pgstrata
A Student's Guide to Startups
2

October 2006

3

(This essay is derived from a talk at MIT.)

4

Till recently graduating seniors had two choices: get a job or go to grad school.

5

I think there will increasingly be a third option: to start your own startup.

6

But how common will that be?

7

I'm sure the default will always be to get a job, but starting a startup could well become as popular as grad school.

8

In the late 90s my professor friends used to complain that they couldn't get grad students, because all the undergrads were going to work for startups.

9

I wouldn't be surprised if that situation returns, but with one difference: this time they'll be starting their own instead of going to work for other people's.

10

The most ambitious students will at this point be asking: Why wait till you graduate?

11

Why not start a startup while you're in college?

12

In fact, why go to college at all?

13

Why not start a startup instead?

14

A year and a half ago I gave a talk [blocked] where I said that the average age of the founders of Yahoo, Google, and Microsoft was 24, and that if grad students could start startups, why not undergrads?

15

I'm glad I phrased that as a question, because now I can pretend it wasn't merely a rhetorical one.

16

At the time I couldn't imagine why there should be any lower limit for the age of startup founders.

17

Graduation is a bureaucratic change, not a biological one.

18

And certainly there are undergrads as competent technically as most grad students.

19

So why shouldn't undergrads be able to start startups as well as grad students?

4–9

Graduating seniors used to have two choices: a job or grad school. There's increasingly a third: start your own startup. In the late 90s undergrads went to work for startups instead of grad school; that may return, but this time they'll start their own.

10–13

The most ambitious students will ask: Why wait till you graduate? Why not start a startup in college? In fact, why go to college at all?

14–19

A year and a half ago I noted that the founders of Yahoo, Google, and Microsoft averaged 24, and asked: if grad students can start startups, why not undergrads? Graduation is a bureaucratic change, not a biological one, and plenty of undergrads are as technically competent. So why shouldn't they?

2–19

Graduating seniors used to choose between a job and grad school; a third option, starting a startup, is becoming real. If grad students can do it, why not undergrads—or college students at all?

21

I now realize that something does change at graduation: you lose a huge excuse for failing.

22

Regardless of how complex your life is, you'll find that everyone else, including your family and friends, will discard all the low bits and regard you as having a single occupation at any given time.

23

If you're in college and have a summer job writing software, you still read as a student.

24

Whereas if you graduate and get a job programming, you'll be instantly regarded by everyone as a programmer.

25

The problem with starting a startup while you're still in school is that there's a built-in escape hatch.

26

If you start a startup in the summer between your junior and senior year, it reads to everyone as a summer job.

27

So if it goes nowhere, big deal; you return to school in the fall with all the other seniors; no one regards you as a failure, because your occupation is student, and you didn't fail at that.

28

Whereas if you start a startup just one year later, after you graduate, as long as you're not accepted to grad school in the fall the startup reads to everyone as your occupation.

29

You're now a startup founder, so you have to do well at that.

30

For nearly everyone, the opinion of one's peers is the most powerful motivator of all—more powerful even than the nominal goal of most startup founders, getting rich. [1] About a month into each funding cycle we have an event called Prototype Day where each startup presents to the others what they've got so far.

31

You might think they wouldn't need any more motivation.

32

They're working on their cool new idea; they have funding for the immediate future; and they're playing a game with only two outcomes: wealth or failure.

33

You'd think that would be motivation enough.

34

And yet the prospect of a demo pushes most of them into a rush of activity.

35

Even if you start a startup explicitly to get rich, the money you might get seems pretty theoretical most of the time.

36

What drives you day to day is not wanting to look bad.

37

You probably can't change that.

38

Even if you could, I don't think you'd want to; someone who really, truly doesn't care what his peers think of him is probably a psychopath.

39

So the best you can do is consider this force like a wind, and set up your boat accordingly.

40

If you know your peers are going to push you in some direction, choose good peers, and position yourself so they push you in a direction you like.

41

Graduation changes the prevailing winds, and those make a difference.

42

Starting a startup is so hard that it's a close call even for the ones that succeed.

43

However high a startup may be flying now, it probably has a few leaves stuck in the landing gear from those trees it barely cleared at the end of the runway.

44

In such a close game, the smallest increase in the forces against you can be enough to flick you over the edge into failure.

45

When we first started Y Combinator we encouraged people to start startups while they were still in college.

46

That's partly because Y Combinator began as a kind of summer program.

47

We've kept the program shape—all of us having dinner together once a week turns out to be a good idea—but we've decided now that the party line should be to tell people to wait till they graduate.

21–24

Something does change at graduation: you lose a huge excuse for failing. A summer job writing software still reads as student; graduate and program for a job, and you're instantly a programmer.

25–29

The problem with starting a startup in school is the built-in escape hatch. Start one before senior year and it reads as a summer job; if it goes nowhere, you're still a student. Start it after you graduate and the startup is your occupation—so you have to do well at it.

30–34

For nearly everyone, peer opinion is the most powerful motivator of all—more powerful even than getting rich. At Prototype Day, startups show each other what they've got. They already have a cool idea, funding, and a game with two outcomes—yet the prospect of a demo pushes most into a rush of activity.

35–36

Even if you start a startup explicitly to get rich, the money seems theoretical most of the time. What drives you day to day is not wanting to look bad.

37–40

You can't change that, and wouldn't want to; someone who truly doesn't care what his peers think is probably a psychopath. So treat this force like a wind: choose good peers and position yourself so they push you where you want to go.

41–47

Graduation changes the prevailing winds. Starting a startup is a close call even for the ones that succeed—however high one flies, it has leaves stuck in the landing gear from trees it barely cleared. The smallest increase in the forces against you can flick you over the edge. When we started Y Combinator we encouraged starting in college; now we tell people to wait till they graduate.

