicon Valley after its professor of engineering, Fred Terman, encouraged two of students, Bill Hewlett and Dave Packard, to set up a technology company in their garage in the 1930s and which helped act as a locus for other start-ups to form.
Stanford remains easily the most important academic institution for turning out globally-relevant entrepreneurs with disruptive ideas. Since Hewlett-Packard’s formation Stanford alumni/ae and faculty have started close to 40,000 companies with annual revenues of $2.7 trillion, according to its survey published in October, while research this year into venture rounds tracked on database CrunchBase by blogger Max Woolf shows those affiliated to Stanford have raised the most by value and volume.
There has, therefore, been a slight paradox as to why Stanford has avoided directly funding its student and faculty entrepreneurs, which takes us back to the Nelson analogy.
Superficially, it looks careless at best for a university to trumpet its 40,000 start-ups it has helped spawn only to then say it will now fund a university venturing vehicle taking equity stakes. But a look through the journey Stanford has been on to reach this milestone is instructive.
Before 1981, Stanford in its official history of its Office of Technology Licensing (OTL) noted there were plenty of arguments about the pros and cons of commercialising research and academia (see summary below ^). The history said it “had a blanket policy forbidding the taking of equity, and for years afterward it simply was not done.
“Indeed, Niels Reimers, founder of OTL [in 1970], resigned in 1991 over the insistence by Stanford’s president at the time, Donald Kennedy, that he submit all equity deals to administrative scrutiny for potential conflicts of interest.”
As Greg Franklin, co-founder of venture capital firm Intellect Partners with Reimers, by email said: “During Niels’ SU [Stanford University] tenure, the university was very cautious about taking equity ownership in start-ups – so not to risk exposure of the large university endowment to frivolous lawsuits. Ie being a shareholder with deep pockets, the concern was aggressive plaintiffs suing the start-up but then seeking damages from its (well heeled) shareholders. Those days appear to be history.”
Stanford has benefited from start-up’s equity growth in three primary ways:

• Philanthropy from shareholders back to their alma mater, which reached $6.2bn in its last phase of seeking donors;
• The university’s $17bn endowment, run by Stanford Management Company since 1991, committing about 3.75% of the total to venture capital funds that back the start-ups, as well as links between university departments and venture capitalists, such as the Stanford Engineering Venture Fund since 1984; and
• Equity as a result of license agreements as, under the Bayh-Doyle Act, since 1980 universities own the rights to intellectual property funded by government money that underpins the majority of research at the institution.

An example of how the model has worked to Stanford’s benefit could be seen search engine Google. In its most recent awards, OTL inducted seven new inventions to the OTL Hall of Fame, for a total of 17 inventions out of over 9,200 cumulative inventions received. These Hall of Fame inductees included hypertext searching – patented in 1996 – developed by Larry Page and Sergey Brin, co-founders of Google.
OTL coyly says each inductee has “generated over $5m each in cumulative royalties to date” – Google’s original exclusive license in return for cash, stock and annual royalty payments ran to 2011. But Stanford also sold 184,207 shares, earning $15.6m, in Google’s stock market flotation in 2004, with a further 7,574 shares of class A Google stock and 1,650,289 B shares locked up and sold soon afterwards, according to news provider Wall Street Journal at the time. SMC, meanwhile, was a limited partner in Google’s two primary venture capital backers, Sequoia and Kleiner Perkins Caufield & Byers, that was estimated to be worth hundreds of millions of dollars to the university.
So why change this trinity of positive returns now?
John Hennessy, current president of Stanford – and who assumed a seat on Google’s board in March 2004 ahead of its initial public offering that brought with it 65,000 shares, gave a clue in an interview to Strategy Business magazine in 1998.
Then-dean of Stanford’s school of engineering, Hennessy said: “There’s two kinds of technologies in the world: stuff that is patentable and broadly applicable and the right thing to do is to give it to OTL.
“Then there’s stuff that is more a preliminary proof of a concept. It’s not patentable, and the real value is in the people and their understanding of that technology and how it can develop into a useful product.
“OTL’s role there is not to get in the way. That’s when the right thing to do is to say, ‘Godspeed, go do it’.”
Hennessy was referring to what the magazine described as a parallel, informal path from the university’s biochemistry labs and computer science classes to commercial markets, instead of the formal way through OTL.
Even in 1995, 25 years after its formation, Stanford’s OTL had equity in only a “couple of dozen of the hundreds of companies it had licensed” as taking equity was seen as the “icing on the cake” of negotiations that reaped $44m that year from royalties in 220 technologies. As of 31 August 2012, Stanford held equity in 124 companies as a result of license agreements struck by OTL.
