STANFORD UNIVERSITY AND HIGH-TECH ENTREPRENEURSHIP: AN EMPIRICAL STUDY

Herv Lebret, Ecole Polytechnique Fdrale de Lausanne, Switzerland ABSTRACT This study examines more than 2700 companies founded by alumni of Stanford University or having licensed a technology from this university. Stanford University is with MIT one of the most entrepreneurial university in the world, and surprisingly not much data is available on its spin-offs and start-ups. Some important features are described such as the use of venture capital, the dynamics of growth and exits through acquisition or initial public offering. Some characteristics of the founders are also considered such as the time lag between their academic activity and the start-up creation as well as the characteristics of serial entrepreneurs. INTRODUCTION Academic entrepreneurship as well as the role of universities in high-tech entrepreneurship through their alumni has been a much-studied topic in the recent past. Two extensive studies (Shane, 2004 and Djokovic & Souitaris, 2008) illustrate the amount of work done recently. Many of these analyses (Shane, 2004; Roberts, 1991; Hsu et al. 2007; Roberts & Eesley, 2009) were focused on the Massachusetts Institute of Technology (MIT). Other authors (Saxenian, 1994; Zhang, 2003, 2009) have compared the Boston Area and Silicon Valley in particular through the angle of venture capital funding and have shown the critical role of both MIT and Stanford University in academic entrepreneurship. It would be impossible to make here a list of all papers published on the topic and Djokovic has done a very interesting compilation of papers studying spinouts from academic institutions. Another synthesis summarizing lessons learnt on universities and start-ups (Lerner, 2005) was also published after many articles on the topics related to spinoffs and venture capital. Whereas Silicon Valley has been extensively studied (Saxenian 1994, 1999; Kenney, 2000; Lee et al., 2000), it appears that Stanford University has not been studied as much as MIT or many other universities, which have been much less entrepreneurial than Stanford. Here can be mentioned the cases of UT-Austin (Smilor, 1990), the University of Cambridge in the UK (Garnsey & Heffernan, 2005), Oxford University (Lawton Smith & Ho, 2006), ETH Zurich (Oskarsson & Schlpfer, 2008) or the broader subject of universities and venture capital (Zhang,

2009). Stanford has been studied through the limited cases of departments or laboratories (Kenney & Goe, 2004; Jong, 2006, Lebret, 2007) including some unpublished work (Lenoir, 2002). A broader view of Stanford University and its connection with the military-industrial complex is the book by Lowen (1997). Stanford remains however some lesser-known territory that deserves more attention. Start-up and Spin-off The definition of a spin-off has also been the subject of many studies and the debate may not be totally closed. Djokovic and Zhang among others have shown the variety of definitions used. They usually include the transfer of technology and/or people but the definition of transfer of technology may be formal (through a contract or license process) or informal. A good example of the difficulty is the famous example of Google vs. Yahoo at Stanford University (Ku, 2002). Yahoo was not considered as a Stanford spin-off because the two founders built the web site as a hobby on their spare time so that no license was needed from Stanford when Yahoo was moved out of the laboratory to a stand-alone company. Stanford had filed a patent on the PageRank technology, which was licensed to Google when the company was incorporated. Fundamentally, however there was the same transfer of people and there was a similar transfer of technology even if it was not formalized through a patent application in the Yahoo case. If the definition of a spinoff looks quite clear and simple, it does not mean that Yahoo was not created thanks to the university facilities and (cultural) ecosystem. This is another motivation for studying not only Stanford spin-offs but also the related start-ups founded by alumni without a license from Stanford. Venture Capital and Founders of Start-Ups Why are start-ups scrutinized so much? One important reason is certainly the value creation of these companies in the last fifty years. From Intel to Google, and in between companies such as Genentech, Apple, Microsoft, Oracle, Cisco, the American economy has positively benefited from these fast growing companies. Billions of dollars of sales and hundreds of thousands of jobs have been created by a relatively modest number of companies in a very short period of time. Silicon Valley and its informal ecosystem have both been at the origin and the beneficiaries of this value creation that very few other regions on the planet have experienced. Some interesting and rather unique characteristics of these companies have also explained this attention. Venture capital has been a critical tool for the growth of these companies and it has become a structured activity in parallel to the development of the Silicon Valley and Boston technology clusters and in particular their start-ups. However, even if venture capital has been extensively studied, the author is not aware of studies that link start-ups and venture capital in a systematic manner. What about start-ups which do not use venture capital? Do venture-backed companies succeed better than others? The founders of start-ups are also critical. The names of the founders of the companies mentioned above are all famous. Why is this? There are certainly elements of leadership and charisma with start-up founders which may be much more important when companies are small and fast-growing. More importantly, these founders have become role models for the new entrepreneurs. Steve Jobs had Robert Noyce as a mentor, Brin and Page met Andy Grove. Founders are important outside their own companies.