21–47

Graduation removes a huge excuse for failing: once you're out, a failed startup is your occupation, not a summer job. Peer opinion, not money, is the real day-to-day motivator, so YC now tells people to wait.

49

Does that mean you can't start a startup in college?

50

Not at all.

51

Sam Altman, the co-founder of Loopt, had just finished his sophomore year when we funded them, and Loopt is probably the most promising of all the startups we've funded so far.

52

But Sam Altman is a very unusual guy.

53

Within about three minutes of meeting him, I remember thinking "Ah, so this is what Bill Gates must have been like when he was 19."

54

If it can work to start a startup during college, why do we tell people not to?

55

For the same reason that the probably apocryphal violinist, whenever he was asked to judge someone's playing, would always say they didn't have enough talent to make it as a pro.

56

Succeeding as a musician takes determination as well as talent, so this answer works out to be the right advice for everyone.

57

The ones who are uncertain believe it and give up, and the ones who are sufficiently determined think "screw that, I'll succeed anyway."

58

So our official policy now is only to fund undergrads we can't talk out of it.

59

And frankly, if you're not certain, you should wait.

60

It's not as if all the opportunities to start companies are going to be gone if you don't do it now.

61

Maybe the window will close on some idea you're working on, but that won't be the last idea you'll have.

62

For every idea that times out, new ones become feasible.

63

Historically the opportunities to start startups have only increased with time.

64

In that case, you might ask, why not wait longer?

65

Why not go work for a while, or go to grad school, and then start a startup?

66

And indeed, that might be a good idea.

67

If I had to pick the sweet spot for startup founders, based on who we're most excited to see applications from, I'd say it's probably the mid-twenties.

68

Why?

69

What advantages does someone in their mid-twenties have over someone who's 21?

70

And why isn't it older?

71

What can 25 year olds do that 32 year olds can't?

72

Those turn out to be questions worth examining.

49–53

Can't you start a startup in college? Of course. Sam Altman, co-founder of Loopt, had just finished his sophomore year when we funded them. But Sam is unusual—within three minutes I thought, "Ah, so this is what Bill Gates must have been like at 19."

54–57

If it can work in college, why tell people not to? For the same reason the apocryphal violinist told everyone they lacked the talent to go pro. Succeeding takes determination as well as talent: the uncertain give up, and the determined think "screw that, I'll succeed anyway."

58–63

So we only fund undergrads we can't talk out of it. If you're not certain, you should wait. The opportunities won't be gone—historically they've only increased with time.

64–72

Why not wait even longer—work a while, or go to grad school first? Maybe. If I had to pick the sweet spot, I'd say the mid-twenties. What can a 25 year old do that a 21 year old can't—and why isn't it older, what can 25 year olds do that 32 year olds can't? Questions worth examining.

49–72

You can start in college—Sam Altman did—but we tell people not to, like the violinist who discourages everyone: the determined ignore it. If you're unsure, wait; the sweet spot is probably the mid-twenties.

74

If you start a startup soon after college, you'll be a young founder by present standards, so you should know what the relative advantages of young founders are.

75

They're not what you might think.

76

As a young founder your strengths are: stamina, poverty, rootlessness, colleagues, and ignorance.

77

The importance of stamina shouldn't be surprising.

78

If you've heard anything about startups you've probably heard about the long hours.

79

As far as I can tell these are universal.

80

I can't think of any successful startups whose founders worked 9 to 5.

81

And it's particularly necessary for younger founders to work long hours because they're probably not as efficient as they'll be later.

74–76

If you start soon after college you'll be a young founder. Your strengths aren't what you might think: stamina, poverty, rootlessness, colleagues, and ignorance.

77–81

Stamina shouldn't be surprising. The long hours are universal—I can't think of a successful startup whose founders worked 9 to 5. It's especially necessary for younger founders, who aren't yet as efficient as they'll be.

74–81

As a young founder your strengths aren't what you'd think: stamina, poverty, rootlessness, colleagues, and ignorance. The first, stamina, is unsurprising—startups demand long hours, especially for the not-yet-efficient.

83

Your second advantage, poverty, might not sound like an advantage, but it is a huge one.

84

Poverty implies you can live cheaply, and this is critically important for startups.

85

Nearly every startup that fails, fails by running out of money.

86

It's a little misleading to put it this way, because there's usually some other underlying cause.

87

But regardless of the source of your problems, a low burn rate gives you more opportunity to recover from them.

88

And since most startups make all kinds of mistakes at first, room to recover from mistakes is a valuable thing to have.

89

Most startups end up doing something different than they planned.

90

The way the successful ones find something that works is by trying things that don't.

91

So the worst thing you can do in a startup is to have a rigid, pre-ordained plan and then start spending a lot of money to implement it.

92

Better to operate cheaply and give your ideas time to evolve.

93

Recent grads can live on practically nothing, and this gives you an edge over older founders, because the main cost in software startups is people.

94

The guys with kids and mortgages are at a real disadvantage.

95

This is one reason I'd bet on the 25 year old over the 32 year old.

96

The 32 year old probably is a better programmer, but probably also has a much more expensive life.

97

Whereas a 25 year old has some work experience (more on that later) but can live as cheaply as an undergrad.

98

Robert Morris and I were 29 and 30 respectively when we started Viaweb, but fortunately we still lived like 23 year olds.

99

We both had roughly zero assets.

100

I would have loved to have a mortgage, since that would have meant I had a house.

101

But in retrospect having nothing turned out to be convenient.

