Paul Graham

Startups

Here you'll find pertinent information regarding frequently asked questions relating to startups. I wish you luck on your next venture!

Will you invest in and/or advise our startup?

I invest in and advise startups through Y Combinator. If you want investment, the way to get it is to apply for our next round of funding. If we fund you, you can have all the advice you want (if not more).

I'm sorry, but I can't give advice to every startup that writes to me. It's as much as I can do to keep up with the ones we've funded.

Back to top

Where do you find co-founders?

Most successful startups have more than one founder, and usually the founders seem to have been friends for at least a year before starting the company. The best way to meet co-founders is to go to school with them, so recent grads have a big advantage there. You can also meet co-founders at work, but be careful not to violate whatever noncompete you signed. In the old days, co-founders often met through user groups, but this seems less common now.

Back to top

I want to start a startup, but I don't know how to program. How long will it take to learn?

I would guess a smart person can learn to hack sufficiently well in 6 months to a year. The best way to do it would be to find some startup to hire you in an initially menial capacity, and start learning to program on the side. Then gradually work your way up from answering phones through system administration to actual software design. There is always so much to do in a startup that people won't be too picky about your paper qualifications, if you can solve problems for them without screwing up.

Back to top

How can someone start a startup if they have a family to support?

The best plan might be to start a consulting business that you can gradually morph into a startup. This way you always have a source of income.

Back to top

Do you know of any good books about startups?

The how-to type of books are generally wretched. Many are downright mistaken.

The best source of information about startups is probably not business books, but histories of particular startups and industries. The most famous is Tracy Kidder's Soul of a New Machine, but there are many good books of this type. I particularly liked Sorensen's My Forty Years with Ford.

The one book we encourage startup founders to read is Dale Carnegie's How to Win Friends and Influence People. It's critically important for anyone in business. Try to get a used copy printed before the 1960s; after Carnegie died, the book continued to be "updated" by a committee, and the changes were not for the better. I'd also recommend Franklin's Autobiography.

Back to top

I have multiple startup ideas. How do I decide which to work on?

Work on the one that will cause the most immediate, concrete improvement in users' lives. Don't worry too much at first about competitors, or how users will find out about it, or how to make money. But don't work on something that's going to take ten years, either. (In technology, ten years rounds up to never.)

Google is a good example. Everyone needs Web search, and the founders probably had something that significantly improved their own ability to find stuff online within the first couple months. And once something starts to work (a) it's enormously encouraging, and (b) it's much clearer what direction to take it in.

Back to top

How much should a startup worry about being sued for patent infringement?

There are two kinds of patent suits:

1. Random people suing you in the hope of getting money.

2. Competitors suing you in order to destroy you.

Until you're successful, you won't have to worry about the first type. And once you are successful, you'll inevitably have a lot of them to deal with, valid or not. So fatalism is the right plan for type 1 suits.

How much you should worry about type 2 suits depends on what kind of business you're in. You probably have to worry more about it in hardware than in software. In software such suits are rare, so far.

So if you're doing software, I'd try to build something people want, and then hope for the best. The best, in this case, being that you get big enough that you have to worry about type 1 suits.

In software, patent suits are rarely to never the deciding factor in whether a startup succeeds or not. Founders should probably spend 100x as much time worrying about building something users won't want, becausethat kills 100x more startups than patent problems.

Back to top