Last year, according to its annual report, Stanford received $76.7m in gross royalty revenue from 660 technologies, $1.23m from selling equity in four companies and shares in 17 new companies.
But with potentially hundreds of companies formed each year from people affiliated to the university, the institution has been exploring how to support them without undermining its mission of education first, and financial returns second.
This is where the new university venturing fund, Stanford-StartX, now comes in.
While some StartX companies may be licensees of Stanford technology through OTL, the terms of the OTL license would be the same irrespective of the StartX relationship, the university said. Katherine Ku, who has been director of Stanford’s OTL since 1992, by email said: “No overlap at all [between StartX and OTL]; we’re not coordinating or anything.”
According to StartX’s frequently-asked questions section: “This [uncapped fund] is the first time that Stanford University or Stanford Hospital has earmarked a fund specifically for the purpose of investing in early-stage companies founded by members of the Stanford community.”
The funding will come from unrestricted university funds, separate from the endowment, managed by Randy Livingston, vice-president for business affairs at Stanford.
The fund aims to provide 10% of the investment amount being raised by each StartX company. The idea is the fund will help entrepreneurs save time by raising the final bit of the round which often takes as much time to source as the first 80% to 90%. Solving this “pain point” helps answer concerns, such as news provider TechCrunch journalist Billy Gallagher’s article that said: “Right off the bat, I don’t understand why students would want Stanford’s money over other VC firms.”
(Read student reaction supporting the fund at bottom of article*.)
The stage of the investment can be seed or any later round provided the company is raising a minimum of $500,000 and professional venture capital or angel investors lead the funding syndicate.
Now legally and financially separate from the university having been formed out of the Stanford Student Enterprises organisation, StartX is a non-profit that runs an accelerator programme with three classes per year for Stanford-affiliated entrepreneurs.
The Stanford-StartX fund has already approved commitments in six companies from the 109 graduated companies from StartX and the 27 still in the programme, including:
Knotch – a network that allows like-minded people to connect based on shared interests founded by Stanford alumna Anda Gansca; and
Cytobank – a big data and analytics platform for single-cell technologies in healthcare co-founded and led by Nikesh Kotecha, a consulting faculty member in the Computational and Systems Immunology program at Stanford University.
These are in addition to recent investment rounds for StartX-backed companies, such as:
○ Genapsys’ $13.5m series A round;
○ InstartLogic’s $17m series B round;
○ NuMedii’s $3.5m series A round;
○ Clinkle’s $25m seed round, which included company Intuit (itself co-founded by two Stanford alumni);
○ Breakthrough’s $5m series A round;
○ Propeller’s $1.35m prize from Google and space agency NASA; and
○ Heap’s $2m seed round.
In total, nearly $200m has been invested in StartX’s graduated companies at an average of more than $1.8m each. And the graduated companies have also quickly had 10 exits for an estimated aggregate $130m to $150m, including:
○ Luma Camera (Instagram), 2013, co-founded by Stanford drop-out Alex Karpenko and acquired by Stanford alumni;
○ WifiSlam (Apple), 2013, for a reported $20m after funding from AngelList’s Naval Ravikant, Google’s Don Dodge and Start Fund’s Felix Shipman;
○ Loki Studios (Yahoo), 2013;
○ Shopwell (HarvestMark), 2013, founded by Stanford design school alumnus Brian Witlin and incubated at Ideo;
○ 6Dot (ProxTalker) 2013;
○ Accevia, 2011, a financial services company that was acquired by its first undisclosed client;
○ Stypi (Salesforce.com), 2012; and
○ Thinkbulbs (Megatasty Labs), 2010.
StartX selects about 9% of the applicants – “those who are both smart and understand what it means to be a founder” – from the 6% to 7% of the student population who apply each year. The selection process is aided by StartX’s connections to the student population that intern at the non-profit. These students tend to be aware of the reputation and potential for the applicants’ ideas, which VCs, such as managers of Dorm Room Fund have regarded as a useful vetting process.
And by waiting for institutional investors to lead the round, the Stanford-StartX fund will capture the experience of the VCs that decide what could make a good bet.
As Brad Hayward, senior director of strategic communications at Stanford University, by email said: “The timing [of the fund and grant] is principally a function of the development and maturation of StartX itself over the last few years.”
And also that StartX is aligned to Stanford’s core goal of education-first. In its FAQs, StartX says: “StartX is, at heart, an education-focused entity. The fund was initiated as a vehicle to help StartX companies succeed.