102

I wasn't tied down and I was used to living cheaply.

103

Even more important than living cheaply, though, is thinking cheaply.

104

One reason the Apple II was so popular was that it was cheap.

105

The computer itself was cheap, and it used cheap, off-the-shelf peripherals like a cassette tape recorder for data storage and a TV as a monitor.

106

And you know why?

107

Because Woz designed this computer for himself, and he couldn't afford anything more.

108

We benefitted from the same phenomenon.

109

Our prices were daringly low for the time.

110

The top level of service was $300 a month, which was an order of magnitude below the norm.

111

In retrospect this was a smart move, but we didn't do it because we were smart. $300 a month seemed like a lot of money to us.

112

Like Apple, we created something inexpensive, and therefore popular, simply because we were poor.

113

A lot of startups have that form: someone comes along and makes something for a tenth or a hundredth of what it used to cost, and the existing players can't follow because they don't even want to think about a world in which that's possible.

114

Traditional long distance carriers, for example, didn't even want to think about VoIP. (It was coming, all the same.)

115

Being poor helps in this game, because your own personal bias points in the same direction technology evolves in.

83–88

Poverty might not sound like an advantage, but it's huge: you can live cheaply. Nearly every startup that fails, fails by running out of money, and a low burn rate gives you room to recover from the mistakes every startup makes at first.

89–92

Most startups end up doing something different than they planned; the successful ones find what works by trying things that don't. So the worst thing you can do is follow a rigid plan and spend a lot of money on it. Better to operate cheaply and let your ideas evolve.

93–97

Recent grads can live on practically nothing, an edge over older founders, since the main cost in software startups is people. I'd bet on the 25 year old over the 32 year old: probably a worse programmer, but a much cheaper life.

103–107

Even more important than living cheaply is thinking cheaply. The Apple II was popular partly because it was cheap—off-the-shelf peripherals, a cassette recorder, a TV for a monitor—because Woz designed it for himself, and couldn't afford more.

108–112

We benefitted from the same thing. Our top tier was $300 a month, an order of magnitude below the norm—not because we were smart, but because $300 seemed like a lot of money to us. Like Apple, we made something inexpensive, and therefore popular, simply because we were poor.

113–115

A lot of startups have that form: someone makes something for a tenth of what it cost, and incumbents can't follow because they won't imagine that world—as long-distance carriers didn't want to think about VoIP. Being poor helps, because your own bias points where technology is heading.

83–115

Poverty is a huge advantage: you can live cheaply, and a low burn rate buys room to recover from the mistakes every startup makes. Thinking cheaply matters even more—being poor points your bias where technology is heading.

117

The advantages of rootlessness are similar to those of poverty.

118

When you're young you're more mobile—not just because you don't have a house or much stuff, but also because you're less likely to have serious relationships.

119

This turns out to be important, because a lot of startups involve someone moving.

120

The founders of Kiko, for example, are now en route to the Bay Area to start their next startup.

121

It's a better place for what they want to do.

122

And it was easy for them to decide to go, because neither as far as I know has a serious girlfriend, and everything they own will fit in one car—or more precisely, will either fit in one car or is crappy enough that they don't mind leaving it behind.

123

They at least were in Boston.

124

What if they'd been in Nebraska, like Evan Williams was at their age?

125

Someone wrote recently that the drawback of Y Combinator was that you had to move to participate.

126

It couldn't be any other way.

127

The kind of conversations we have with founders, we have to have in person.

128

We fund a dozen startups at a time, and we can't be in a dozen places at once.

129

But even if we could somehow magically save people from moving, we wouldn't.

130

We wouldn't be doing founders a favor by letting them stay in Nebraska.

131

Places that aren't startup hubs [blocked] are toxic to startups.

132

You can tell that from indirect evidence.

133

You can tell how hard it must be to start a startup in Houston or Chicago or Miami from the microscopically small number, per capita, that succeed there.

134

I don't know exactly what's suppressing all the startups in these towns—probably a hundred subtle little things—but something must be. [2]

135

Maybe this will change.

136

Maybe the increasing cheapness of startups will mean they'll be able to survive anywhere, instead of only in the most hospitable environments.

137

Maybe 37signals is the pattern for the future.

138

But maybe not.

139

Historically there have always been certain towns that were centers for certain industries, and if you weren't in one of them you were at a disadvantage.

140

So my guess is that 37signals is an anomaly.

141

We're looking at a pattern much older than "Web 2.0" here.

117–119

Rootlessness works like poverty. When you're young you're more mobile—lacking a house or much stuff, and less likely to have serious relationships. That matters, because a lot of startups involve someone moving.

120–122

The founders of Kiko, for example, are en route to the Bay Area for their next startup. It was easy to decide: neither has a serious girlfriend, and everything they own fits in one car—or is crappy enough they don't mind leaving it.

125–134

Even if we could save founders from moving, we wouldn't—we'd be doing them no favor by letting them stay in Nebraska. Places that aren't startup hubs [blocked] are toxic to startups; you can tell from the tiny number that succeed, per capita, in Houston or Chicago or Miami.

135–141

Maybe this will change—maybe cheaper startups will survive anywhere, and 37signals is the future. But historically there have always been towns that were centers for certain industries. So my guess is 37signals is an anomaly: a pattern much older than "Web 2.0."

117–141

Rootlessness, like poverty, makes you mobile, which matters because startups often involve moving—and they have to move to a hub. Places that aren't startup hubs are toxic to startups; the pattern is older than Web 2.0.

143

Perhaps the reason more startups per capita happen in the Bay Area than Miami is simply that there are more founder-type people there.

144

Successful startups are almost never started by one person.

