Tag Archives: Serial Entrepreneur

Stanford and Startups

Stanford is in the top2 universities with MIT for high-tech entrepreneurship. There is not much doubt about such statement. For the last ten years, I have been studying the impact of this university which has grown in the middle of Silicon Valley. After one book and a few research papers, here is a kind of concluding work.

A little less than 10 years ago, I discovered the Wellspring of Innovation, a website from Stanford University listing about 6’000 companies and founders. I used that list in addition from data I had obtained from OTL, the Stanford office of technology licensing as well as some personal data I had compiled over years. The report Startups and Stanford University with subtitle “an analysis of the entrepreneurial activity of the Stanford community over 50 years”, is the result of about 10 years of research. Of course, I did not work on it every day, but it has been a patient work which helped me analyze more than 5’000 start-ups and entrepreneurs. There is nearly not storytelling but a lot of tables and figures. I deliberately decided not to draw many conclusions as each reader might prefer one piece to another. The few people I contacted before publishing it here twitted about it with different reactions. For example:

Katharine Ku, head of OTL has mentioned another report when I mentioned mine to her: Stanford’s Univenture Secret Sauce – Embracing Risk, Ambiguity and Collaboration. Another evidence of the entrepreneurial culture of that unique place! I must thank Ms Ku here again for the data I could access thanks to her!

This report is not a real conclusion. There is still a lot to study about high-tech entrepreneurship around Stanford. With this data only. And with more recent one probably too. And I will conclude here with the last sentence of the report: “How will it develop in the future is obviously impossible to predict Therefore a revisited analysis of the situation in a decade or so should be very intersting.”

Myths and Realities of Serial Entrepreneurs

This is my latest contribution to Entreprise Romande. This is a topic I have covered from time to time and though that it would be interesting to share it with a wider audience thanks to the FER newsletter.


I have always been suspicious of the concept of serial entrepreneur, this creator who, according to Wikipedia, “continuously comes up with new ideas and starts new businesses, as opposed to a typical entrepreneur, who will often come up with an idea, start the company, and then see it through and play an important role in the day to day functioning of the new company.” Why such a bias if one considers exceptional serial entrepreneur like Steve Jobs (Apple, Next, Pixar), Elon Musk (PayPal, Tesla, SpaceX) or English “rock star” Richard Branson who declined Virgin in music, retail, air transportation and mobile communications? Because from experience, the idea of going from one idea to another seems far from sufficient if the entrepreneur does not invest an enormous and sustained activity to market and commercialize her project? Not really, since the three examples show that it can be not superficial hyper-activity, but the success of products or services consecutive to a total commitment of their creators.

My suspicion was built over time, because with the exception of some mythical characters always cited as examples for good reasons, I had the belief of recurring “patterns” and the example of Steve Jobs is a good illustration: he has never done as well as with Apple, his first creation. A few years ago, I dedicated some time to a statistical study about the “performance” of these serial entrepreneurs in comparison with their more conventional counterparts. [1] The study of some 450 serial entrepreneurs out of a group of more than 2700 founders gave some interesting results: if on average, serial entrepreneurs do better than others with their first business (the value created is larger, and with less investment), the trend is reversed with the following ventures, and beginning with the third, they do less well, while raising more money from their investors. QED! This study was perhaps the result of a particular situation in Silicon Valley and Stanford? A 2011 study of some 600 British entrepreneurs [2] shows that 60% of the founders having experienced failure were serial entrepreneurs while they accounted for half of the sample. The authors are known as experts on the subject and their many studies do not show that experience is a real advantage.

If the facts seem somewhat scuttle the myth, it is also interesting to analyze a little further. A serial entrepreneur, and more if she had a successful career, will have an enormous self-confidence and probably a seduction power to attract investors and talent for her future projects. She will be ready to take risks, even greater as she has already succeeded, so that failure may have a lower financial impact. The authors of the British study add that those who have failed have experienced such trauma that they repress this failure to the point not to learn anything from the experience …