This article does not have the ambition to consider hypotheses that could be validated or not but has the objective of illustrating some features of start-ups (that the author believes are important and even if somehow quite well-known, not always described with facts and figures). It also has the ambition of showing that some of these features might be used as success or growth measures of start-ups: these are venture capital resources, value creation, time-to-exit for example. DATA AND RESULTS This paper studies three different groups of start-ups linked to Stanford University. The first one is the group of start-ups which obtained a license from the Office of Technology Licensing (OTL) of Stanford University (called the spin-offs). The second one is based on a study commissioned by OTL (Leone et al. 1992). The third group known as the Wellspring of Innovation (http://www.stanford.edu/group/wellspring) is a list of companies founded by Stanford Alumni and was retrieved on February 6, 2009 (this web site is an ongoing project). This makes a total of more than 2700 start-ups. Essentially the names of the companies and Stanford founders were available in these lists. We have empirically built consistent data over the three groups: the fields of activities, the resources provided by venture capital and other investors, the year of foundation, the year of a liquidity event if any (Initial Public Offering IPO, Trade Sale M&A or Cessation of Activity). The value creation is also studied in three ways: the sales, the employment and the value creation (market capitalization) when the company is public or the value of the M&A if the company was acquired. The time span between activity at Stanford and creation of the start-up, and between creation and liquidity of the start-up has been studied. The last but related features are linked to the founders: are these serial entrepreneurs? What about their past experience before becoming founders. What about the role of professors as founders? Value creation For the sake of efficiency and space available in this article, we will compare in this first part of our results the spin-offs and a subset of the Wellspring of Innovation (WI). Stanford University generated 204 spin-offs (see Table 1). The number of start-ups in WI which did not belong to the two other groups is 2140. Out of these, 1467 can be considered as high-tech companies as the creators of the WI list also included entrepreneurs in non technical activities such as consulting, finance. Table 1 indicates the fields of activities of the companies and the number of VC-backed companies. Biotechnologies and medical devices (life sciences or LS) represent about 50% of the spin-offs and information technologies (IT) about 45%; however for the WI group, once the non high-tech companies are excluded, LS account for 15% and IT for 85%. A second important comment is that in high-tech, and in both groups, about 50% of the companies are venture-backed. More precisely, since 1985, 4 out of the spin-offs founded each year raised on average $30 million during their lifetime. In the WI group, 26 start-ups founded per year raised $41 million each. Table 2 summarizes the value creation of these start-ups. It also includes the third group not mentioned until now. The group of spin-offs raised a total of $2.9 billion of venture-capital. The acquisitions represent a cumulative value of $8.2 billion and the value of public companies as of October, 3, 2009 was $22.4 billion, excluding Cisco ($131 billion) and Google ($153 billion). The WI group in its high-tech part had respectively $27 billion of venturecapital, $173 billion of M&A and $183 billion of public value. Dynamics of growth The numbers shown in Tables 1 and 2 are not fundamentally new. The value creation of hightech start-ups is well-known, even if it may have not been linked to Stanford in such a systematic