145

Usually they begin with a conversation in which someone mentions that something would be a good idea for a company, and his friend says, "Yeah, that is a good idea, let's try it."

146

If you're missing that second person who says "let's try it," the startup never happens.

147

And that is another area where undergrads have an edge.

148

They're surrounded by people willing to say that.

149

At a good college you're concentrated together with a lot of other ambitious and technically minded people—probably more concentrated than you'll ever be again.

150

If your nucleus spits out a neutron, there's a good chance it will hit another nucleus.

151

The number one question people ask us at Y Combinator is: Where can I find a co-founder?

152

That's the biggest problem for someone starting a startup at 30.

153

When they were in school they knew a lot of good co-founders, but by 30 they've either lost touch with them or these people are tied down by jobs they don't want to leave.

154

Viaweb was an anomaly in this respect too.

155

Though we were comparatively old, we weren't tied down by impressive jobs.

156

I was trying to be an artist, which is not very constraining, and Robert, though 29, was still in grad school due to a little interruption in his academic career back in 1988.

157

So arguably the Worm made Viaweb possible.

158

Otherwise Robert would have been a junior professor at that age, and he wouldn't have had time to work on crazy speculative projects with me.

159

Most of the questions people ask Y Combinator we have some kind of answer for, but not the co-founder question.

160

There is no good answer.

161

Co-founders really should be people you already know.

162

And by far the best place to meet them is school.

163

You have a large sample of smart people; you get to compare how they all perform on identical tasks; and everyone's life is pretty fluid.

164

A lot of startups grow out of schools for this reason.

165

Google, Yahoo, and Microsoft, among others, were all founded by people who met in school. (In Microsoft's case, it was high school.)

166

Many students feel they should wait and get a little more experience before they start a company.

167

All other things being equal, they should.

168

But all other things are not quite as equal as they look.

169

Most students don't realize how rich they are in the scarcest ingredient in startups, co-founders.

170

If you wait too long, you may find that your friends are now involved in some project they don't want to abandon.

171

The better they are, the more likely this is to happen.

172

One way to mitigate this problem might be to actively plan your startup while you're getting those n years of experience.

173

Sure, go off and get jobs or go to grad school or whatever, but get together regularly to scheme, so the idea of starting a startup stays alive in everyone's brain.

174

I don't know if this works, but it can't hurt to try.

175

It would be helpful just to realize what an advantage you have as students.

176

Some of your classmates are probably going to be successful startup founders; at a great technical university, that is a near certainty.

177

So which ones?

178

If I were you I'd look for the people who are not just smart, but incurable builders.

179

Look for the people who keep starting projects, and finish at least some of them.

180

That's what we look for.

181

Above all else, above academic credentials and even the idea you apply with, we look for people who build things.

143–150

Successful startups are almost never started by one person: someone says an idea would make a good company, and a friend says "let's try it." Miss that second person and it never happens. That's another undergrad edge—at a good college you're concentrated with ambitious, technical people more than you'll ever be again. If your nucleus spits out a neutron, there's a good chance it hits another nucleus.

151–153

The number one question people ask us is: Where can I find a co-founder? That's the biggest problem for someone starting at 30. In school they knew lots of good co-founders, but by 30 they've lost touch, or those people are tied down by jobs.

159–165

There's no good answer to the co-founder question. They should be people you already know, and by far the best place to meet them is school: a large sample of smart people, performing identical tasks, lives still fluid. Google, Yahoo, and Microsoft were all founded by people who met in school. (In Microsoft's case, high school.)

166–171

Many students feel they should wait and get more experience. All else equal, they should—but it isn't. Most don't realize how rich they are in the scarcest ingredient, co-founders. Wait too long and your friends may be in some project they won't abandon—and the better they are, the likelier this is.

175–181

Just realize what an advantage you have as students. Some of your classmates are going to be successful founders. So which ones? Look for people who are not just smart but incurable builders—who keep starting projects and finish some. Above all, above credentials and even the idea you apply with, we look for people who build things.

143–181

The scarcest startup ingredient is co-founders, and school is by far the best place to find them—a concentration of smart people you'll never have again. Don't wait too long, and look for incurable builders.

183

The other place co-founders meet is at work.

184

Fewer do than at school, but there are things you can do to improve the odds.

185

The most important, obviously, is to work somewhere that has a lot of smart, young people.

186

Another is to work for a company located in a startup hub.

187

It will be easier to talk a co-worker into quitting with you in a place where startups are happening all around you.

188

You might also want to look at the employment agreement you sign when you get hired.

189

Most will say that any ideas you think of while you're employed by the company belong to them.

190

In practice it's hard for anyone to prove what ideas you had when, so the line gets drawn at code.

191

If you're going to start a startup, don't write any of the code while you're still employed.

192

Or at least discard any code you wrote while still employed and start over.

193

It's not so much that your employer will find out and sue you.

194

It won't come to that; investors or acquirers or (if you're so lucky) underwriters will nail you first. Between t = 0 and when you buy that yacht, someone is going to ask if any of your code legally belongs to anyone else, and you need to be able to say no. [3]

195

The most overreaching employee agreement I've seen so far is Amazon's.

196

In addition to the usual clauses about owning your ideas, you also can't be a founder of a startup that has another founder who worked at Amazon—even if you didn't know them or even work there at the same time.

197

I suspect they'd have a hard time enforcing this, but it's a bad sign they even try.

198

There are plenty of other places to work; you may as well choose one that keeps more of your options open.

199

Speaking of cool places to work, there is of course Google.

200

But I notice something slightly frightening about Google: zero startups come out of there.

201

In that respect it's a black hole.

202

People seem to like working at Google too much to leave.