What lessons for those – investors and employees – who are willing to blindly follow such a hero? Probably the need to show a little caution and analyze with some rationality if the project makes sense and if the creator seems a minimum rational in his vision of the development of this new project. In reality, success will always remain the domain of the exception, an unlikely alignment of the planets. An entrepreneur must always be optimistic, but if he loses too much sight of these realities, blindness can be fatal. And I would add a message to the entrepreneur without experience: by listening too much to the advice of those who “know”, the entrepreneur may forget his inner voice, the intuition so fundamental for all creative people. This myth of the serial entrepreneur perhaps shows that talent is more important than experience …

[1] Serial Entrepreneurs: Are They Better? – A View from Stanford University Alumni – Babson Conference “Frontiers of Entrepreneurship Research” 2012. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2133127
[2] Why Serial Entrepreneurs Don’t Learn from Failure. Par Deniz Ucbasaran, Paul Westhead et Mike Wright . https://hbr.org/2011/04/why-serial-entrepreneurs-dont-learn-from-failure

The unusual and amazing success of two serial entrepreneurs: Andy Bechtolsheim and David Cheriton

Serial entrepreneur is a buzz word. I have never been convinced by the link between serial entrepreneur and success. I even made an analysis for the ones linked to Stanford University (check Serial entrepreneurs: are they better?). But from time to time, you see such amazing and rare success stories.

Andy Bechtolsheim (left) and David Cheriton (right) [with Arista’s co-founder, Ken Duda).

Andy Bechtolsheim‘s is a Silicon Valley icon. In 1982, he co-founded Sun Microsystems. Born in Germany in 1995, he moved to the USA at age 20 for his master at CMU. He moved to Silicon Vallley to work at Intel but ended up at Stanford for his PhD. Sun came thereafter. He stayed there until 1995…

David Cheriton is a Stanford professor. Born in 1951, he got his BS from UBC and his PhD from the University of Waterloo. He moved to Stanford in 1981. I am not sure how they met, but they co-founded Granite Systems in 1995. A year later, it was bought by Cisco for $220M. Bechtolsheim stayed with Cisco until 2003. Cheriton is still a Stanford professor. Two years later, they met with two unknown Stanford students, Larry Page and Sergei Brin. Both invested $100’000 each in their start-up, but this is another story…

In February 2001, they co-founded another networking start-up, Kealia. In April 2004, “Sun issued an aggregate of approximately 20,000,000 shares of common stock (including assumed options) in exchange for all outstanding stock and options of Kealia” (Newswire reference). At that time, Sun’s share was worth about $4, so it would have been an $80M acquisition. That same year, Google went public (on August 19) at $85/share. They had received 1’600’000 shares for their $100k investment (i.e. $0.0625 per share, a multiple of 1’360 and with a six month lock-up, the share value more than doubled…) The Kealia success is all but relative…

Granite might have had a logo, but I could not find it on the web. Kealia was apparently always in stealth mode. No logo available either

But it did not stop them. In October 2004, they co-founded Arista Networks. The name at the time was Arastra. The company just went public which is the motivation for this post. My usual cap. table follows. And because they made so much money, the two serial entrepreneurs nearly funded it entirely… Not the smallest success of all!

Click on picture to enlarge

PS: Are Cheriton and bechtolsheim good friends? I have not clue, but the Arista IPO document mentions a litigation:

On April 4, 2014, Optumsoft filed a lawsuit against us in the Superior Court of California, Santa Clara County titled Optumsoft, Inc. v. Arista Networks, Inc., in which it asserts (i) ownership of certain components of our EOS network operating system pursuant to the terms of a 2004 agreement between the companies, and (ii) breaches of certain confidentiality and use restrictions in that agreement. Under the terms of the 2004 agreement, Optumsoft provided us with a non-exclusive, irrevocable, royalty-free license to software delivered by Optumsoft comprising a software tool used to develop certain components of EOS and a runtime library that is incorporated into EOS. The 2004 agreement places certain restrictions on our use and disclosure of the Optumsoft software and gives Optumsoft ownership of improvements, modifications and corrections to, and derivative works of, the Optumsoft software that we develop.

