Category Archives: Start-up data

nVidia, the new giant

Leave a reply

nVidia has made the headlines recently as its stock value jumped by 25% to reach a valuation close to $1T ($1’000B) joining a small club of companies generally called the GAFA(M) or BigTech. I knew nVidia as just another Silicon Valley success story, a big one, but just one more. It belongs to my 800+ startup list and here is my typical cap. table.

Nvidia was founded in 1993 by Jen-Hsun “Jensen” Huang, Curtis Priem, and Chris Malachowsky and is headquartered in Santa Clara, California.


There would be so many little things to mention about how typical it is, but here are a few:
– The founders were young engineers (29, 33 and 33), one from Stanford University, the two others from solid even if lesser known schools. One is of Taiwenese origin. They worked in big tech companies before founding their startup, and they are still leading it. They had equal ownership at foundation.
– There was a typical support of venture capital, a total of $20M in 4 rounds between 1993 (the foundation) and 1997 (the IPO), followed by an IPO in 1999, less than 6 years after the incorporation. The VCs were Sequoia (which also funded Apple and Google), and Sutter Hill. The board included experts from Synopsys (its cofounder) and Avid.
– Employees owned at least 20% of the company through stock options (and maybe even 35%+ throug additional common shares).
– It went public at a $500M valuation, more than decent and was a leader in computer graphics chips until nVidia applied its technology to AI. Hence its current popularity.

Equal ownership of founders in startups ?

Leave a reply

Yesterday I had a short debate about Wozniak and Jobs initial ownership in Apple Computer. It is true that at the IPO Wozniak had much less ownership than Jobs, but this can be explained by the fact that he gave or sold at a low price shares to employees (whom he thought deserved it and Jobs did not). But at the origin, they had equal shares as the extract from the prospectus shows.

So I thought of having a look at my startup database (currently having 890 cap. tables) and I studied the numbers. Here they are:

So what are the lessons?

First majority of startup have between 1 and 3 founders, and 1 founder (contrarily to intuition maybe) is not so rare. Now there is a caveat: the history of a startup is never fully known. Apple had initially (and for 2 weeks) 3 founders! The third one was Ronald Wayne

Second, equal ownership is not the majority but it is not rare. Around 15-20%.

But this does not mean, one founder owns more than 50%. Of course yes with 2 founders. But for 3 founders, this happpens in 41% of the cases. When more than 3 founders, this is 31% of the cases. I did not check (yet) if geography or fields of activities have an impact…

Finally, if you read this blog, you should know that statistics do not say it all. Startups are a world of exceptions (and the statistics are seldom Gaussian but follow a power low, so beware of means of %). Therefore more anecdotally, but still important, here are some famous examples:

Famous startups – 2 founders with equal shares
Adobe
Akamai
Apple
Atlassian
Broadcom
Cisco
Genentech
Google
Intel
Netscape
Riverbed
Skype
Soitec
Spotify
Tivo
Yahoo
Zalando

Famous startups – 3 founders with equal shares
Airbnb
Checkpoint
Compaq
DoubleClick
Equinix
Marimba
nVidia
Palantir
Revolut
RPX
WeWork

Famous startups – 3+ founders with equal shares
AMD
Regulus
ROLM
Xiaomi

Famous startups – founders with non-equal shares
Cypress
DropBox
Etrade
Eventbrite
Facebook
Lyft
Microsoft
Mysql
Oracle
Pinterest
Salesforce
Sun Microsystems
Twitter
Uber

Figma, a new cap. table and there is much more to Dylan Field’s story

Leave a reply

Figma is the latest startup success story. Not an IPO, there are so few in 2022, but a $20B acquisition by Adobe. I did not have much data to build its cap. table so this is mostly tentative. Still it is interesting. So here it is. However there is much more to the story of its founders Dylan Field and Evan Wallace. Read below.

Not much to add to the numbers except the founders (including possibly some earnout shares) & investors stake, 20% & 50% respectively as well as the time it took for all this. A few months for seed, 10 years for an exit.

Photographer: David Paul Morris/Bloomberg

So let us have a look more specifically at Dylan Field. Again the typical even if rare school drop-out in his early twenties who is still the CEO 10 years later. His cofounder is a friend. The rest is history. Well not really. Read his Wikipedia page for more or this article from Business Insider.

First, Field received the $100k that Peter Thiel was offering to young students ready to drop out of school. Field’s parents were against the idea but Field dropped out of Brown University. I have always been intrigued with the idea of pushing people out of school. Will Field ever go back there?

Second, he found some VC money despite the fact that Field recalled that he experienced a “wake up call” when [a potential] investor turned down the chance to invest in Figma’s seed round and said, “I don’t think you know what you’re doing yet.”

