Y Combinator's Paul Graham: We're Looking For People Like Us

Absolutely.

Ah, good morning, we, it's-it's-it's very good for me to be back after last year and, and what we want to do this morning is talk to two people who are both investors and they are both interested in other things. Paul, as you know, I am not, we couldn't really hear what Mike said so, so I'm just gonna add a couple of things here.

Number one, he was a guy who went to Cornell and got interested in programming and then got a PhD at Harward. Also had a kind of collateral interest in Art, took some courses at The Rhode Island School of Design while he was there. He had an interesting idea which we'll talk about as to, as to whether, if you, the difference between being a hacker and a painter and how you would become one and not the other.

We'll talk about that too. What we want to get out of this conversation, I think most of all is to see the world through his eyes and what is he looking for and what is he experiencingof looking at so many projects, how has it sharpened his vision and his experience in terms of identifying those people most likely to succeed?

You know, and what is the criteria, and what's the profile and those kinds of things? But let me begin with the most obvious question of all, what is YCombinator?

It 's a trick in the lambda calculus. Just kidding. It's a company that invests in a lot of start ups at once.

Yes.

And then.

What's a lot?

This summer sixty-two. so far. So we fund sixty start ups at a time, and for three months they come to Silicon Valley if they don't live there already. We work really closely with them and at the end we have an event called Demo Day.

Demo Day.

They all present your investment.

You have two of those a year?

Yeah.

All right. But let me back up a little bit. There's a famous story about you making a speech at Harvard. I guess the Harvard Business School.

No, actually it was to undergrads.

It was to undergrads.

Yeah.

Yeah. And tell us what the results were and what it did - said to you.

Well.

This was '95 or so.

Yeah, March '95. It was We would talk about how to start a start up to the undergrads at Harvard. And in the talk I said if you are trying to raise money, it is better to raise money from someone who made money themselves through a start up because then they would be able to give you advise. And I suddenly noticed they were all looking at me expectantly and I thought oh my god!

I have an idea. You have money. I had this terrifying vision that they would all send me their business plans. So I said but not me, right but then afterwards I felt guilty. So, I said, "Alright, alright, I should finally do some angel investing. So, we setup YCombinator. Initially, we were just going to do normal angel investing.

Invest asynchronously.

And so why did you call it Y Combinator?

Well the domain name was not taken. And it is a really cool mathematical concept actually.


Which is?

Well, you really want me to explain?

Is there not a simplified version.

It's a way to make a recursive function even when functions can't have names.

There you go. I got it completely. Yeah.

So when you look at it when you started it verses today, how has it changed?

It's a lot bigger. It's a lot bigger. It's sort of like the difference between seeing some like mom and pop restaurant and seeing like mom and Pop co, some kind of gigantic machine or something like that. It's got the same essential structure, but it's just giant.

Is it easier to do a start-up today than it was when you started?

Yeah I think so.

Much easier?

There's a lot more knowledge out there, and it's a lot easier to raise money.

Why is that?

Partly because more people want to invest in start-ups, partly because the procedure for raising money is better understood, like the two parties, the protocol for the founders and investors to talk to one another, it's smoother and more established. There's less misconnections.

It's also cheaper, yeah.

It's cheaper to do a start-up, yes it's much cheaper.

It's essentially-essentially free, right, because it used to be that what cost you money were, you had to pay money for computers, well everybody's got a computer now, for wasting their time now with video games.

A fast computer, now.

Yeah, and like you need a powerful internet connection, but everybody's already got a big internet action for pirating movies.

So where does angel investing end and venture capital begin?

A million dollars.

About a million dollars.

Yeah.

So now people now want your, they want your ou know, they come to you and they say "I've got an idea". You seem to say that the founder is more important than the idea?

When people come to me and they say, "I've got an idea", I say, "Tell me about your co-founders."

Is that right?

Oh yeah, yeah. I don't care about the idea! I mean, I care a little bit. If it's-if it's really great, I might be excited if it's really terrible, that might be a bad reflection on the guy, but what I care about is the founders.

And the most important thing you want to know from founders is-is how they plan to make money?

No, I want to know what kind of people they are.

Oh.

Right, there are-there are some people who just get what they want in the world.

Yeah.

Isaac Stern, a great pianist and great violin player once said to me that he could always tell who the prodigies were gonna to be because they would walk in front of him to do, just an exhibition,  and they would basically be  saying to him if they were ten years old "Sit down and listen to me, I have something to play".

Yeah, well, I would not want a bit of that.

I would not want to encourage that too much in the people who come to my company here. I mean were already sitting down and listening.

But-but you did layout in Fortune magazine, in Forbes magazine, 5 things. Let's go through those.

Well I hope you remember what the word because I do not remember.

I do remember all of them determination. You can spot that, if they don't have that.

Well I wish I could spot it. Actually the hardest thing about doing Y combinator interviews, which are only ten minutes long, we decided whether or not to fund people.