manner. A related feature of this value creation is the speed at which the value is created. The author tried to systematically look for the year of incorporation of the companies as well as the time of exit, if any, i.e. the year of an acquisition, of an initial public offering or of a liquidation. Many studies focus on the survival rate of start-ups (e.g. Oskarsson & Schlpfer, 2008) but Table 3 shows that the dynamics of exits may be much more relevant. Only about a third of the companies remain privately-held and much less (16%) in the VC-backed companies of the WI group. Another third has been acquired (55% of the VC-backed in WI). A smaller group (5 to 15%) is public and the difference is made of liquidated companies. Figures 1 and 2 show the time to liquidity in years of the companies which are not privately held anymore. For the spin-off case (Figure 1), the number of companies is 61 for the VC-backed ones and 31 for the others. The average number of years to liquidity is 5.97 for start-ups with VC money and 6.55 for the others, the overall average being 6.17 years. The WI group (Figure 2) has 630 VC-backed companies and 574 start-ups in the second group. The average time to liquidity is 5.3 years for the first subgroup and 8 years for the other one, with an overall average of 6.6 years. One clear difference between spin-offs and start-ups is the impact of non-tech companies. First, as Figure 4 shows, 55% of the WI non-tech companies are still private (vs. 21% of the WI high-tech ones). Secondly, the average time to liquidity for non-tech companies is 10 years. Founders Founders are a critical component of companies. However, no formal definition exists. Founders should not be confused with entrepreneurs who usually work in the companies they start nor with managers who may not be part of the founding team, even if some were early employees. The only simple definition of the group of founders is the group of people who recognize themselves as such. The author could identify 2711 unique names of individuals for the 2727 companies (Table 4). However, for 62 companies, no founder could be identified as a member of the Stanford community (i.e. professor, staff or alumnus). 2203 companies had one Stanford founder and 462 had more than one. Professors are active founders: 167 unique professors were founders in 243 companies. In 140 of these companies, they were the only Stanford founder. Only 82 of these companies had a license and therefore belong to the spin-off group. The experience of founders is one of their key features. It is usually illustrated by their prior professional activity or their age. The author did not have access to such information. However, thanks to data available through the Stanford Alumni association, it was possible to find when a founder graduated from Stanford University and therefore obtain the number of years between the activity at Stanford (graduation year if an alumnus or latest date of activity if a professor or staff) and the year of incorporation of a company. In the case of the spin-offs however no information could be obtained for 42 out of 204 companies (some of these companies do not have Stanford founders) and 163 companies had missing information out of the 2523 other start-ups. When a company had several founders, the time difference was taken as the year of foundation minus the average of the activity years of all founders. Figure 4 is the histogram of these time differences for the spin-offs. Even if the average value is 2 years, it appears clearly that a majority of spin-offs is created by individuals active at Stanford at the time of incorporation. Figure 5 gives the information for the start-ups, i.e. the WI and the 1992 study. The average is 9.2 years. Figure 5 shows two interesting features: first, nearly 250 companies (a little under 10% of the group) were created at year 0; the numbers decrease immediately around 100 for years 1-3 but increase again for years 4-6, then decrease smoothly thereafter. Figure 6 shows the three groups (spin-offs and start-ups) by field of activities. High-tech companies are created in the range of 6-8 years (e.g. 5.7 for biotech, around 7 for medical technologies or medtech, semiconductor, IT, software and

