Stop Trying to Think Up Startup Ideas
The best companies began as side projects, so build skill, curiosity, and a circle of collaborators first.
YC Root Access · Lecture 3 - Before the Startup (Paul Graham)Paul Graham, the founder of the startup backer Y Combinator, has a warning for Stanford students who want to found a company one day: do not sit down and "try to think of startup ideas."
Deliberate brainstorming, he told them, produces "ideas that are not only bad, but bad and plausible sounding." You fool yourself and everyone else, and you "waste a lot of time before realizing they're no good."
The alternative sounds passive, but it is a method. Graham said to "turn your brain into the type that has startup ideas unconsciously", so unconsciously that you do not at first realize they are company ideas. Yahoo, Google, Facebook, and Apple all began that way, he said. None was supposed to be a company. Each was a side project. "The very best ideas almost have to start as side projects" because they are "such outliers that your conscious mind would reject" them as businesses.
Graham put the prescription in three parts. First, "learn a lot about things that matter". Second, "work on problems that interest you". Third, do it "with people you like and respect".
That last part is also how you find co-founders. And the first part is broader than it sounds. Graham said he first wrote "become good at some technology" in his essay, then decided that was too narrow. Brian Chesky and Joe Gebbia, the founders of Airbnb, went to art school. Their edge was not code. It was design, and an ability to organize people and get projects done.
How do you know you are working on something real? Graham said his test is shamelessly personal. Real problems are interesting, and he follows his curiosity "even if no one else cares about them". When the team behind Justin.tv decided to focus on people watching video games, a business that became Twitch, his first reaction was confusion. It worked anyway.
Graham said he seems to have "some sort of internal compass" that helps him, but cautioned that this is true for him and he does not know what is in listeners heads. So the practical move now, in college, is to follow that curiosity hard. Pick a hard, real, current problem you actually enjoy, and build something with friends you admire. The startup comes later, as the side effect.
Questions people ask
If I shouldn't brainstorm ideas, how do I pick what to work on?
Graham's advice is to learn about things that matter and work on problems you genuinely find interesting, even if no one else cares about them. His example is Twitch: when Justin.tv focused on people playing video games, he was confused, but it turned out to be a good business.
Does this only work for programmers?
No. Graham says becoming good at technology is sufficient but too narrow. He points to the Airbnb founders, who went to art school and brought expertise in design and in organizing people and getting projects done. What matters is work that stretches you.
How do I find co-founders this way?
Graham says working with people you like and respect is how you get an idea and co-founders at the same time. As an editor's suggestion for trying this, not Graham's words, you could work on demanding projects with classmates you admire to learn how you work together.