In ten minutes?

Yes, yes, so it's hard to tell if they are determined in ten minutes. People often fool us. I mean if they didn't, every startup we funded would succeed. Right? And that's certainly not the case. So we can be fooled about determination. You can usually tell how smart somebody is in ten minutes. You know, you hit a few balls across and see how hard they come back.

What did they have to explain?

But people can put on an act, about determination, for ten minutes a lot of the time.

How about the next one was flexibility.

Yeah.

What does that mean?

Well, that we can test, actually.

Yeah.

Sort of mental flexibility. So, people come in and they say, "We want to do such and such idea" and then we'll say, "Oh have you considered the possibility of like rotating at ninety degrees and doing that instead".

And how they respond to that?

Right, some people will take it up and say, "Yeah, you know and we could also do such and such, and such and such" and other people will just like, "Oh no, actually we wanted to do the other thing".

And that shows inflexibility and they're lesslike fib, that is not good. But, people are often nervous in their interviews, so we have to, the thing is and this is a lot to put in ten minutes. The third thing is imagination. Yeah, very similar kind of conversation to the one about flexibility. We try to say, "Have you tried this, have you tried that?".

"What do you think about such they are not only be able to grasp what we are saying but will take it further. And, other people just do their best to answer the question.

When you say "we," what do you mean? Well, there are six YCombinator partners. Yeah, so when you come in to be interviewed a YCombinator, there is going to be so many of this. It is just like a court martial. Rather intimidating.

Yeah, we do not mean it to be intimidating, but the ten minute format and the numerical disparity tend to make it that way. This is the one I did not really understand, naughtiness. Naughtiness, well start-ups often have to do with slightly dubious things.Right .

So, like you know, you can kind of tell if people have a gleam in their eye.

Yeah, right.

We don't want the kind of people who would be obedient employees.

Yeah.

You want people who understand and are willing to do, that sort of, that kind of stuff held together by duct tape. You have to do to get things done.

That is almost an entrepreneurial quality ...by definition.

You can't, we're not looking for people who are sort of, did what they were told, in life.

But team building is important.

Well ...

Or not?

They don't, you know, it's not team building so much, I mean, you just want founders who are already friends, you want founders who have a relationship.

Which is the last one, friendship.

Yeah.

Was is?

There you go.

Yes indeed.

So let me just talk. One of your big success stories is Sam Altman, one of them.

Yeah.

Give us... who were those that... tell us what you saw, in those that you have recently invested in, that make a... difference to you, what was it about them?

Well Sam in particular, when I met Sam he was 19 going on 40.

Right.

And I remember..

Going on forty.

Right, right. It's.. seems like, it's weird, you know, you don't realize this until you talk to these, You can kind of guess what somebody's age is based on talking to them right. Even though you couldn't see them whether they are young or old. Old people won't take as much shit but maybe they're not as mentally flexible either, right?

Sam, when I talked to him he pushed back like he was forty. I couldn't treat him like a nineteen year old. He was like "Screw you dude, I'm forty inside. Never mind what I look like".

He was in Stanford then or not?

Yeah, he was an undergrad. He was a sophomore. I tried to reject him actually. I sent him an email saying look, you know, what's the rush stay in college? You're only a sophomore. Just do this later and he said you know, actually, I insulted him. I said he was a freshman and he sent me a reply back saying I'm a sophomore and I'm coming, you know, and he was really good.

I'm glad you did!

In all the list of things you cited, friendship, determination, flexibility, imagination and naughtiness, intelligence is not there.

Yeah.

And so your suggesting that, what? By not putting intelligence as one of the things your trying to look at.

Well, partly that, it 's somewhat taken for granted. If you're a programer you're not probably at least above average intelligence. But also, it really isn't as important. When we first started, we thought that was going to be the key. Everyone in the startup likes to believe that's what makes startup succeed, But it isn't.

There are plenty of smart people who get no where.

Each one is not a lot of money.

You mean each start up Angel investing yeah.

Yeah because we, our goal is only to invest enough money in them to cover their expenses until demo day, when they can raise some more.

So they can live until demo day.

It's just first year.

And What's the relationship you develop?

Because the promise for them is that they get, a, imprimatur because you've invested in them, they also get a sense of opportunity to meet people, there is a connection, to get a part of a community.

It had better not be just those two things though, because imprimaturs live and die pretty quickly. Brands. Right. Introducing people, you know.

So, what is it?

You help them understand the Silicon Valley culture.

I would like to believe we actually give them good advice.

About their business.

Yeah. Yeah.

But tell me about the t-shirts that you have.

We have t-shirts that says on the front, "Make something people want".

So, what do you mean by that?

Well, because the the biggest mistake that start-up founders make is to be disengaged from their users, right?

Yeah.

The thing I tell people above all, is go tell your users, you know, go talk to your users. Find out what your users want, because the amount of value you create is sort of this rectangle where the number of users on one axis.