Internet) where as non-technical are above 10 years (about 12 for finance and non-technical services). Serial entrepreneurship (founders in our study) is an interesting topic as it is regularly mentioned in the general press. Many investors claim they favor entrepreneurs with experience, even if they have failed in their prior venture. The literature is surprisingly not very rich and two recent works (Bengtsson, 2008 and Gompers et al., 2009) focus on venture-backed companies but seem to reach slightly different conclusions. This article does not focus on serial entrepreneurship only and a dedicated article is under preparation. However, it is of interest to show some results. Among the 2711 founders, 445 individuals (including 44 professors) created more than one company. The total number of companies launched by these serial founders is 988 (some companies have several serial entrepreneurs). Table 5 compares the resources and value creation of companies which did not have serial entrepreneurs, as well as the first, second, third and fourth companies created by the serial founders. The results tend to show that serial entrepreneurs have more resources with their new ventures in terms of venture capital money but do not create on average more value with their new companies than with the prior ones or compared to one-time entrepreneurs. The only exception is the M&A value of the second ones compared to that of onetime entrepreneurs. DISCUSSION AND CONCLUSION The results of the analysis of Stanford high-tech entrepreneurship are manifold. The value creation is exceptionally high thanks, in part, to a very high level of venture-capital money. Venture capital does not explain alone this success. Hewlett-Packard belongs to the 1992 study, it is the largest of the companies considered in the three groups and was not financed by venture capital. However the levels of money raised (on average more than $30 million for the companies accessing VC money) are high and should be an indicator for many academic spin-offs of the levels of resources used to succeed. The value creation is also extremely high. Table 2 shows the amounts of sales and employment generated by the existing public companies. In high-tech, the sales were close to $350 billion in 2008 and the employment was close to 1 million jobs. Nevertheless the top 5 tech companies generated about 2/3 of the value creation and the top 10, about 3/4. Many smaller companies contribute to this creation but are not small companies even if they are called start-ups. The resources used by these top 5 and top 10 companies in terms of venture-capital are relatively much smaller. Finally, a total of 1050 companies could be identified as VC-backed out of the 2727. The value creation of this subset is $186 billion in M&A and $543 billion in public value (against $82 billion of M&A and $286 billion for those which were not associated with venture capital). Table 1 also shows which fields are most developed by the spin-offs and financed by venturecapital: they are the same. Indeed life sciences and information technologies represent 99% of the spin-offs as well as the companies backed by venture capital both in the spin-off and WI groups. An interesting feature of spin-offs is the very high share of life sciences, more than 50% of the spin-offs and the VC-backed companies whereas they represent less than 15% of the WI group! There is no doubt that intellectual property licensed from universities play a role in this difference. Many studies (e.g. Oskarsson & Schlpfer, 2008; Roberts and Eesley, 2009) do not show the same field repartitions. This feature may explain why some universities do not experience the same ratios of fast growing companies. Growth is not measured only by the available resources. Time to exit is another such measure. Whereas some studies emphasize the survival rates of start-ups as a measure of success, this article