In its lawsuit, Optumsoft has asked the Court to order us to (i) give Optumsoft copies of certain components of our software for evaluation by Optumsoft, (ii) cease all conduct constituting the alleged confidentiality and use restriction breaches, (iii) secure the return or deletion of Optumsoft’s alleged intellectual property provided to third parties, including our customers, (iv) assign ownership to Optumsoft of Optumsoft’s alleged intellectual property currently owned by us, and (v) pay Optumsoft’s alleged damages, attorney’s fees, and costs of the lawsuit. David Cheriton, one of our founders and a former member of our board of directors who resigned from our board of directors on March 1, 2014 and has no continuing role with us, is a founder and, we believe, the largest stockholder and director of Optumsoft. The 2010 David R. Cheriton Irrevocable Trust dtd July 27, 2010, a trust for the benefit of the minor children of Mr. Cheriton, is our largest stockholder.

Optumsoft has identified in confidential filings certain software components it claims to own, which are generally applicable tools and utility subroutines and not networking specific code. We cannot assure which software components Optumsoft may ultimately claim to own in the litigation or whether such claimed components are material.

On April 14, 2014, we filed a cross-complaint against Optumsoft, in which we assert our ownership of the software components at issue and our interpretation of the 2004 agreement. Among other things, we assert that the language of the 2004 agreement and the parties’ long course of conduct support our ownership of the disputed software components. We ask the Court to declare our ownership of those software components, all similarly-situated software components developed in the future and all related intellectual property. We also assert that, even if we are found not to own any particular components at issue, such components are licensed to us under the terms of the 2004 agreement. However, there can be no assurance that our assertions will ultimately prevail in litigation.

On the same day, we also filed an answer to Optumsoft’s claims, as well as affirmative defenses based in part on Optumsoft’s failure to maintain the confidentiality of its claimed trade secrets, its authorization of the disclosures it asserts and its delay in claiming ownership of the software components at issue. We have also taken additional steps to respond to Optumsoft’s allegations that we improperly used and/or disclosed Optumsoft confidential information. While we believe we have strong defenses to these allegations, we believe we have (i) revised our software to remove the elements we understand to be at issue and made the revised software available to our customers and (ii) removed information from our website that Optumsoft asserted disclosed Optumsoft confidential information.

We intend to vigorously defend against Optumsoft’s lawsuit. However, we cannot be certain that, if litigated, any claims by Optumsoft would be resolved in our favor. For example, if it were determined that Optumsoft owned components of our EOS network operating system, we would be required to transfer ownership of those components and any related intellectual property to Optumsoft. If Optumsoft were the owner of those components, it could make them available to our competitors, such as through a sale or license. In addition, Optumsoft could assert additional or different claims against us, including claims that our license from Optumsoft is invalid. Additionally, the existence of this lawsuit could cause concern among our customers and potential customers and could adversely affect our business and results of operations. An adverse litigation ruling could also result in a significant damages award against us and the injunctive relief described above. In addition, if our license was ruled to have been terminated, and we were not able to negotiate a new license from Optumsoft on reasonable terms, we could be required to pay substantial royalties to Optumsoft or be prohibited from selling products that incorporate Optumsoft intellectual property. Any such adverse ruling could materially adversely affect our business, prospects, results of operation and financial condition. Whether or not we prevail in the lawsuit, we expect that the litigation will be expensive, time-consuming and a distraction to management in operating our business.

We do not believe a loss is probable; however, it is reasonably possible. Due to the early stage of this matter, no estimate of the amount or range of possible amounts can be determined at this time.

Serial entrepreneurs: are they better?

Are serial entrepreneurs better than novices? This is a classical topic in entrepreneurship and it seems to me that urban legend says yes! There has been academic research going that way, with one major argument being that experience matters. However, I just finished my own analysis which I presented at the BCERC conference in Fort Worth (Texas). It is based on my previous work related to Stanford related start-ups: Stanford University and High-Tech Entrepreneurship: An Empirical Study (you can have a look here at the presentation and the article). The article on serial is available on the SSRN network and you can have a look at the presentation in pdf below.

click on picture to view the pdf slides

And the conclusion? I do not find evidence that novice entrepreneurs would be less performant. It is a “work in progress” but if you have a look at the slides, you might see it in particular on slides 7, 9 or even 20. Slide 7 shows average performances according to experience. Slide 9 (q-q graphs) sshows something else: serial founders would do worse with time. Finally slide 20 that serial successful in the past have a new success rate of about 28-29 % which is similar to novices (the novice figure is not on the slide), whereas serial who failed before have a lower success rate. As if talent mattered more than experience…