203

So if you hope to start a startup one day, the evidence so far suggests you shouldn't work there.

204

I realize this seems odd advice.

205

If they make your life so good that you don't want to leave, why not work there?

206

Because, in effect, you're probably getting a local maximum.

207

You need a certain activation energy to start a startup.

208

So an employer who's fairly pleasant to work for can lull you into staying indefinitely, even if it would be a net win for you to leave. [4]

209

The best place to work, if you want to start a startup, is probably a startup.

210

In addition to being the right sort of experience, one way or another it will be over quickly.

211

You'll either end up rich, in which case problem solved, or the startup will get bought, in which case it it will start to suck to work there and it will be easy to leave, or most likely, the thing will blow up and you'll be free again.

183–187

Co-founders also meet at work, fewer than at school. Work somewhere with a lot of smart, young people, ideally in a startup hub, where it's easier to talk a co-worker into quitting with you.

188–194

Read the employment agreement you sign. Most claim any ideas you have while employed, though the line gets drawn at code. So don't write any code while still employed—or discard it and start over. Before you buy that yacht, someone doing due diligence will ask whether your code belongs to anyone else, and you need to say no.

195–198

The most overreaching agreement I've seen is Amazon's: besides owning your ideas, you can't co-found a startup with anyone who worked at Amazon, even if you never met or overlapped. They'd struggle to enforce it, but it's a bad sign they try.

199–203

Speaking of cool places, there's Google—but something slightly frightening: zero startups come out of there. In that respect it's a black hole; people like working there too much to leave. So if you hope to start a startup, the evidence suggests you shouldn't work at Google.

204–211

Odd advice, I know: if it's so good, why leave? Because you're getting a local maximum. You need a certain activation energy to start a startup, and a pleasant employer can lull you into staying even when leaving would be a net win. The best place to work is a startup—one way or another it's over quickly: you end up rich, it gets bought, or it blows up and you're free again.

183–211

Co-founders also meet at work, so work somewhere with smart young people in a startup hub—and read your employment agreement, since it likely claims your ideas. Avoid black holes like Google that no startups escape.

213

Your final advantage, ignorance, may not sound very useful.

214

I deliberately used a controversial word for it; you might equally call it innocence.

215

But it seems to be a powerful force.

216

My Y Combinator co-founder Jessica Livingston is just about to publish a book of interviews with startup founders, and I noticed a remarkable pattern in them.

217

One after another said that if they'd known how hard it would be, they would have been too intimidated to start.

218

Ignorance can be useful when it's a counterweight to other forms of stupidity.

219

It's useful in starting startups because you're capable of more than you realize.

220

Starting startups is harder than you expect, but you're also capable of more than you expect, so they balance out.

221

Most people look at a company like Apple and think, how could I ever make such a thing?

222

Apple is an institution, and I'm just a person.

223

But every institution was at one point just a handful of people in a room deciding to start something.

224

Institutions are made up, and made up by people no different from you.

225

I'm not saying everyone could start a startup.

226

I'm sure most people couldn't; I don't know much about the population at large.

227

When you get to groups I know well, like hackers, I can say more precisely.

228

At the top schools, I'd guess as many as a quarter of the CS majors could make it as startup founders if they wanted.

229

That "if they wanted" is an important qualification—so important that it's almost cheating to append it like that—because once you get over a certain threshold of intelligence, which most CS majors at top schools are past, the deciding factor in whether you succeed as a founder is how much you want to.

230

You don't have to be that smart.

231

If you're not a genius, just start a startup in some unsexy field where you'll have less competition, like software for human resources departments.

232

I picked that example at random, but I feel safe in predicting that whatever they have now, it wouldn't take genius to do better.

233

There are a lot of people out there working on boring stuff who are desperately in need of better software, so however short you think you fall of Larry and Sergey, you can ratchet down the coolness of the idea far enough to compensate.

234

As well as preventing you from being intimidated, ignorance can sometimes help you discover new ideas. Steve Wozniak put this very strongly:

235

All the best things that I did at Apple came from (a) not having money and (b) not having done it before, ever. Every single thing that we came out with that was really great, I'd never once done that thing in my life.

236

When you know nothing, you have to reinvent stuff for yourself, and if you're smart your reinventions may be better than what preceded them.

237

This is especially true in fields where the rules change.

238

All our ideas about software were developed in a time when processors were slow, and memories and disks were tiny.

239

Who knows what obsolete assumptions are embedded in the conventional wisdom?

240

And the way these assumptions are going to get fixed is not by explicitly deallocating them, but by something more akin to garbage collection.

241

Someone ignorant but smart will come along and reinvent everything, and in the process simply fail to reproduce certain existing ideas.

213–217

Your final advantage, ignorance, may not sound useful; call it innocence. My co-founder Jessica Livingston is about to publish a book of interviews with founders, and one after another said that if they'd known how hard it would be, they'd have been too intimidated to start.

218–220

Ignorance is useful as a counterweight to other forms of stupidity. Starting startups is harder than you expect, but you're also capable of more than you expect, so they balance out.

221–224

Most people look at a company like Apple and think, how could I ever make such a thing? But every institution was once a handful of people in a room deciding to start something. Institutions are made up, by people no different from you.

225–233

I'm not saying everyone could—most couldn't. But I'd guess a quarter of CS majors at top schools could make it if they wanted. Past a certain intelligence the deciding factor is how much you want to. You needn't be a genius: start in some unsexy field with less competition, like software for HR departments, and ratchet the coolness down to compensate.

234–241

Steve Wozniak put it strongly: the best things he did at Apple came from not having money and not having done it before. When you know nothing you reinvent things, and if you're smart your reinventions may be better—especially where the rules change. Obsolete assumptions get fixed not by deallocating them but by garbage collection: someone ignorant but smart reinvents everything and fails to reproduce them.