shows that fast growth is a general feature of Stanford high-tech companies. More surprisingly maybe, biotech companies do not show longer time to exits. This apparent mystery may be easily explained by the fact that many biotech companies go public without any sale at an early stage of their development. However even if some non VC-backed companies are slower to exit (medtech or electronics), others are as fast with or without VC Money (software and biotech) as Figures 1 and 2 show it. As interesting to illustrate the fast growth is the number of privately held companies. A note of caution is necessary: the spin-off and WI groups are different in nature; spin-offs include companies founded until 2008 whereas the WI group includes companies founded before 2005, therefore the dynamics of exits are obviously different for the most recent companies. Furthermore, the WI group includes non-tech companies which may survive more easily and longer with revenues generated from their customers in sectors such as finance or nontech services (Figure 4). Therefore even if the explanations for the number of private companies may be diverse, the numbers are very low in both cases. High value creation and fast growth thanks to adequate resources appear clearly. These results should not be surprising. The contribution of Silicon Valley to the American economy is well documented but what may have been often neglected is how much a university may directly (licenses) or indirectly (alumni) contribute to a dynamic ecosystem. Similarly to MIT for the Boston Area, Stanford is a major contributor to Silicon Valley, but it less clear that other universities or clusters have been as successful. The correlations between lesser entrepreneurial regions and their university spin-off success might be similar. In a much smaller-scale study, it was described how some US start-ups went public five years on average after their incorporation, whereas it took ten years for European start-ups to become public (Lebret, 2007). The classical 5 to 7 years that venture capitalists look for as a holding horizon when they invest in start-ups could also explain the results. In terms of recommendations for metrics and benchmarks of academic innovation, the dynamics of growth and value creation should therefore be used and not only quantitative measures such as company creation or patenting and licensing activities. The level of resources used by start-ups may be an indication that many non-US start-ups are undercapitalized to succeed. The characteristics of the founders represent the second part of this study. One key feature is the founders experience after leaving or not Stanford University. A little less than 340 companies were created at year 0, another 360 companies were created between year 1 and year 3 of the average activities at Stanford (average on all founders) and 390 between years 4 and 6. Companies created at year 0 (12% of total) represent about 50% of the value of public companies, 25% of the M&A value and 28% of the VC money. The second group (years 1-3) and third group (years 4-6) represent respectively 13% and 14% of the number of start-ups, 6% and 20% of the public value, 13% and 20% of the M&A value, and 12% and 15% of the VC money. Though these comments would require a deeper analysis, companies created at year 0 seem to create much more value than others. Experience may not matter so much as possibly the quality of the technologies. This could be correlated to the young age of many extremely successful entrepreneurs in Silicon Valley. Experience of founders may not be a fundamental requirement. In terms of research on technology transfer, the comparison of Figures 4 and 5 indicates that formalizing the spin-off creation through licensing may be a necessary process but probably insufficient in describing the value creation of universities. Though difficult to quantify, it is likely that many start-ups created at year 0 in Figure 5 include companies which could have been considered as spin-off similarly to the YahooGoogle analogy considered in our introduction. A final and interesting feature is the serial entrepreneurship factor. Whereas the general agreement seems to claim that serial founders would be important because of the experience they

bring on board of start-ups, the results of this study does not seem to confirm this general belief. What is also interesting is that serial entrepreneurs are more successful with their first companies than one-time entrepreneurs and on average they raise less VC money. However the situation is inverted with the following ones with two exceptions: the average M&A value of the 2nd one is still higher than that of one-time entrepreneurs and the VC money raised by the 4th ones is smaller! It is also worth noticing as described in the previous paragraph that most of the value creation seems to be linked to the time proximity to Stanford. Some explanations have been given that serial entrepreneurs could be over-optimistic and less motivated. One other explanation might be that more disruptive (and therefore promising) technologies belong to companies close to Stanford. One cannot avoid thinking that the luck factor may have an important role in high-tech entrepreneurship. The value creation would therefore be explained by the statistical effect of the large number of ventures. As a conclusion, we would like to propose some areas for future research and also mention some limitations and difficulties linked to the data. With the exception of public companies which disclose an enormous and rich amount of information in their SEC documents and in particular in their IPO prospectus, high-tech start-ups do not disclose much information. It is the authors experience that information is not only difficult to find for private companies but it is also sometimes doubtful. Money raised through venture capital, value of M&A transaction should be treated with some caution. A company may announce numbers which are not always accurate (because of milestones-based financing as an example). The revenue and employment figures were provided on the basis on public companies only and the author considered that private companies have lower numbers, therefore the numbers give values lower than the real ones. Similar difficulties arise with founders as we mentioned earlier in our introduction. Who knows that Apple Computer did not have two founders (Wozniak and Jobs) but three (with the addition of Ronald Wayne)? Building a database of founders and start-ups was not an easy task and the author will not claim that it is void of mistakes or inaccuracies. Stanford founders are not the only founders of these companies, which is another limitation of the study. There might also be a bias in favor of successful companies and founders who may be easier to identify so that this may explain a rather high level of success rate. When it was possible, all data were double checked but the author recognizes that there are strong limitations. It is an intuition he had when observing the start-up world and its studies. The examples of experiences of founders and serial entrepreneurship are illustrations that general beliefs may have to be reconsidered. In terms of recommendations and future research, the author believes that the present work may help in reassessing what could be benchmarks and good metrics for (academic) start-ups in terms of value creation and growth. Stanford University and MIT are obviously exceptional universities, but because innovation is a global phenomenon, it is not clear why other universities should not measure their results according to the performance of these two institutions. One interesting work might be to compare MIT and Stanford and analyze how similar or different they are and if so why. Another possible study that the author did not have data to analyze would be the age of the founders and not their experience. The topic of high-tech entrepreneurship is fascinating and even if decently well-known, opened to many new directions of research. What is the impact of venture capital in the value creation? What is the real impact of professors (vs. their students) in academic start-ups? What is the role of luck vs. experience? Are there areas of activities which are more favorable to start-ups than others? The author believes that he has contributed in a modest but valuable manner to a better understanding of the dynamics of high-tech entrepreneurship. CONTACT: Herv Lebret, herve.lebret@epfl.ch; (T): +41 21 693 7054; (F): +41 21 693 14 89; EPFL, 1015, Lausanne, Switzerland

