Is two founders also one too many? Sometimes! At least with one founder you actually solve the problem of founders not agreeing on a course of action (which surely is a bigger probably two founders than an odd number like three).
I'll concede that I'm biased, but my startup has three founders and there is absolutely no doubt in my mind that we would be worse off with fewer people.
It's more like Hacker, Business Guy ("Hustler" sounds a little pejorative), and Sales Guy. But I'd point out that I consider us a B2B media company that makes clever use of technology, more than a "technology startup." Nailing enterprise sales is crucial to our success, which is probably not the case for consumer internet startups (or for startups that are, uh, not yet trying to make money).
In our experience, 3 is as good as 2 if the people are as good. It's harder to find 3 good people than 2 though.
If you decide three is one too many, the new pitfall you have to watch out for is tiebreaking. Suppose there's a fundamental disagreement, like whether to sell out or whether to fire one of the two founders and find someone new. You think you'll always be able to talk things through and come to a consensus, but let's say you're wrong. A naive 2-person setup simply has no way to handle a situation like that -- it's paralyzing. So you need a way to break ties. And now instead of a fair three-way split, you have one founder with more power (and investment) than the other, or you have some outside party (played by Justin Timberlake, presumably) with control over your future if they like the other guy more than you.
So maybe two founders is one too many. Going with one founder shouldn't cause any problems ... right?
Here is a test for the next time you are founding something. If you weren't a founder, you were just hired labor, and someone tried to give you 93% of this company and total autonomy in exchange for running it full-time, but NO other support of any kind and you would be completely on your own in every aspect of running this company - would you accept this?
For most ideas, the answer is, "Hell no. I have a lot better things to be doing with your time than running your worthless company without any support from you, full time in exchange for 93% of it."
People somehow fail to apply the same standard to their own ideas. They accept this bad deal just because the person who "gives" them 93% (or 100%) of the company is one and the same person, themselves.
Let me illustrate with an example. Think of an idea for a web site that you have.
If someoe approached you with that idea and asked you to code it, but they would not help in any way, would you do so in exchange for nearly 100% of it? Probably not.
But when you make YOURSELF the same proposition, you often end up saying, "Yes. I will code this since the result is mine."
You forget that the result is not worth the work and you would not accept it under more neutral terms.
This is simply illogical. Stop thinking like a founder, and start thinking like a managing director.
I think one of the biggest mistakes that startups make is not getting on the same page from day one, and letting small differences become large ones. This applies equally to any sized team, but may get exacerbated as your team increases in size, and there may be a huge jump in this problem as you get to size 3.
Then again, every team size comes with its own set of issues. I wouldn't say 1 is too few, 3 is too many, and 2 is just right. I'd take this article more as, 'the perils of a 3-person startup.'
Any headline which ends in a question mark can be
answered by the word 'no'
AirBnB & Twitter each had three founders. VMWare had five founders[2], and they've done alright.[1] http://en.wikipedia.org/wiki/Betteridges_Law_of_Headlines
[2] http://www.quora.com/Startups/Can-a-startup-be-successful-wi...
http://www.quora.com/Startups/What-would-the-ideal-web-techn...
The truth of the matter is there's no "one size fits all". Ask 100 founders and you'll get 100 different personal experiences and storied of how one structure worked for them, while other structures destroyed them.
The Hacker, Hustler, Designer model seems to hold strong with incubators, investors etc. as it seems, at the very least, to convey the atmosphere of being well rounded, fully capable and fast to deliver. Obviously this is not always the case.
The nature of past relationship between the founders, plays a large role in how well they can be expected to perform together and survive long term stress. There's no wonder YC dedicated a question to learning the nature of this relationship in their application form.
Some relationships make it, and others won't - the best you can do is try to truly and honestly asses how your relationships might hold up through hard times. Examine all aspects of it in as objective way as you can and make the decision on a case by case basis.
Finding one other person is hard enough -- trying to find three seems really challenging. There's ways to make decisions and break tie-breakers without having a third co-founder.
I believe that it comes down to expertise. Its very likely that you're not going to have domain expertise in everything, so you'll have to find some other co-founder that does. I'd think that you'd want to add more co-founders to ensure you have the domain expertise covered.
There's a blog post on our website that discusses "Why Three Founders in Better than Two". Might want to check that out for ideas on why it is better to have more than two.
http://thinkspace.com/why-three-founders-is-better-than-two/
Increase the team size by one and you go from one relationship to three.