Pick startup partners like friends, not resumes
Graham says the most successful startups tend to hire college friends.
YC Root Access · Lecture 3 - Before the Startup (Paul Graham)Paul Graham, the investor and startup adviser speaking to Stanford students, has watched the same hiring regret play out again and again. A founder meets someone who looks impressive on paper, feels uneasy, ignores the feeling, and later says, “I knew there was something wrong about that guy.”
His fix is simple and deliberately personal. Pick co-founders the way you would pick friends. Graham put it this way: “Work with people you genuinely like and respect and that you have known long enough to be sure.”
That last part matters, Graham says, because a lot of people are good at seeming likable for a while. Wait until interests are opposed, he says, and then you will see who someone really is.
The trap is sharpest for engineers, he adds. They often assume business is supposed to be slightly distasteful, so when they meet someone smart but distasteful they think, well, okay “this must be normal for business,” in Graham’s words. It is not normal, and it is not required.
Apply the same standard to the first hires. The first two or three people should be almost like founders, motivated by the same things as you. They should be peers you barely have to manage, not employees who need supervision. An example raised in the lecture was Airbnb waiting five months to make its first hire, a sign that ideally you succeed before you even need a small team.
Asked whether hiring people you like creates a monoculture with blind spots, Graham answered that in a startup “many things will be going wrong” and you cannot expect it to be perfect. As Graham said, “the advantages of hiring people that you know and like are far greater than, you know, the small disadvantages of having some monoculture.” In an offhand remark to close, Graham said all the most successful startups hire pals from college.
Questions people ask
Doesn't hiring friends lead to groupthink?
Graham himself plays down that worry, saying the trust advantage outweighs the monoculture risk. Beyond Graham, the general advice from managers is to keep that trust but add different skills and user experience, and to invite disagreement about the work.
What should I do if I feel uneasy about an impressive candidate?
Pause and check the feeling before credentials impress you into saying yes. Talk to people who have worked with them under stress, and ask what happened when money, credit, or direction was disputed. If the unease persists, pass.
When should I make those first hires?
Later than you think, after you have made something people want and the work is overflowing. Early hires add communication costs, so each one should unlock progress you cannot make alone. If you have to manage them closely, they are the wrong early hire.