ACKNOWLEDGEMENTS The author would like to thank Katarine Ku, head of the Office of Technology Licensing at Stanford University for providing data on the spin-offs and the links to the 1992 study. He would also like to thank the organizers of the 2009 Society for Entrepreneurship Scholars Meeting sponsored by the Marion Ewing Kaufman Foundation, where data used in this study were first shown and discussed. REFERENCES Bengtsson, O. (2008). Relational Venture Capital Financing of Serial Founders. Under Review. Carey, P. (2001, December 1). A Start-Up's True Tale: Often-told story of Cisco's launch leaves out the drama, intrigue. San Jose Mercury News. Di Gregorio, D. & Shane, S. (2003). Why Do Some Universities Generate More Start-ups than Others. Research Policy 32, 209-227. Djokovic, D. & Souitaris, V. (2008). Spinouts from academic institutions: a literature review with suggestions for further research. Journal of Technology Transfer 33, 225247. Garnsey, E. & Heffernan, P. (2005). High-technology Clustering through Spin-out and Attraction: The Cambridge Case. Regional Studies 39 (8), 11271144. Gibbons, J. F. (2000). The Role of Stanford University: A Dean's View. In: Lee, C., Miller, W., Hancock, M., Rowen, H. (Eds.) The Silicon Valley Edge (pp. 200-217). Stanford, CA: Stanford University Press. Gompers, P., Kovner, A., Lerner, J., & Scharfstein, D. (2009). Performance Persistence in Entrepreneurship. Harvard Business School. Working Paper 09-028 Hsu, D. H. Roberts, E. B. & Eesley, C. E. (2007). Entrepreneurs from Technology-Based Universities: Evidence from MIT. Research Policy 36, 768788. Jong, S. (2006). How organizational structures in science shape spin-off firms: the biochemistry departments of Berkeley, Stanford, and UCSF and the birth of the biotech industry. Industrial and Corporate Change 15 ( 2), 251283 Kenney, M. (2000). Understanding Silicon Valley. Stanford, CA: Stanford University Press. Kenney, M. & Goe, W.R. (2004). The role of social embeddedness in professorial entrepreneurship: A comparison of electrical engineering and computer science at UC Berkeley and Stanford. Research Policy, 33, 691 - 707. Ku, K. (2002). Software Licensing in the University Environment. Computing Research News 14 (1), 3,8. Lawton Smith, H. & Ho, K. (2006). Measuring the performance of Oxford University, Oxford Brookes University and the government laboratories spin-off companies. Research Policy 35, 15541568. Lebret, H. (2007). Start-Up, What We May Still Learn From Silicon Valley. Scotts Valley, CA: CreateSpace. Lee, C. et al. (editors), (2000). The Silicon Valley Edge. Stanford CA: Stanford University Press. Lenoir, T. (2002). The Stanford Startup Report project - Inventing the Entrepreneurial University: Stanford and the Co-Evolution of Silicon Valley. Retrieved from <http://www.stanford.edu/dept/HPS/TimLenoir/Startup/QuarterlyRpts> Leone, A., Vamos, J., Keeley, R., & Miller, W. (1992). Technology-based Companies Founded by Members of the Stanford Community. A Study Commissioned by the Stanford University Office of Technology Licensing.