”As a mission-driven non-profit, StartX is able to prioritize the value of the programme for the participating entrepreneurs over everything else, including profit.
“The non-profit charges no equity for participating in the programme, which is industry and stage-agnostic. StartX can focus on developing the best entrepreneurs regardless of whether they have just developed a prototype or are raising their series A round of funding. StartX leverages the Stanford community to help run the programme, including students of all backgrounds.”
Cameron Teitelman, founder and chief executive of StartX, who graduated 2010 with a degree from Stanford in Management Science & Engineering, said: “In the past two years we had offers to set up a fund but we held stubbornly to our values of being a non-profit and putting founders first.
“Limited partners want financial returns while accelerators often try to be educational, which causes a tension. We knew Stanford was the right partner as it has not put financial returns but education is first, which avoids conflicts. Stanford was aware of students dropping out [such as Clinkle and Luma Camera] or starting companies [without their support], so this fund is a way for them to double down on helping them be successful because this helps source the best students and gain donations in future. But Stanford did not want an in-house incubator as professors saw in the [1990s dot.com] bubble students were overfunded and wasted five years of their lives on start-ups. People think we [at StartX] have built an incubator but we are providing training.”
The fund is also a partial answer to concerns, such as in news provider New Yorker’s April feature, “The End of Stanford?”, about how “the leadership of a university has encouraged an endeavour [Clinkle] in which students drop out in order to do something that will enrich the faculty”.
Teitelman said no individuals from StartX or the university would take stakes in the start-ups backed by the fund. The non-profit gains from carried interest on the fund’s performance but these returns are likely to be small at least initially requiring it to rely on grants.
Since 2009, StartX has been funded by $1.65m grants from foundations, such as Kauffman Foundation and Blackstone Charitable Foundation; corporate partnerships, including Microsoft, Intuit, Cisco, AOL (where StartX is based), Groupon and AT&T; and venture partners, for example Greylock Partners and Founders Fund.
As part of the fund announcement, Stanford University and Stanford Hospital said they would jointly provide a $3.6m, three-year grant to StartX, to cover about half its costs for its team of eight staff and 18 volunteers/interns/part-timers as it looks to sail to other universities in the US and round the world.
* Comments from the TechCrunch article
Andrew Lockhart
As a StartX founder who has Stanford as an investor, I can speak personally to why I opted in to taking Stanford’s money – This arrangement is helping to ensure financial sustainability for the StartX program, which was critical to our development as a company. If my company does end up being successful, I would love to know that the program who helped us get to where we are is benefitting as well. I will (and did) take that over another value-add investor any day.
Mike Dorsey
@AndrewLockhart, I agree wholeheartedly.   As a current StartX founder (DataFox), I am happy to see this in place.  It’s on friendly terms and it’s efficient.  If proceeds from my company’s success are going to go somewhere, I’m more than happy for StartX and Stanford to benefit (as they were both core to our company’s formation and acceleration).  And after all, a real portion of Stanford’s endowment $ goes into the top venture capital firms in Silicon Valley anyway (Stanford is an LP in that case).  The VCs take a cut of that $ and re-allocate it to startups.  Either way, you’re raising some of your $ from Stanford.  In this new format, Stanford is cutting out the middleman (VCs) and honoring StartX’s position as a group of students that created value for us without taking equity. Entrepreneurs get capital on reasonable terms. Stanford gets direct participation (in thrice vetted companies). StartX can continue being a world class accelerator that offers its service for free (no fees, no equity).
^ Stanford discusses pros and cons of research commercialisation
The pros and cons of licensing inventions
In short, there are three primary arguments against licensing university inventions.
First, it could promote conflicts of interest, tempting faculty to slant their research in the direction of commercial return. In this case, basic research might suffer because applied research tends to be more easily licensed for immediate commercial use. Graduate student researchers might be exploited by being steered into commercially promising projects as cheap labor for sponsoring companies. The university researcher’s traditional role to disseminate new knowledge by publication might be subordinated to the secrecy requirements of the patenting process, or to the sponsor’s desire to maintain secrecy for competitive advantage – both of which clearly would have a negative effect on the free dissemination of scientific research.
Second, since licensing imposes financial burdens on industry in the form of license fees and royalties, products may become more expensive, diminishing our country’s competitiveness.
Third, since most scientific research is funded by taxpayers through federal funding, one might expect that products developed from such research should be available to taxpayers free of charge.
There are, however, many more arguments in support of licensing university inventions.
