Number of founders - statistics
+--------------------+-----------+
| number of founders | frequency |
+--------------------+-----------+
| 1 | 40 |
| 2 | 31 |
| 3 | 17 |
| 4 | 6 |
| 5 | 2 |
| 6 | 1 |
| 7 | 1 |
| 8 | 1 |
| 13 | 1 |
+--------------------+-----------+
The number of sole founder companies is probably slightly exaggerated because sometimes one person is so much more prominent that the others are never mentioned. I did a little double checking though. Another important thing is that some companies were omitted because I knew only that they had several founders but I wasn't able to find out exactly how many. I suspect there are more companies with 4, 5, 6 or more founders.
Unfortunately without numbers about failed companies we still don't know the success ratio of each frequency group, which is what's most interesting for startup founders here. It is well possible that sole founders fail 10 times as often as 2 founder companies, or vice versa. For anyone interested in double checking (yes I'm sure there are errors ;-) here's the full list:
+-----------------------------+--------------------+
| company | number of founders |
+-----------------------------+--------------------+
| ACS | 1 |
| Activision | 5 |
| Adobe | 2 |
| Akamai | 4 |
| Altera | 3 |
| Amazon | 1 |
| AMD | 8 |
| Amphenol | 1 |
| Analog Devices | 2 |
| Ansys | 1 |
| Apple | 3 |
| Applied Biosystems | 2 |
| Aspen Technology | 1 |
| Autodesk | 13 |
| Avnet | 1 |
| Baidu | 2 |
| BEA | 3 |
| Beckman Coulter | 1 |
| BMC | 3 |
| Broadcom | 2 |
| Brocade | 4 |
| CA | 1 |
| Canon | 2 |
| Cerner | 3 |
| Check Point Software | 3 |
| CheckFree | 1 |
| Cisco | 2 |
| Citrix | 1 |
| Cognos | 2 |
| Compuware | 3 |
| CSC | 3 |
| Cypress Semiconductor | 1 |
| Cytyc | 1 |
| Dell | 1 |
| Dolby Laboratories | 1 |
| Eaton Corp. | 3 |
| Ebay | 1 |
| Electronic Arts | 1 |
| EMC | 2 |
| Fiserv | 2 |
| Flextronics | 1 |
| Garmin | 4 |
| Google | 2 |
| Harris Corp. | 2 |
| HP | 2 |
| IBM | 1 |
| IMS Health | 2 |
| Informatica | 2 |
| Infosys | 7 |
| Intel | 2 |
| Intuit | 2 |
| Jabil Circuit | 2 |
| Juniper Networks | 3 |
| Konami | 1 |
| Kyocera | 1 |
| Lam Research | 1 |
| Logitech | 3 |
| LSI Corporation | 3 |
| Man Tech | 1 |
| McAfee | 1 |
| Micron Technology | 2 |
| Microsoft | 2 |
| Mindray | 1 |
| Motorola | 2 |
| NetApp | 3 |
| Nokia | 1 |
| Novell | 4 |
| Nvidia | 3 |
| Open Text | 1 |
| Oracle | 3 |
| Parametric Technology Corp. | 1 |
| Philips | 1 |
| Qualcomm | 2 |
| Quest Software | 2 |
| RIM | 2 |
| Roper Industries | 1 |
| SAIC | 1 |
| Salesforce | 1 |
| SanDisk | 2 |
| SAP | 5 |
| Satyam | 1 |
| Seagate | 2 |
| Siemens | 1 |
| SRA International | 1 |
| Sun Microsystems | 4 |
| SunPower | 1 |
| Sybase | 2 |
| Symantec | 1 |
| Synaptics | 2 |
| Tellabs | 6 |
| Thomson | 1 |
| TI | 4 |
| Tibco | 1 |
| Trimble Navigation | 3 |
| VMware | 2 |
| Waters Corp. | 1 |
| Wipro | 1 |
| Xerox | 2 |
| Xilinx | 3 |
| Yahoo | 2 |
+-----------------------------+--------------------+
I wonder out of these founders how long each worked fulltime for the company. Also, it would be great to know how many key early employees each had who may have not gotten the "co-founder" title.
