Paul Graham, Dropbox and The Single Founder Exception

"Without Paul Allen Microsoft as we know it would not have existed because he named Microsoft Microsoft."

So he named the company? Now you are being totally silly and showing that you are determined to fight with me for just silly reasons.

You are the one being 'personal'. You just don't want the argument made that sole founders can be a good bet and just want to stay with the PG and VC herd that cofounders are crucial, say, for naming the company and cut me down for not joining your herd.

Jobs was a successful sole founder at Next, Pixar, and his return to Apple. The only role from Woz and Jobs's first time at Apple was just cash, not Woz or a 'cofounder'. You are straining to deny that Jobs was successful in his last three gigs as a sole founder.

Gates? He had to refound the company after the success of MS/DOS and make Windows, Windows Server, SQL Server, and Office all real. He did. Alone. While doing this, IBM was laughing at him. He knocked the socks off IBM, DB2, Lotus, WordPerfect, Sun, HP, and more.

Just look around you: Commonly cofounders and coleaders just suck. Big committees suck. Group decision making sucks. Take IBM, AT&T, GM, Sun, HP: In each case, the management suite and top management 'team' was awash in what was regarded as the best qualifications. Still, they all just sucked. The Navy knows better: On the bridge of a ship, there exactly one captain. Although one captain can be bad, two is usually worse.

Then you are ignoring the present for what is central at HN: IT startups, especially Web 2.0 startups. Elsewhere on this thread I've explained in good detail: A founder needs to understand his business; he needs to understand the software; the bottleneck in understanding the software is not the unique software of the company but just understanding the now huge software components available; that bottleneck can be passed by just one person at a time, essentially alone, or learning is not a spectator sport or a team sport; once a founder has passed this bottleneck, a cofounder becomes much less relevant than in the past. You can see this, if you want to.

The main issue here is sole founders or cofounders. The resolution is, net, having all the business between one pair of ears is a great advantage when it can be done, and now for a Web 2.0 startup it not only can be done but should be done. The 'team' of Michelangelo and anyone else would be nowhere nearly as good at painting the ceiling as just Michelangelo alone.

"By definition getting to the same results without co-founders requires a much higher level of ability than if being part of a 'collective'"

I think in order for this to be true, you'd have to assume that the team as a whole is greater than the sum of its parts. This is not always the case with cofounders.

In general, though, I agree with you. Results speak for themselves; build a business for customers, not investors.

Are there any good examples where a single founder made it big, truly on his own? You can imagine that maybe Drew could have done it without Arash, or Gates could have done it without Allen, etc. But they didn't, which validates the exception. So before we casually say /oh if you don't feel like you need a co-founder then don't bother/, it would be good to know if any single founders have been successful.

And maybe we should define success as financial/market success, not emotional fulfilment or getting your first two round of funding.

Once again:

Drew came to PG with a lot of good work done. As is common in VC, at least for funding from YC, PG said that Drew needed a 'cofounder'.

PG in his

http://paulgraham.com/startupmistakes.html

gives 18 reasons a startup might fail with the first reason "Single Founder". Moreover, a 'single founder' is frowned on elsewhere in VC.

So, we have claims that projects with single founders are somehow inferior, to be frowned upon, to be avoided, not to be funded, etc. On these claims I call BS.

Instead, given a project, typically much more information is available for evaluation so that with that extra information 'single founder' should commonly be from a strong advantage to nearly neutral to a strong disadvantage. Net, by itself, 'single founder' says next to nothing important. So, I claim that automatically saying there is something seriously wrong with a single founder project is BS.

There are some dangers in having a cofounder: Likely the most important danger is that cofounders can get into fights. Another danger, broadly, is the one the US Navy recognizes: On a ship, there is exactly one Captain. So, in particular, the ship doesn't continue without good direction while 'co-captains' argue about the direction.

In

http://news.ycombinator.com/item?id=3129873

0x12 says

"The upside (and it's a huge upside)

- you don't have to deal with your co-founders"

I agree.

In

http://news.ycombinator.com/item?id=3128876

staunch asked

"What if Arash had turned out to be a bad match? They didn't know each other, so it was a pretty big gamble."

I agree.

PG answered stauch with in part:

"Oddly enough Drew and I talked about this question tonight on a panel at the Computer History Museum. It was a big risk to recruit someone as a cofounder that he didn't know before (and I've seen a lot of cases where it didn't work out) but in this case the risk was mitigated by how much they had in common."

So here PG recognizes "a big risk". So, it is widely recognized that having a cofounder is risky.

For the "mitigated", I can't take that very seriously. Instead, For Drew to take on a cofounder so quickly was still risky.

Let's examine this issue of risk in more detail: Why accept the risk?

One claimed reason for accepting the risk is that a single founder may lack 'psychological stability and self-discipline' and, thus, go wacko, make bad decisions, get discouraged, give up on the company, go for fast women, slow horses, cheap booze, and suffer "the thousand natural shocks that flesh is heir to". He might. Yet he might not.

