How to Start a Startup: Advice from Paul Graham

At the end of the day, founding a startup primarily requires you to do one thing: learn.

At least that’s what Paul Graham, cofounder of Y Combinator, said in a lecture at Stanford University as part of YC President Sam Altman’s series on How to Start a Startup.

(For more tips, check out our previous coverage of the How to Start a Startup series, including a lecture by Altman and Facebook cofounder Dustin Moskovitz.)

Graham’s advice sounds simple, and maybe it is. But he’s simplified it because founding a startup is a lot more counterintuitive than it seems. You can’t just trust your intuition.

It’s like skiing. Your instinct is to lean back, but if you do, you’ll fly down the hill out of control. You have to suppress those instincts. You have to keep a list in your head of what to do and not to do (alternate feet, do not drag inside foot, etc.). Startups are as unnatural as learning how to ski, and there’s a list of counterintuitive things you have to remember in order to keep your startup from flying down the hill.

1. Work with people who seem likeable

Graham explained that startups are so weird, if you follow your instincts, they’ll lead you astray. You can, however, trust your instincts in terms of dealing with people. Startup founders often don’t do this. People may seem smart and impressive, but that shouldn’t trump your instinct on whether or not you’d be friends with that person. Graham’s advice: work with people who seem likeable.

2. Understand your users

What you need to succeed is not expertise about startups but expertise about your own users.

Graham said that Mark Zuckerberg did not succeed because he was an expert in startups — after all, Facebook was originally incorporated in Florida as an LLC. Instead he succeeded in spite of that, because he understood his users.

It may actually be dangerous to know too much about starting a startup. Graham explained that too many people think it’s as simple as: come up with a great idea, get some investors, hire your friends, and rent space in SOMA. But then they realize they may have created a product that people don’t actually want. It’s otherwise known as “playing house,” and Graham said it’s what they have been trained to do until now.

Young founders try to figure out what the new tricks are, what to do next. But the measure of success for startups is not fundraising. You need to start a startup that’s actually doing well, and then secure funding. And how do you start a startup that’s doing well? Give users something they want!

3. Don’t look for tricks

Starting a startup is where gaming the system stops working. There’s no one to trick, only users. And all they care about is if your product does what they want.

Graham did concede that faking can sometimes work with investors. But he explained that, even if you might be able to fool them for a few times, ultimately it’s not in your best interest. So stop looking for “the trick.”

4. There’s no rush

Startups are all-consuming. It never actually gets easier; only the nature of the problems change. Graham compared the experience to having kids: while it’s the best thing in the world, there are a lot of things that are easier to do before you have kids than after.

People seem to think they should start startups in college, but then, he said, you’re not a student anymore! If the goal, ultimately, is to have a good life, there’s no rush. There are things you can do in your twenties, for example, that you can’t do later if you’ve started a startup. Major success actually takes away some serendipity, which is something you can’t create on your own, and something you can’t get back.

5. You can’t know the outcome in advance

If you’re not up to the challenge, you can’t tell in advance. And it will change your life. You can tell how smart a person is, but you can’t tell how ambitious.

6. Turn your brain off

The way to get startup ideas is not to try to think of startup ideas. Instead of making a conscious effort to think of them, turn your brain off.

Graham contended that the very best ideas have to start as side projects because they’re always such outliers that your conscious mind would reject them as actual companies. So, how do you do this?

Learn a lot about things that matter.
Work on problems that interest you.
Work with people you like and respect (which will also help you find co-founders).

7. Solve real problems

What kind of things matter, Graham asked? Real problems. He explained that he only worked on Y Combinator because it seemed interesting. The best way to prepare yourself for starting a startup is to be interested in solving real problems.

The best startups tend to come from true entrepreneurship, not learned entrepreneurship. And the component of entrepreneurship that really matters, he said, is domain expertise. Larry Page, for example, was genuinely interested in search, not because he thought it would amount to anything but because he wanted to learn. And then it became—well, you know.