Paul Graham - Essays

Some people could probably start a company at 18 if they wanted to. Bill Gates was 19 when he and Paul Allen started Microsoft. (Paul Allen was 22, though, and that probably made a difference.) So if you're thinking, I don't care what he says, I'm going to start a company now, you may be the sort of person who could get away with it.

The other cutoff, 38, has a lot more play in it. One reason I put it there is that I don't think many people have the physical stamina much past that age. I used to work till 2:00 or 3:00 AM every night, seven days a week. I don't know if I could do that now.

Also, startups are a big risk financially. If you try something that blows up and leaves you broke at 26, big deal; a lot of 26 year olds are broke. By 38 you can't take so many risks-- especially if you have kids.

My final test may be the most restrictive. Do you actually want to start a startup? What it amounts to, economically, is compressing your working life into the smallest possible space. Instead of working at an ordinary rate for 40 years, you work like hell for four. And maybe end up with nothing-- though in that case it probably won't take four years.

During this time you'll do little but work, because when you're not working, your competitors will be. My only leisure activities were running, which I needed to do to keep working anyway, and about fifteen minutes of reading a night. I had a girlfriend for a total of two months during that three year period. Every couple weeks I would take a few hours off to visit a used bookshop or go to a friend's house for dinner. I went to visit my family twice. Otherwise I just worked.

Working was often fun, because the people I worked with were some of my best friends. Sometimes it was even technically interesting. But only about 10% of the time. The best I can say for the other 90% is that some of it is funnier in hindsight than it seemed then. Like the time the power went off in Cambridge for about six hours, and we made the mistake of trying to start a gasoline powered generator inside our offices. I won't try that again.

I don't think the amount of bullshit you have to deal with in a startup is more than you'd endure in an ordinary working life. It's probably less, in fact; it just seems like a lot because it's compressed into a short period. So mainly what a startup buys you is time. That's the way to think about it if you're trying to decide whether to start one. If you're the sort of person who would like to solve the money problem once and for all instead of working for a salary for 40 years, then a startup makes sense.

For a lot of people the conflict is between startups and graduate school. Grad students are just the age, and just the sort of people, to start software startups. You may worry that if you do you'll blow your chances of an academic career. But it's possible to be part of a startup and stay in grad school, especially at first. Two of our three original hackers were in grad school the whole time, and both got their degrees. There are few sources of energy so powerful as a procrastinating grad student.

If you do have to leave grad school, in the worst case it won't be for too long. If a startup fails, it will probably fail quickly enough that you can return to academic life. And if it succeeds, you may find you no longer have such a burning desire to be an assistant professor.

If you want to do it, do it. Starting a startup is not the great mystery it seems from outside. It's not something you have to know about "business" to do. Build something users love, and spend less than you make. How hard is that?

[1] Google's revenues are about two billion a year, but half comes from ads on other sites.

[2] One advantage startups have over established companies is that there are no discrimination laws about starting businesses. For example, I would be reluctant to start a startup with a woman who had small children, or was likely to have them soon. But you're not allowed to ask prospective employees if they plan to have kids soon. Believe it or not, under current US law, you're not even allowed to discriminate on the basis of intelligence. Whereas when you're starting a company, you can discriminate on any basis you want about who you start it with.

[3] Learning to hack is a lot cheaper than business school, because you can do it mostly on your own. For the price of a Linux box, a copy of K&R, and a few hours of advice from your neighbor's fifteen year old son, you'll be well on your way.

[4] Corollary: Avoid starting a startup to sell things to the biggest company of all, the government. Yes, there are lots of opportunities to sell them technology. But let someone else start those startups.

[5] A friend who started a company in Germany told me they do care about the paperwork there, and that there's more of it. Which helps explain why there are not more startups in Germany.

[6] At the seed stage our valuation was in principle $100,000, because Julian got 10% of the company. But this is a very misleading number, because the money was the least important of the things Julian gave us.

[7] The same goes for companies that seem to want to acquire you. There will be a few that are only pretending to in order to pick your brains. But you can never tell for sure which these are, so the best approach is to seem entirely open, but to fail to mention a few critical technical secrets.

[8] I was as bad an employee as this place was a company. I apologize to anyone who had to work with me there.

[9] You could probably write a book about how to succeed in business by doing everything in exactly the opposite way from the DMV.

Thanksto Trevor Blackwell, Sarah Harlin, Jessica Livingston, and Robert Morris for reading drafts of this essay, and to Steve Melendez and Gregory Price for inviting me to speak.

A couple months ago I got an email from a recruiter asking if I was interested in being a "technologist in residence" at a new venture capital fund. I think the idea was to play Karl Rove to the VCs' George Bush.

I considered it for about four seconds. Work for a VC fund? Ick.

One of my most vivid memories from our startup is going to visit Greylock, the famous Boston VCs. They were the most arrogant people I've met in my life. And I've met a lot of arrogant people. [1]

I'm not alone in feeling this way, of course. Even a VC friend of mine dislikes VCs. "Assholes," he says.

But lately I've been learning more about how the VC world works, and a few days ago it hit me that there's a reason VCs are the way they are. It's not so much that the business attracts jerks, or even that the power they wield corrupts them. The real problem is the way they're paid.

The problem with VC funds is that they're funds . Like the managers of mutual funds or hedge funds, VCs get paid a percentage of the money they manage: about 2% a year in management fees, plus a percentage of the gains. So they want the fund to be huge-- hundreds of millions of dollars, if possible. But that means each partner ends up being responsible for investing a lot of money. And since one person can only manage so many deals, each deal has to be for multiple millions of dollars.

This turns out to explain nearly all the characteristics of VCs that founders hate.

It explains why VCs take so agonizingly long to make up their minds, and why their due diligence feels like a body cavity search. [2] With so much at stake, they have to be paranoid.

It explains why they steal your ideas. Every founder knows that VCs will tell your secrets to your competitors if they end up investing in them. It's not unheard of for VCs to meet you when they have no intention of funding you, just to pick your brain for a competitor. This prospect makes naive founders clumsily secretive. Experienced founders treat it as a cost of doing business. Either way it sucks. But again, the only reason VCs are so sneaky is the giant deals they do. With so much at stake, they have to be devious.