The poster bases his logic on statistics, saying that many companies were started by a single person--which may indeed be the most popular combination, as it seems that whenever a BusinessWeek slideshow is posted here listing the top 10 hottest startups, half are about companies with single founders. This makes sense--the founder created a project to solve a problem, not to make money; then, it took off--in contrast to gathering three people and "forcing" them to create something that can be sold to a VC three months to two years later.
Now, pg is probably right that most of the people he has ever met had a co-founder (he should know). The problem is, even according to the list above, FIVE of every SIX people on that list will tell you they have a co-founder. 5.00 out of every 6.11 to be exact. That does not mean those companies are more successful. I think that reasoning is flawed. Not only is this misleading numerically, but on top of that, it's probably misleading socially since the single founder is always programming while teams with multiple founders are much more likely to make outside contact with investors like pg, being that there is more of them.
Also, if two or three people believe in the idea versus one, that does not mean the first team has a better chance to make a better product.
However, if you are going to move to the Silicon Valley as a single founder, and expect to get a co-founder or employee later, I'm guessing it will be nearly impossible to get anyone on short notice, if at all. Therefore, if you're going to start with money, you may as well bring a committed team with you. I wouldn't be surprised if it's probably easier to get money in Silicon Valley than to find people to join you once you've already moved there. This is why, although YCombinator is probably better in Silicon Valley than any other place, I wouldn't be surprised to learn that most applicants are from outside the region. Any stats on this, pg?
Thus, where Apple neede hardware and software and marketing guys (really three founders, though Woz should count double)... plentyoffish.com only needs one guy. And when Apple was at the revenue level pof is at now, I think they probably had a thousand employees, probably at least hundreds.
so, the requirement for multiple founders (if it ever existed) is now less likely.
I know I'm not answering your question because its an academic question for me-- I've chosen my path already.
I think really, at the end of the day, number of founders has no factor on the success of startups, but founder fights are a significant factor in startup failure.
So, whether you are solo, or you have 5 other founders, you need harmony between you and your other founders and/or you and your other employees, to have success.
You had to work all the time at your start-up back then and you still have to now. The gains in complexity have kept pace with the gains in efficiency in a sort of self-feeding co-evolution.
Ian
However, I don't know if it's terribly useful to include ancient companies like IBM, Beckman or Avnet. Although it does help show that the trendline is going up -- Salesforce, one of the most recent, had four co-founders in 1999.
Here's a USA Today article on Salesforce. Under the second subheading they say,
"Salesforce was born in March 1999 in cramped, adjoining apartments atop San Francisco's picturesque Telegraph Hill. Benioff, still employed at Oracle, lived in a condo. Fellow co-founders Parker Harris, Dave Moellenhoff and Frank Dominguez set up shop next door. Computer cables ran between the apartments through windows. 'It was the Gold Rush era for tech start-ups then,' Harris recalls."
And if you Google on Parker Harris at least, you can find him described as a co-founder.
By the way, SAIC is an interesting one. The main guy founded it with "a handful of scientists" and one page describes someone as a "Vice Co-Founder" of SAIC. "Vice co-founder!" That's harsh.
PG advocates consensus management, and that's the real sticky wicket.
That said probably it depends of how big the idea is and how many people are required to implement it, so the number of people it's really an illusion.
Relative to that, other interesting data would be to know how early before it was obvious did they go after their idea and if that gave them time to make mistakes and correct them. Minimizing time-to-market, that works probably in favor of no matter how many are the founders.
Thus, as the number/quality of founders decreases so the earlier your implementation should get or be feasible to get to the market.
A few I know the startup story: OpenText | 3 | 1.7 billion | F Janna Systems | 1 | 1.1 billion | T
Facebook - 10? 20? lol