213–241

Ignorance—or innocence—is a powerful force: founders say that had they known how hard it would be, they'd have been too intimidated to start. You're capable of more than you realize, and ignorance lets you reinvent things better.

243

So much for the advantages of young founders.

244

What about the disadvantages?

245

I'm going to start with what goes wrong and try to trace it back to the root causes.

246

What goes wrong with young founders is that they build stuff that looks like class projects.

247

It was only recently that we figured this out ourselves.

248

We noticed a lot of similarities between the startups that seemed to be falling behind, but we couldn't figure out how to put it into words.

249

Then finally we realized what it was: they were building class projects.

250

But what does that really mean?

251

What's wrong with class projects?

252

What's the difference between a class project and a real startup?

253

If we could answer that question it would be useful not just to would-be startup founders but to students in general, because we'd be a long way toward explaining the mystery of the so-called real world.

254

There seem to be two big things missing in class projects: (1) an iterative definition of a real problem and (2) intensity.

255

The first is probably unavoidable.

256

Class projects will inevitably solve fake problems. For one thing, real problems are rare and valuable.

257

If a professor wanted to have students solve real problems, he'd face the same paradox as someone trying to give an example of whatever "paradigm" might succeed the Standard Model of physics.

258

There may well be something that does, but if you could think of an example you'd be entitled to the Nobel Prize.

259

Similarly, good new problems are not to be had for the asking.

260

In technology the difficulty is compounded by the fact that real startups tend to discover the problem they're solving by a process of evolution.

261

Someone has an idea for something; they build it; and in doing so (and probably only by doing so) they realize the problem they should be solving is another one.

262

Even if the professor let you change your project description on the fly, there isn't time enough to do that in a college class, or a market to supply evolutionary pressures.

263

So class projects are mostly about implementation, which is the least of your problems in a startup.

264

It's not just that in a startup you work on the idea as well as implementation.

265

The very implementation is different.

266

Its main purpose is to refine the idea.

267

Often the only value of most of the stuff you build in the first six months is that it proves your initial idea was mistaken.

268

And that's extremely valuable.

269

If you're free of a misconception that everyone else still shares, you're in a powerful position.

270

But you're not thinking that way about a class project.

271

Proving your initial plan was mistaken would just get you a bad grade.

272

Instead of building stuff to throw away, you tend to want every line of code to go toward that final goal of showing you did a lot of work.

273

That leads to our second difference: the way class projects are measured.

274

Professors will tend to judge you by the distance between the starting point and where you are now.

275

If someone has achieved a lot, they should get a good grade.

276

But customers will judge you from the other direction: the distance remaining between where you are now and the features they need.

277

The market doesn't give a shit how hard you worked.

278

Users just want your software to do what they need, and you get a zero otherwise.

279

That is one of the most distinctive differences between school and the real world: there is no reward for putting in a good effort.

280

In fact, the whole concept of a "good effort" is a fake idea adults invented to encourage kids.

281

It is not found in nature.

243–249

So much for the advantages. What goes wrong with young founders is that they build stuff that looks like class projects. We figured this out only recently: we'd noticed similarities among the startups falling behind, then realized they were building class projects.

250–254

But what's the difference between a class project and a real startup? There seem to be two big things missing in class projects: (1) an iterative definition of a real problem and (2) intensity.

255–259

The first is unavoidable. Class projects solve fake problems, because real ones are rare and valuable. A professor wanting students to solve real problems would face the same paradox as someone asked to name the paradigm succeeding the Standard Model—give an example and you'd deserve the Nobel Prize.

260–263

In technology it's compounded because real startups discover their problem by evolution: someone has an idea, builds it, and only then realizes the problem they should solve is another. A college class has no time for that, and no market to supply evolutionary pressure. So class projects are mostly about implementation—the least of your problems.

264–272

And the implementation itself is different: its purpose is to refine the idea. Often the only value of what you build in the first six months is proving your initial idea was mistaken—which is extremely valuable. But in a class project, proving your plan wrong just gets a bad grade, so you want every line of code to show you did work.

273–281

That's the second difference: how you're measured. Professors judge you by the distance from your starting point to now. But customers judge from the other direction—the distance remaining to the features they need. The market doesn't give a shit how hard you worked. The whole concept of a "good effort" is a fake idea adults invented to encourage kids; it is not found in nature.

243–281

What goes wrong with young founders is that they build class projects, which miss two things: an iterative definition of a real problem, and intensity. Real startups discover their problem by evolution, and the market judges only what's left to do.

283

Such lies seem to be helpful to kids.

284

But unfortunately when you graduate they don't give you a list of all the lies they told you during your education.

285

You have to get them beaten out of you by contact with the real world.

286

And this is why so many jobs want work experience.

287

I couldn't understand that when I was in college.

288

I knew how to program.

289

In fact, I could tell I knew how to program better than most people doing it for a living.

290

So what was this mysterious "work experience" and why did I need it?

291

Now I know what it is, and part of the confusion is grammatical.

292

Describing it as "work experience" implies it's like experience operating a certain kind of machine, or using a certain programming language.

293

But really what work experience refers to is not some specific expertise, but the elimination of certain habits left over from childhood.

294

One of the defining qualities of kids is that they flake.

295

When you're a kid and you face some hard test, you can cry and say "I can't" and they won't make you do it.

296

Of course, no one can make you do anything in the grownup world either.

297

What they do instead is fire you.

298

And when motivated by that you find you can do a lot more than you realized.