Third, he remained as CEO despite his lack of business experience. At one point early into Figma’s existence, Field said he once faced the very real risk of an exodus of disaffected employees. He had to learn, quickly, how to be more open to feedback and to empower his teams, while also hiring experienced managers. “I didn’t know how to manage effectively,” he said. “I didn’t know the basics around how to have good judgment around who to hire. When we were 10 people, I was a year into management. Usually if you are a new manager, you manage a few people. I was trying to do this at the same time and get the product to market.” Apparently he survived a few crises and the VCs let him lead.

There is certainly much more to learn from this unique story, but it is enough here. One final point. I would love to know more about VC performance. I worked at Index in the early days so I learnt that a great success does not guarantee a fund performance. But here Index made apparently more than $2B and made at least a 400x multiple in the seed part of its investment in Figma. But information about VC performance is close to impossible to find…

More (interesting?) data about French unicorns

Leave a reply

A month ago, I published data about French startups. I had been surprised to discover that access to data about private companies was finally possible for free in my dear country. So I looked at some (famous) French unicorns with an interest in the shareholder structure and how much money they had raised overall, as well as in their seed and A rounds. You will find the detailed information in a pdf in the bottom of the post.

But before moving on to this analysis, I want to mention an excellent article on seed fundraising, which gives advice and quite rich information. It is in French though and is entitled La levée de fonds seed ou amorçage. So here are the results:

In this first table, I just had a look at their age and fund raising. To give simple rule of thumbs, about the ones which are still private, they are about 5 to 15 years old, they have raised about €200M, with seed rounds of €0.5M and A round of €2-3M. The market capitalization should be (by definition) above one billion euros, but apparently this is not always the case (let us say that the value of a private company is a very volatile metric!) and the ratio of this value to amount raised seems also to be 5 to 15…

I then looked at how much dilution the seed and A rounds correspond to as well as the age of the companies for these rounds. Again, not taking outliers into account, both the seed and A rounds seem to induce a 25% dilution, therefore, with rounds of €0.5M and €2-3M respectively, the value at seed is about €2M and at A round is €8-12M. Finally the startups are less than 1-2 years old at seed and less than 4 years old at A round.

The last table is about the shareholding or equity structure as well as some data about the founders. The founders keep 25-30% of their startups, investors have 60-65% whereas employees have 5-10%.

There are about two founders per startup, they are surprisingly often below 30 with a median and average age of 29 and sadly with not a single woman.

Equity List – French Unicorns

New data about academic technology transfer to startups

Leave a reply

Nathan Benaich is very unhappy with technology transfer in the UK and he is probably right to be. For many years I had noticed that British academic institutions often took more than a 25% ownership in a startup in exchange for a license of intellectual property, whereas the standard figure in the USA and continental Europe is more in the 5-10% range. He had published a very interesting article in May 2021, Rewriting the European spinout playbook where he was complaining about a lack of transparency and very frustrating processes.

He is now working on a new set of data provided by founders that he makes freely available on spinout.fyi. He is asking for help and any interested founder should provide a little if she can. I downloaded his data and provide here my own analysis although Nathan has his own here. You should read it. Here is a first set of tables:

If you do not like tables and even if you do, here are more figures:

And because I had done a similar research some years ago, posted here as How much Equity Universities take in Start-ups from IP Licensing?, I did the exercise of combining his and my data. This is a set of 190+ companies! You will see the equity ownership per domain and per region.

So what are the lessons? The UK is a clear outlier, but what is more striking is the volatility in the numbers. And why is that? Some professional claim each startup is different. I disagree. Strongly! The lack of transparency in the policies is the reason of the volatility. Founders seldom know how they will be treated. This is why I was so happy with EPFL publishing its policy. See my recent post Technology Transfer according to EPFL and Rules for Startups.

I really hope that Nathan Benaich’s effort will help in bringing a much needed transparency in these numbers!

Access to French startups data

Leave a reply

I should have known sooner about new rules on data about French startups. In the past, you had to pay on sites like societe.com or Euridile to get filing documents of private companies from the register of companies. This is the past! Now it is possible to access this data for free. And this is great news. So my favorite exercise which consists in building cap. tables of startups, which had become a habit for companies going public, for Swiss companies in certain cantons like Zurich or Basel, or for British companies thanks to Companies House is now possible in France with Pappers.