So, in short, such reasons are claiming that a person working nearly alone essentially under just their own direction and motivation with their own ideas is likely doomed.

Let's examine this claim; is a person working alone essentially doomed from various 'psychological' issues?

First, let's consider the case of nearly anyone who, during the years they are working, is by a wide margin the best in the world at what they are doing. So, by being so good in their work, they really can't have a 'partner' or 'coworker' to share the crucial parts of the work. That is, being the best in the world is not a pairs competition, partnership, joint work, group activity, or team sport. Or, as is often said, "It's lonely at the top.". So people who are the best in the world at what they do are essentially necessarily working alone but are definitely not "doomed".

Second, for some examples, consider some world famous people in the arts, Michelangelo, Bach, Mozart, Beethoven, Tchaikovsky. Also in musical performance consider Heifetz, Rostropovich, Horowitz. And in the sciences consider Newton, Maxwell, Einstein. In each case, they worked essentially alone. E.g., consider Bach and compare his music with that of any of his children; the children were hardly up to copying Dad's manuscripts. These people were working essentially alone but definitely were not "doomed".

Are there any successful startups by a single founder? Sure: FedEx. While I don't know the fine details of the early days at Oracle, a guess is that Ellison was a sole founder there. Viterbi may have been a sole founder at QUALCOMM. We certainly have to count Dell. HotMail? Lotus? So, these sole founders in business were not doomed.

But there are millions more: Just pick a Main Street anywhere in the US, border to border. There will find millions of successful entrepreneurs running 'sole proprietorships' and doing all the important work themselves. Will find such people running restaurants, big truck-little truck operations, gas stations, auto body shops, auto repair shops, doing kitchen renovation, dentistry, tax accounting, family law, various trades, and more. Then outside of town, look at farmers growing fruits, vegetables, grains, raising poultry, dairy cattle, beef cattle, sheep, etc. These people usually are not in high margin, high growth, high potential businesses, but they are often 'sole proprietors' and are certainly not doomed.

For more, in business, is a coworker important or can one guy hope really to do the work essentially alone? I suggest considering Steve Jobs, Bill Gates, and Warren Buffet. Did they do IT startups always alone? No. But in the world class work they did, and/or are doing, they did it alone or nearly so. So, working without a coworker is not doomed.

For the more narrow activity of IT startups, especially for Web 2.0, can a sole founder really do all the crucial work himself or is he doomed? I claim that now he can expect to do all the crucial work and, indeed, has one heck of an advantage over a 'team'.

For the advantage, he won't get into fights with a cofounder, won't have his best work diluted by compromising with a cofounder, and won't have to spend the time convincing, communicating with, or coordinating with a cofounder.

For being able to get the work done, there have been some revolutions in the foundation 'resources' of IT from Moore's law, the similar law for disk space, optical fibers for digital communications, and infrastructure software, say, based on Linux or Windows. So, the founder can be the chief programmer on his 'chief programmer team' and also the whole team except for, say, some occasional very focused technical support from Internet fora, Microsoft, Cisco, HP, etc. The CEO of a startup really should know all the work of his company like the back of his hand; basically the way for him to do this is just to do the work himself.

When should a VC set aside a project from a single founder?

(1) The founder should have some good work to show and, otherwise, a VC should 'set aside'. (2) The founder should appear to be able to get the rest of the work done through a company big enough really to need to hire. So, the founder should be able to (a) give a good pitch to the VCs, (b) show the good work to date, (c) have solid plans for the next steps, (d) clearly be able to work with his lawyer, bookkeeper, tax accountant, recruiting firm, office landlord, and paying customers, and (e) manage at least a few people before it is time to get him, say, a COO. (3) The founder should not have any obvious disqualifications.

If these criteria (1)-(3) are met, then a sole founder stands to be a better bet than a 'team'.

Drew came to PG with a lot of good work done. As is common in VC, at least for funding from YC, PG said that Drew needed a 'cofounder'.

PG in his

http://paulgraham.com/startupmistakes.html

gives 18 reasons a startup might fail with the first reason "Single Founder". Moreover, a 'single founder' is frowned on elsewhere in VC.

So, we have claims that projects with single founders are somehow inferior, to be frowned upon, to be avoided, not to be funded, etc. On these claims I call BS.

Instead, given a project, typically much more information is available for evaluation so that with that extra information 'single founder' should commonly be from a strong advantage to nearly neutral to a strong disadvantage. Net, by itself, 'single founder' says next to nothing important. So, I claim that automatically saying there is something seriously wrong with a single founder project is BS.

There are some dangers in having a cofounder: Likely the most important danger is that cofounders can get into fights. Another danger, broadly, is the one the US Navy recognizes: On a ship, there is exactly one Captain. So, in particular, the ship doesn't continue without good direction while 'co-captains' argue about the direction.