299

So one of the things employers expect from someone with "work experience" is the elimination of the flake reflex—the ability to get things done, with no excuses.

300

The other thing you get from work experience is an understanding of what work is, and in particular, how intrinsically horrible it is.

301

Fundamentally the equation is a brutal one: you have to spend most of your waking hours doing stuff someone else wants, or starve.

302

There are a few places where the work is so interesting that this is concealed, because what other people want done happens to coincide with what you want to work on.

303

But you only have to imagine what would happen if they diverged to see the underlying reality.

304

It's not so much that adults lie to kids about this as never explain it.

305

They never explain what the deal is with money.

306

You know from an early age that you'll have some sort of job, because everyone asks what you're going to "be" when you grow up.

307

What they don't tell you is that as a kid you're sitting on the shoulders of someone else who's treading water, and that starting working means you get thrown into the water on your own, and have to start treading water yourself or sink.

308

"Being" something is incidental; the immediate problem is not to drown.

309

The relationship between work and money tends to dawn on you only gradually.

310

At least it did for me.

311

One's first thought tends to be simply "This sucks.

312

I'm in debt.

313

Plus I have to get up on monday and go to work."

314

Gradually you realize that these two things are as tightly connected as only a market can make them.

315

So the most important advantage 24 year old founders have over 20 year old founders is that they know what they're trying to avoid.

316

To the average undergrad the idea of getting rich translates into buying Ferraris, or being admired.

317

To someone who has learned from experience about the relationship between money and work, it translates to something way more important: it means you get to opt out of the brutal equation that governs the lives of 99.9% of people.

318

Getting rich means you can stop treading water.

319

Someone who gets this will work much harder at making a startup succeed—with the proverbial energy of a drowning man, in fact.

320

But understanding the relationship between money and work also changes the way you work.

321

You don't get money just for working, but for doing things other people want.

322

Someone who's figured that out will automatically focus more on the user.

323

And that cures the other half of the class-project syndrome.

324

After you've been working for a while, you yourself tend to measure what you've done the same way the market does.

325

Of course, you don't have to spend years working to learn this stuff.

326

If you're sufficiently perceptive you can grasp these things while you're still in school.

327

Sam Altman did.

328

He must have, because Loopt is no class project.

329

And as his example suggests, this can be valuable knowledge.

330

At a minimum, if you get this stuff, you already have most of what you gain from the "work experience" employers consider so desirable.

331

But of course if you really get it, you can use this information in a way that's more valuable to you than that.

283–290

When you graduate no one hands you the list of lies; you have to get them beaten out of you by the real world. That's why so many jobs want work experience. I couldn't understand that in college—I knew how to program, better than most people doing it for a living. So what was this mysterious thing?

291–293

Part of the confusion is grammatical. "Work experience" sounds like operating a machine. But what it really refers to isn't expertise—it's the elimination of certain habits left over from childhood.

294–299

One defining quality of kids is that they flake. Face a hard test and you can cry "I can't." No one can make you as a grownup either—instead they fire you, and motivated by that you find you can do far more than you realized. So employers want the flake reflex gone: getting things done, no excuses.

300–303

The other thing is understanding how intrinsically horrible work is. The equation is brutal: you spend most of your waking hours doing what someone else wants, or starve. A few interesting jobs conceal this—but imagine what you want and what others want diverging, and you see the reality.

304–308

Adults don't so much lie about this as never explain the deal with money. As a kid you're on the shoulders of someone treading water, and starting work means you get thrown in to tread water yourself. "Being" something is incidental; the problem is not to drown.

309–318

So the most important advantage 24 year olds have over 20 year olds is knowing what they're trying to avoid. To an undergrad, getting rich means Ferraris; to someone who's learned, it means opting out of the brutal equation. Getting rich means you can stop treading water.

319–331

Someone who gets this works much harder—with the energy of a drowning man. It also changes how you work: you get money not for working but for doing what others want, so you focus more on the user. And that cures the other half of class-project syndrome. You needn't spend years to learn this—Sam Altman grasped it in school, which is why Loopt is no class project.

283–331

Work experience isn't expertise but the elimination of childhood habits: the flake reflex, and ignorance of the brutal equation that you must do what others want or starve. Knowing this makes you focus on users—curing the other half of class-project syndrome.

333

So suppose you think you might start a startup at some point, either when you graduate or a few years after.

334

What should you do now?

335

For both jobs and grad school, there are ways to prepare while you're in college.

336

If you want to get a job when you graduate, you should get summer jobs at places you'd like to work.

337

If you want to go to grad school, it will help to work on research projects as an undergrad.

338

What's the equivalent for startups?

339

How do you keep your options maximally open?

340

One thing you can do while you're still in school is to learn how startups work.

341

Unfortunately that's not easy.

342

Few if any colleges have classes about startups.

343

There may be business school classes on entrepreneurship, as they call it over there, but these are likely to be a waste of time.

344

Business schools like to talk about startups, but philosophically they're at the opposite end of the spectrum.

345

Most books on startups also seem to be useless.

346

I've looked at a few and none get it right.

347

Books in most fields are written by people who know the subject from experience, but for startups there's a unique problem: by definition the founders of successful startups don't need to write books to make money.

348

As a result most books on the subject end up being written by people who don't understand it.

349

So I'd be skeptical of classes and books.

350

The way to learn about startups is by watching them in action, preferably by working at one.

351

How do you do that as an undergrad?

352

Probably by sneaking in through the back door.

353

Just hang around a lot and gradually start doing things for them.

354

Most startups are (or should be) very cautious about hiring.

355

Every hire increases the burn rate, and bad hires early on are hard to recover from.