I obviously tried with some of the current famous private startups. I failed with Dataiku, probably because it moved to the USA, but could build some partial tables for Doctolib, Mirakl, Alan, Ledger and BlaBlaCar. It is far from perfect because you need to read many documents. I had to go through 68 ones for BlaBlaCar. I did not go into the details of stock options, granted or exercised. But I could get the info about the founders and the funding rounds. Here is a summary:

and here are the individual tables. QUite fascinating to see the recent trends in France through 5 examples:

The (Recent) Impact of Venture Capital on Startups

Leave a reply

I have regularly been puzzled with the (real) impact of venture capital on startups, their growth or even their success. A few days ago, I received an email from a friend with a very interesting table.

The measure of capital productivity is given by the ratio a/b where a is the startup revenue at the time of IPO and b is the amount of venture capital raised by the startup before going public. The 4 big tech companies are Apple, Microsoft, Amazon and Google, 4 companies founded before 2000. The ratio a/b is greater than 10. The recent VC or IPO deals give ratios below 1 and closer to 0.1.

So it was easy for me to look at my usual cap. tables (check here if you do not know what I talk about). I have now 880 companies and I sorted this a/b ratio over time. Here is the result:

I have 272 biotech startups and 361 in the Software and Internet fields (the others are hardware, semiconductor, energy, medtech companies mostly). I have the a/b ratio (which I call Sales to VC) by period of 5 years (the years of foundations of the startups). I also pu the PS ratio (the famous Market Capitalization to Sales or “Price to Sales” ratio). Indeed the ratio is “collapsing” from above 5 before the 90s to below 1 after 2000.

I have also separated biotech which is known to have startups going public with low revenue and the group of Software and Internet. The curves which follow are probably better illustrations. Quite striking!

I also looked at the profits (or losses) of the startups and computed the Profit to VC amount ratio. Here it is:

Biotech is different as companies were rarely profitable when going public and the ratio is quite stable since 1990. But overall, companies were profitable at IPO before 2000 (and sometimes highly profitable, so that I could not find a good axis scale for my figure). They are losing money since 2005 and apparently losing more and more.

All this is no real surprise. Mallaby in his recent book The Power Law has described the new trends in venture capital. Funds like Softbank Vision or Tiger Global are pouring tons of money in startups which try to capture market dominance, whatever the cost. So the capital productivity is decreasing at IPO with the hope of huge gains in the future. A very, very risky bet…

PS (dated April 22, 2022) : I was asked a question about startups in the energy / greentech field. This is indeed interesting. Couple of comments before providing an answer from my data. Greentech has never been a stable or profitable segment. Kleiner Perkins or Khosla Ventures, early entrants in the field, seem to have suffered a lot. In addition I have only 21 companies in my DB. You are right, it is stable but from the low range, with low sales to VC ratios and negative profits…

La question était : On a different note you have data only on Energy/green tech? what would you expect to see? I was wondering if for capex intensive businesses the trend is more weak as they already needed to raise a lot of capital.

Silicon Valley will soon be 65. Should it be Retired ? – The Darwinian Dynamics of the Region

1 Reply

Silicon Valley will soon be 65. 65? Yes, I usually say that the region began its growth with the foundation of Fairchild Semiconductor in 1957 (even if the name itself was created in 1971).

endeavor-insight-sv-2-retina

The region is increasingly criticized for both good and bad reasons (see for example here and there) and perhaps it is a little out of breath. Too old ? Ten years ago I had looked at its “Darwinian dynamics” in Darwinian and Lamarckian innovation – by Pascal Picq. I had noted there the remarkable dynamics of creation (and destruction) of businesses. “Twenty of the top 40 SV companies in 1982 did not exist anymore in 2002 and twenty one of the 2002 top 40 companies had not been created in 1982.” So I just did the exercise again.

The table below gives the data for 1982 and 2002 again, then those for 2020. I should have waited for 2022 and the 65th birthday of Silicon Valley, but I didn’t have the patience! Ten of the 40 largest companies did not exist in 2000 and seven more did not exist in 1995. Sixteen of the top 40 of 2002 no longer exist in 2021. The region is therefore a little less dynamic but it remains quite remarkable… The retirement seems to me far away in reality !

As a final comment, five years ago, I had mentioned the evolution of the American capitalism in The top US and European (former) start-ups and in particular The Largest Companies by Market Cap Over 15 Years. You could compare it with the dynamics of French CAC40.