Lerner, J. (2005). The University and the Start-Up: Lessons from the Past Two Decades. Journal of Technology Transfer 30 (1/2) 4956. Lowen, R. (1997). Creating the Cold War University. Berkeley, CA: University of California Press. Oskarsson, I. & Schlpfer, A. (2008). The performance of Spin-off companies at the Swiss Federal Institute of Technology Zurich. Zurich, Switzerland: ETH Transfer. Roberts, E., B. (1991). Entrepreneurs in High Technology: Lessons from MIT and Beyond. Oxford, UK: Oxford University Press. Roberts E. B. & Eesley, C. E. (2009). Entrepreneurial Impact: The Role of MIT. Kansas City, MO: the Ewing Marion Kauffman Foundation. Saxenian, A. (1994). Regional Advantage. Cambridge, MA: Harvard University Press, Saxenian, A. (1999). Silicon Valley New Immigrants Entrepreneur.s San Francisco, CA: Public Policy Institute of California. Shane, S. (2004). Academic Entrepreneurship, University Spinoffs and Wealth Creation. Cheltenham, UK: Edward Elgar. Smilor R., Gibson D., Dietrich G., (1990) University Spin-Out Companies: Technology Start-Ups from UT-Austin. Journal of Business Venturing 5, 63-76 Zhang, J. (2003). High-Tech Start-Ups and Industry Dynamics in Silicon Valley. San Francisco, CA: Public Policy Institute of California. Zhang, J. (2009). Why do some US universities generate more venture-backed academic entrepreneurs than others? Venture Capital 11 (2), 133162.

Table 1: Stanford Spin-offs and Start-ups

Fields of activity All Biotech Medtech Computers Semiconductor Electronics Telecom IT & SW Internet Energy Env Manufacturing Eng. Services Others Unknown Subtotal (high-tech) Total 204 100% 100 100% 11 5% 1 1% 3 1% 1 1% 40 20% 26 26% 42 21% 19 19% 72 34 2 Spin-offs % 35% 17% 1% VCbacked 38 13 2 % 38% 13% 2% All 66 113 30 106 114 184 310 356 23 20 145 628 45 1467 2140 % all 3% 5% 1% 5% 5% 9% 14% 17% 1% 1% 7% 29% 2% 69% 100% 100% Wellspring of Innovation % hightech 4% 8% 2% 7% 8% 13% 21% 24% 2% 1% 10% VCbacked 41 64 18 70 44 129 153 211 0 1 6 24 0 737 761 % all 5% 8% 2% 9% 6% 17% 20% 28% 0% 0% 1% 3% 0% 97% 100% 100% % hightech 6% 9% 2% 9% 6% 18% 21% 29% 0% 0% 1%

Table 2: Value Creation of Stanford Start-ups and Spin-offs

Group Stanford Spin-offs 1992 Study (not including licenses) Wellspring of Innovation (tech. only) Total-Tech WI (non tech) Total Top 5 high-tech Number 204 383 VC ($M) 2'969 1'842 M&A ($M) 8'214 75'406 Public ($M) 307'136 185'175 Sales ($M) 65'410 171'579 Jobs 105'281 454'082

1467 2077 673 2727

27'125 31'936 272 32'208 1'719 5%

173'375 256'995 11'892 268'887 75'800 29% 110'200 43%

183'615 675'926 154'413 830'339 445'000 66% 497'010 74%

111'696 348'685 46'348 395'033 223'929 64% 259'828 75%

401'453 960'816 204'895 1'165'711 603'528 63% 680'087 71%

Top 10 high-tech

2'794 9%

Table 3: Status of spin-offs and start-ups Status All Public Private M&A Ceased Unknown Total 8% 39% 29% 16% 8% 204 Spin-offs VC-backed 14% 37% 36% 13% 100 Wellspring of Innovation All VC-backed 5% 33% 34% 21% 7% 2140 10% 16% 55% 19% 754