First, while the federal government has traditionally been the major sponsor of basic research conducted in universities, the current trend is to limit such funding. Universities thus are faced with the need to develop alternate sources of funding or to curtail their research activities. Licensing income can be a critical source of much-needed unrestricted funding.
Second, these constraints on university research funding come at a time when global competition makes innovation the key ingredient for competitiveness in most industries. American innovation has historically been provided by established industrial companies with large research and development capabilities. Today, however, innovation is increasingly provided by small but dynamic companies established for the sole purpose of developing a new idea or technology. In contrast to the large companies, these emerging companies do not have the cash reserves needed to generate new ideas and technologies in their own research laboratories. Instead, they raise the capital to develop ideas which were generated elsewhere, often in universities. As a result, universities have become an ever-more important source of the new ideas needed by American industry to stay competitive.
Third, licensing often is the only way a new invention will ever become a product. University inventions are typically in the very early stages of development – nowhere close to commercial reality. A licensee thus must be prepared to invest significant resources in further development, product design, applications engineering, and quality testing before bringing it to the market. Unless a licensee is assured of a proprietary position in the marketplace that will enable it to recoup its investment, the company has little incentive to pursue the project.
Fourth, the possibility of a financial return may create some incentive for the academic researcher who will share in the profits. Scientists often are reluctant to spend a great deal of time on non research-related tasks. A financial incentive sometimes encourages researchers to make the extra effort to disclose their inventions, assist with patent preparation, and provide support to the licensing personnel, patent attorneys, and licensees in the evaluation, development, and patenting of the new technology. Often even more important is the fact that shares of all licensing income are paid to the inventor’s department and school. This usually enhances a researcher’s standing among his or her peers who may benefit from such additional funds.
Other reasons to license university technology include strengthened relationships with industry, which can provide productive cooperative arrangements for researchers; jobs for graduates; and new sources of research funding. Some researchers also find that exposure to real world applications provides them with valuable knowledge and stimulates their thinking in new directions.
Some Special Challenges of Licensing University Inventions
The first challenge is the breadth of research in which a university is likely to be engaged. A commercial company generally concentrates its efforts in a relatively limited area, which its licensing personnel understand well. At universities, licensing personnel must comprehend and manage inventions in an enormous range of technical industries.
A second challenge is that academic scientists are far more independent than industrial researchers – yet they must be willing to take on all the same responsibilities. The inventor must be motivated to file an invention disclosure, detailing the history of the invention, co-inventorships, publications, research proposals, sponsorships, and possible applications. He or she will need to work with the patent attorney to provide background material, review drafts, and execute the necessary documentation. And potential licensees often ask the inventor to make additional demonstrations, provide more detailed information and the most recent experimental results, or attend meetings to educate their own people.
A third challenge results from the fact that academic researchers’ foremost interest and focus is to gain new scientific knowledge and publish the results of their research as soon as possible. As a result, our experience at Stanford has shown that some inventions can no longer be patented or licensed by the time they are disclosed to us because proprietary aspects of the invention have already been published or presented in meetings.
A related issue is that researchers often have close contacts with colleagues in other institutions or industry with whom they collaborate and talk freely about their ideas and work in progress. This can cause problems in determining the actual inventors on a project. In cases in which scientists from industrial companies have enjoyed frequent access to research groups at Stanford, company attorneys have claimed the inventions as well, complicating negotiations to the point where the inventions could no longer be patented or licensed.
In working with academic inventors, it is particularly important to clarify the distinction between co-authorship and co-inventorship. For publication purposes, it is traditional practice to list as co-authors everyone who has worked on a project. In filing a patent, however, only inventors who have made a distinctly inventive contribution may be listed or the patent may not hold up to challenges later. And since Stanford inventors receive a share of all royalty income, it is important to carefully determine each co-inventor’s creative participation. Finally, when co-inventors from other universities are involved, an inter-institutional agreement is created that specifies which university will handle the licensing and how the income will be split between the universities.
A fourth challenge is that university inventions typically are far from full-fledged, commercial products. In many cases, their feasibility has not even been proven experimentally when the disclosure is filed. As we mentioned earlier, a licensee must generally spend significant resources and time to develop them. On the other hand, a company taking a license at such an early stage may have a greater competitive advantage and, often, the inventor will cooperate with the licensee in such development. Such a relationship can be beneficial to both, but the university and the inventor must take steps to ensure that no conflict of interest results.
Finally, our primary objective at Stanford has been to get an invention into widest possible use, rather than to seek maximum financial return. In many cases this has meant granting non-exclusive licenses to several companies, instead of a more lucrative, exclusive license to one.