Forty Largest Technology Companies in Silicon Valley
(the same data are provided in jpg format at the end of the post)

1982 2002 2021 Revenue Market Cap
1. Hewlett-Packard 1. Hewlett-Packard 1. Apple $294,135 $2,153,363
2. National Semiconductor 2. Intel 2. Alphabet e $182,527 $1,169,351
3. Intel 3. Cisco b 3. Facebook d $85,966 $787,268
4. Memorex 4. Sun bc 4. Intel $77,867 $194,491
5. Varian 5. Solectron c 5. HP Inc. $57,667 $31,545
6. Environtech a 6. Oracle 6. Cisco $48,026 $192,007
7. Ampex 7. Agilent b 7. Oracle $39,403 $191,539
8. Raychem a 8. Applied Materials 8. Tesla d $24,578 $77,574
9. Amdahl a 9. Apple 9. HP Enterprises $26,866 $15,677
10. Tymshare a 10. Seagate Technology 10. Netflix e $24,996 $236,117
11. AMD 11. AMD 11. Gilead $24,689 $74,058
12. Rolm a 12. Sanmina-SCI 12. SYNNEX $23,757 $6,588
13. Four-Phase Systems a 13. JDS Uniphase c 13. PayPal e $21,454 $277,047
14. Cooper Lab a 14. 3Com c 14. salesforce.com e $21,252 $208,200
15. Intersil 15. LSI Logic 15. Applied Materials $18,202 $78,716
16. SRI International 16. Maxtor b 16. NVIDIA $16,675 $328,615
17. Spectra-Physics 17. National Semiconductor c 17. Western Digital $16,327 $16,183
18. American Microsystems a 18. KLA Tencor 18. Adobe $12,868 $241,275
19. Watkins-Johnson a 19. Atmel b 19. Uber d $12,078 $93,549
20. Qume a 20. SGI c 20. Lam Research $11,929 $69,264
21. Measurex a 21. Bell Microproducts bc 21. eBay e $10,713 $36,576
22. Tandem a 22. Siebel bc 22. AMD $9,763 $115,364
23. Plantronic a 23. Xilinx bc 23. Square d $9,498 $106,173
24. Monolithic 24. Maxim Integrated b 24. Intuit $7,717 $99,872
25. URS 25. Palm bc 25. Opendoor d $7,324 $1,221
26. Tab Products 26. Lam Research 26. Sanmina $6,875 $2,117
27. Siliconix 27. Quantum c 27. KLA Tencor $6,073 $40,492
28. Dysan a 28. Altera bc 28. Equinix e $5,999 $63,238
29. Racal-Vadic a 29. Electronic Arts b 29. Electronic Arts $5,670 $41,368
30. Triad Systems a 30. Cypress Semiconductor bc 30. NetApp $5,590 $14,480
31. Xidex a 31. Cadence Design b 31. Agilent $5,530 $36,607
32. Avantek a 32. Adobe Systems b 32. Intuitive Surgical e $4,551 $92,762
33. Siltec a 33. Intuit b 33. ServiceNow d $4,519 $110,315
34. Quadrex a 34. Veritas Software bc 34. Juniper e $4,445 $7,478
35. Coherent 35. Novellus Systems b 35. Workday d $4,318 $57,934
36. Verbatim 36. Yahoo bc 36. Synopsys $3,821 $39,023
37. Anderson-Jacobson a 37. Network Appliance b 37. Autodesk $3,790 $67,066
38. Stanford Applied Eng. 38. Integrated Device 38. Palo Alto Net. d $3,783 $33,851
39. Acurex a 39. Linear Technology 39. Twitter d $3,716 $44,436
40. Finnigan 40. Symantec b 40. Airbnb d $3,378 $94,765


NOTES: This table was compiled using 1982 and 2002 Dun & Bradstreet (D&B) Business Rankings data and Blommberg data for 2020. Companies are ranked by sales.
a – No longer existed by 2002.
b – Did not exist before 1982.
c – No longer existed by 2021.
d – Did not exist before 2000.
e – Did not exist before 1995.

Same table in jpg format

Female founders – an analysis from 800 (former) startups

Leave a reply

I just decided to add a new analysis to my recent study of 800 (former) startups. Although the topic is an important one in high-tech entrepreneurship, I had never looked at it except anectoticaly in the posts with the tag #women-and-high-tech.


Eight female founders or entrepreneurs. I am not sure how many I would have automatically recognized. And you?

And here are the results I found. My apologies in advance as this work is far from perfect: I tried to identify female founders from their name and this is not always easy. I believe however I cannot be too far from the exact results.

So what does this say?

– There are 76 female founders in 825 companies, which says 9% of these former startups had a female founder. Now to make it worse, the total number of founders identified is 1644.
– It is in the biotech field, that they are most represented (hence Boston, Switzerland, California outside Silicon Valley)
– The good news is that the number is up to 15% for the last decade. Still…
– Now there are only 31 female CEOs, this is only 4% (remember that founding CEOs are a little more than 60% so this is even worse as some of these female CEOS are not even founders – see here if you don’t know what I am talking about). In fact, 20 of these women were founders and 11 were not…