Figure 1: Number of years from foundation to liquidity event of spin-offs

Unknown Energy Env IT&SW Electronics Computers Medtech Biotechnology 0 2 4 6 8 10 12

NoVCor unknown All VCbacked

Nb. of companies Unknown Energy Env. IT & SW Electronics Computers Medtech Biotech Total

VC backed

No VC or unknown 2 3

16 16 2 2 25 61

5 6 0 4 11 31

Figure 2: Number of years from foundation to liquidity event of WI companies

Consumergoods Finance Nontechservices Eng.Services Others Manuf. Energy Env Internet IT&SW Telecom Electronics Semiconductor Computers Medtech Biotech 0 2 4 6 8 10 12 14 16 18 20
WithoutVC All VCbacked
Nb. of companies Consumer Goods Finance Non Tech Services Engineering services Other tech Manufacturing Energy env. Internet IT & SW Telecom Electronics Semiconductor Computers Medtech Biotech Total VC backed No VC or unknown

8 0 2 5 0 1 0 177 125 107 35 65 18 51 36 630

45 52 67 37 5 10 7 102 93 34 43 26 8 27 18 574

Figure 3: Status of WI companies by field of activities (N=2140)

100% 90% 80% 70% 60% 50% Unknown 40% 30% 20% 10% 0% Public Private M&A Ceased

Table 4: Founders by company (including professors) Stanford founders All Founders including one professor by company Companies Individuals Companies Individuals 0 1 2 3 4 5 6 7 8 9 Total Unique names 62 2203 300 113 29 12 5 1 1 1 2727 0 2203 600 339 116 60 30 7 8 9 3372 2711 140 49 38 9 5 1 1 243 140 52 43 16 6 1 2 260 167

Figure 4: Histogram of time between activity at Stanford and spinoff foundation (N=142).
100 90 80 70 60 50 40 30 20 10 0 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 36 Years

Figure 5: Histogram of time between activity at Stanford and start-up foundation (N=2523)
250

200

150

100

50

0 17 15 13 11 9 7 5 3 1 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 Years

Figure 6: Average time between activity at Stanford and foundation of start-ups (by field)
Total Consumergoods Finance Nontechservices Eng.Services Manuf. Energy Env Internet IT&SW Telecom Electronics Semiconductor Computers Medtech Biotech 0 2 4 8.86 9.81 13.04 12.63 9.97 9.59 11.95 6.95 7.68 9.37 8.8 7.11 6.43 7.05 5.71 6 8 10 12 14

Table 5: Value Creation by Serial Founders

Data on non-serial VC-backed 1739 Number 474 Data on serial 988 VC-backed Number 386 1st comp 445 VC-backed Number 147 2nd comp 445 VC-backed Number 202 3rd comp 128 VC-backed Number 57 4th comp 46 VC-backed Number 23 Average Average Average Average Average M&A Number Average Public Number Average $4'929'000'000 370 Ceased Average $5'955'000'000 232 Ceased Average $11'934'000'000 92 Ceased Average $3'371'000'000 107 Ceased Average $2'324'000'000 39 Ceased Average $1'109'000'000 12 Ceased

$36'081020 253 M&A Number

$520'000'000 102 Public Average Number

$39'132'000 220 M&A Number

$624000'000 55 Public Average Number

$28'466'000 120 M&A Number

$900'000'000 30 Public Average Number

$42'042'000 93 M&A Number

$617'000'000 20 Public Average Number

$54'251'000 18 M&A Average Number $38'867'000 13

$277'000'000 6 Public Average Number $165'000'000 3