In

http://news.ycombinator.com/item?id=3129873

0x12 says

"The upside (and it's a huge upside)

  - you don't have to deal with your co-founders"

I agree.

In

http://news.ycombinator.com/item?id=3128876

staunch asked

"What if Arash had turned out to be a bad match? They didn't know each other, so it was a pretty big gamble."

I agree.

PG answered stauch with in part:

"Oddly enough Drew and I talked about this question tonight on a panel at the Computer History Museum. It was a big risk to recruit someone as a cofounder that he didn't know before (and I've seen a lot of cases where it didn't work out) but in this case the risk was mitigated by how much they had in common."

So here PG recognizes "a big risk". So, it is widely recognized that having a cofounder is risky.

For the "mitigated", I can't take that very seriously. Instead, For Drew to take on a cofounder so quickly was still risky.

Let's examine this issue of risk in more detail: Why accept the risk?

One claimed reason for accepting the risk is that a single founder may lack 'psychological stability and self-discipline' and, thus, go wacko, make bad decisions, get discouraged, give up on the company, go for fast women, slow horses, cheap booze, and suffer "the thousand natural shocks that flesh is heir to". He might. Yet he might not.

So, in short, such reasons are claiming that a person working nearly alone essentially under just their own direction and motivation with their own ideas is likely doomed.

Let's examine this claim; is a person working alone essentially doomed from various 'psychological' issues?

First, let's consider the case of nearly anyone who, during the years they are working, is by a wide margin the best in the world at what they are doing. So, by being so good in their work, they really can't have a 'partner' or 'coworker' to share the crucial parts of the work. That is, being the best in the world is not a pairs competition, partnership, joint work, group activity, or team sport. Or, as is often said, "It's lonely at the top.". So people who are the best in the world at what they do are essentially necessarily working alone but are definitely not "doomed".

Second, for some examples, consider some world famous people in the arts, Michelangelo, Bach, Mozart, Beethoven, Tchaikovsky. Also in musical performance consider Heifetz, Rostropovich, Horowitz. And in the sciences consider Newton, Maxwell, Einstein. In each case, they worked essentially alone. E.g., consider Bach and compare his music with that of any of his children; the children were hardly up to copying Dad's manuscripts. These people were working essentially alone but definitely were not "doomed".

Are there any successful startups by a single founder? Sure: FedEx. While I don't know the fine details of the early days at Oracle, a guess is that Ellison was a sole founder there. Viterbi may have been a sole founder at QUALCOMM. We certainly have to count Dell. HotMail? Lotus? So, these sole founders in business were not doomed.

But there are millions more: Just pick a Main Street anywhere in the US, border to border. There will find millions of successful entrepreneurs running 'sole proprietorships' and doing all the important work themselves. Will find such people running restaurants, big truck-little truck operations, gas stations, auto body shops, auto repair shops, doing kitchen renovation, dentistry, tax accounting, family law, various trades, and more. Then outside of town, look at farmers growing fruits, vegetables, grains, raising poultry, dairy cattle, beef cattle, sheep, etc. These people usually are not in high margin, high growth, high potential businesses, but they are often 'sole proprietors' and are certainly not doomed.

For more, in business, is a coworker important or can one guy hope really to do the work essentially alone? I suggest considering Steve Jobs, Bill Gates, and Warren Buffet. Did they do IT startups always alone? No. But in the world class work they did, and/or are doing, they did it alone or nearly so. So, working without a coworker is not doomed.

For the more narrow activity of IT startups, especially for Web 2.0, can a sole founder really do all the crucial work himself or is he doomed? I claim that now he can expect to do all the crucial work and, indeed, has one heck of an advantage over a 'team'.

For the advantage, he won't get into fights with a cofounder, won't have his best work diluted by compromising with a cofounder, and won't have to spend the time convincing, communicating with, or coordinating with a cofounder.

For being able to get the work done, there have been some revolutions in the foundation 'resources' of IT from Moore's law, the similar law for disk space, optical fibers for digital communications, and infrastructure software, say, based on Linux or Windows. So, the founder can be the chief programmer on his 'chief programmer team' and also the whole team except for, say, some occasional very focused technical support from Internet fora, Microsoft, Cisco, HP, etc. The CEO of a startup really should know all the work of his company like the back of his hand; basically the way for him to do this is just to do the work himself.

When should a VC set aside a project from a single founder?

(1) The founder should have some good work to show and, otherwise, a VC should 'set aside'. (2) The founder should appear to be able to get the rest of the work done through a company big enough really to need to hire. So, the founder should be able to (a) give a good pitch to the VCs, (b) show the good work to date, (c) have solid plans for the next steps, (d) clearly be able to work with his lawyer, bookkeeper, tax accountant, recruiting firm, office landlord, and paying customers, and (e) manage at least a few people before it is time to get him, say, a COO. (3) The founder should not have any obvious disqualifications.

If these criteria (1)-(3) are met, then a sole founder stands to be a better bet than a 'team'.