356

However, startups usually have a fairly informal atmosphere, and there's always a lot that needs to be done.

357

If you just start doing stuff for them, many will be too busy to shoo you away.

358

You can thus gradually work your way into their confidence, and maybe turn it into an official job later, or not, whichever you prefer.

359

This won't work for all startups, but it would work for most I've known.

360

Number two, make the most of the great advantage of school: the wealth of co-founders.

361

Look at the people around you and ask yourself which you'd like to work with.

362

When you apply that test, you may find you get surprising results.

363

You may find you'd prefer the quiet guy you've mostly ignored to someone who seems impressive but has an attitude to match.

364

I'm not suggesting you suck up to people you don't really like because you think one day they'll be successful.

365

Exactly the opposite, in fact: you should only start a startup with someone you like, because a startup will put your friendship through a stress test. I'm just saying you should think about who you really admire and hang out with them, instead of whoever circumstances throw you together with.

366

Another thing you can do is learn skills that will be useful to you in a startup.

367

These may be different from the skills you'd learn to get a job.

368

For example, thinking about getting a job will make you want to learn programming languages you think employers want, like Java and C++.

369

Whereas if you start a startup, you get to pick the language, so you have to think about which will actually let you get the most done.

370

If you use that test you might end up learning Ruby or Python instead.

371

But the most important skill for a startup founder isn't a programming technique.

372

It's a knack for understanding users and figuring out how to give them what they want.

373

I know I repeat this, but that's because it's so important.

374

And it's a skill you can learn, though perhaps habit might be a better word.

375

Get into the habit of thinking of software as having users.

376

What do those users want?

377

What would make them say wow?

378

This is particularly valuable for undergrads, because the concept of users is missing from most college programming classes.

379

The way you get taught programming in college would be like teaching writing as grammar, without mentioning that its purpose is to communicate something to an audience.

380

Fortunately an audience for software is now only an http request away.

381

So in addition to the programming you do for your classes, why not build some kind of website people will find useful?

382

At the very least it will teach you how to write software with users.

383

In the best case, it might not just be preparation for a startup, but the startup itself, like it was for Yahoo and Google.

333–339

So suppose you might start a startup someday. What should you do now? For jobs and grad school you prepare with summer jobs and research projects. What's the equivalent for startups?

340–349

One thing is to learn how startups work, which isn't easy. Few colleges have classes, and business-school entrepreneurship classes are likely a waste. Most books are useless too: the founders of successful startups don't need to write books for money, so the books get written by people who don't understand the subject.

350–359

The way to learn is by watching startups in action, preferably by working at one. As an undergrad, sneak in the back door: hang around and gradually start doing things for them. Startups are cautious about hiring, but they're informal and always have a lot that needs doing—start doing stuff and many will be too busy to shoo you away.

360–365

Number two, make the most of school's wealth of co-founders. Look at the people around you and ask which you'd like to work with—you may prefer the quiet guy you've ignored to someone impressive with an attitude to match. Only start a startup with someone you like, because it will stress-test the friendship.

366–370

Another thing is to learn startup-useful skills, which differ from job skills. Thinking about a job makes you learn languages employers want, like Java and C++. But in a startup you pick the language, so you choose what lets you get the most done—maybe Ruby or Python.

371–383

But the most important skill isn't a programming technique. It's a knack for understanding users and giving them what they want. Get into the habit of thinking of software as having users: what would make them say wow? Users are missing from most college programming classes—taught like writing as grammar, without the audience. But an audience is now only an http request away, so why not build a website people will find useful? At best it becomes the startup itself, as it did for Yahoo and Google.

333–383

To keep your options open now: learn how startups work by sneaking into one, make the most of school's wealth of co-founders, and learn startup-useful skills—above all the knack for understanding users and giving them what they want.

385

[1] Even the desire to protect one's children seems weaker, judging from things people have historically done to their kids rather than risk their community's disapproval. (I assume we still do things that will be regarded in the future as barbaric, but historical abuses are easier for us to see.)

386

[2] Worrying that Y Combinator makes founders move for 3 months also suggests one underestimates how hard it is to start a startup. You're going to have to put up with much greater inconveniences than that.

387

[3] Most employee agreements say that any idea relating to the company's present or potential future business belongs to them. Often as not the second clause could include any possible startup, and anyone doing due diligence for an investor or acquirer will assume the worst.

388

To be safe either (a) don't use code written while you were still employed in your previous job, or (b) get your employer to renounce, in writing, any claim to the code you write for your side project.

389

Many will consent to (b) rather than lose a prized employee.

390

The downside is that you'll have to tell them exactly what your project does.

391

[4] Geshke and Warnock only founded Adobe because Xerox ignored them. If Xerox had used what they built, they would probably never have left PARC.

392

Thanks to Jessica Livingston and Robert Morris for reading drafts of this, and to Jeff Arnold and the SIPB for inviting me to speak.

393

Comment on this essay.

385

Even the desire to protect one's children seems weaker than peer approval, judging from things people have historically done to their kids rather than risk their community's disapproval.

386

Worrying that Y Combinator makes founders move for three months also suggests you underestimate how hard a startup is—you'll face far greater inconveniences than that.

387–390

Most employee agreements claim any idea relating to the company's present or future business—which could include almost any startup. To be safe, either don't use code written while employed, or get your employer to renounce, in writing, any claim to your side project.

391

Geschke and Warnock only founded Adobe because Xerox ignored them; had Xerox used what they built, they'd probably never have left PARC.

385–393

Footnotes: peer approval can outweigh even protecting one's children; moving three months is trivial next to a startup's inconveniences; employment IP clauses are dangerous; and Adobe exists only because Xerox ignored what its founders built.