Yeah.

And how much you improve their lives on the other axis, right? So ideally, you want to be something like Google where it improves a lot of people's lives a lot, right? But the area of that rectangle is the value you create and ultimately your net-worth.

Yeah. But all of them that you've funded have that the drive to make something that people wanted.

No, we thought all of them that we funded, had that drive after interviewing them for ten minutes.

So, what's your percentage of success?

You know, it's hard to tell ya. It takes years!

You used to be said like one in every ten investments of a venture capitalist will be hugely successful, a few will be moderately successful, and the last five will fail.

You know our definitions of success are lower too. A venture capitalist: if a start-up gets acquired for 30 or 50 million dollars, that's a failure to them. It's certainly not to us. And not the founders either, if they're not very diluted, that's enough, definitely enough to make them rich.

Take off for us, where the investments are, and why you thought these were good ideas, and what made a difference? Who is in the portfolio?

Well, there's are 313 companies.

Well, give me the best ten.

The best ten. Oh my God. Oh Drop Box.

Drop Box.

Drop Box is very popular, right. Airbnb Right, Airbnb. being Loopt, Heroku which just got acquired, Scribd, Greplin, Xobni, Justin.tv,
I don't know if I can name ten or not.

O.k. but we can find it.

They all share these qualities of the founder, number one.

The founders of those companies are all driven.

Yeah.

But do they share something else about where we are in terms of this moment in the evolving history of technology and the web?

A lot of them were building something that they themselves wanted.

Right.

A product that they wanted or service they wanted.

Right, like Drew made Dropbox first for himself and even the founders of Airbnb, they were their first users too, right.

Sam Altman was, built Loopt for himself, the original incarnation of Loopt.

It's so important to understand your users that, if you build something for which you you are yourself the user, then of course you understand them. You have like, the recipient for the ideas and the creator of the ideas in one head.

How do you keep up?

Practice and software. YCombinator includes a lot of software for keeping track of everything.

Yeah And today The relationship you've maintained with these investments is what?

It depends what they want. Right? Like Sam Hoffman I still talk to regularly and reinvested in them 6 years ago.

And they come back when they are successful and provide C capital to be invested through you?

Sometimes, sometimes they come back and do YCombinator again.

Is that right?

Yeah!

For another deal?

Yeah! This-this summer, we have nine alumni coming back.

Yeah.

When you at the point I made earlier about. Tell me this story, when you, somebody once asked you, I think it is on your website about questions I might have asked or something like that, which was, whether a hacker could become a painter, or a painter could become a hacker, remember that?

Yeah, vaguely, yeahBasically, you said that and I thought it was fascinating, is that you have to choose something that's mysterious to you and if you have to do it when you're really young, Oh, because you're too stupid to realize how hard it is.

Exactly right.

Do most people have that quality when they come to you;
they are too stupid to realize how hard it is and all they want to do who has worked as harder as I can.

It's not all of them.

Not all of them because they're not all.

Yeah.

That's one of the misconceptions about YCombinator,
the founders are all 22. We funded-

Because part of the conventional wisdom was anything over 30.

I don't know yet.

I don't know.

A lot of our goal, YCombinator are actually one of our secrets is that our goal is not only to make money, it's also to get more data about what works.

Yeah.

Originally that was our entire goal when we first started.

The only reason we started funding a bunch of start-ups at once was to teach ourselves how to be angel investors.

It was a complete-

So we'll learn from those you've invested in. Yeah! This new business model is like funding a bunch of startups all at the same time. That was a complete accidental discovery.

Let me move to broader thing. Do you worry about a tech bubble?

I definitely worry prices are high.

Yeah.

I'm always reluctant to use the word bubble. Because I was there for the last bubble. And, like, this is, things are not crazy like they were then.

Yeah.

I warned people, like, prices are high. Raise money now. Things might not be so great.

That was a great tip. Well tell us more about that. You're warning people what?

Raise money now.

Get the money now.

Yeah? Because like, who knows?

Yeah. Because it may not be there when the bubble bursts.

Tomorrow, the Chinese economy collapses and all the money goes away again. I mean, things have these ups and down, and it's an up now, no question.

Any investments you have in Chinese start-ups?

Well, there might be some. I'm not sure, there might be some Chinese ones this summer.

Yeah.

I'm not sure.

Oh, they're coming here now?

Yeah.

To Demo Day.

Yeah, I'm not sure. I'm not absolutely sure. I mean, we only, some of these people this summer, I haven't talked to since we interviewed them.

How many application do come to Demo Day?

Well, they don't apply to Demo Day, they apply to YCombinator.

Indeed.
A bit over 2,000?

a 2,000?, yeah The other question I'm fascinated more, because of all the things happening in social media, and we'll talk later about that Where do you see the opportunities in social media now?

Boy. Where is the. People always want to know, what are the good startup idea is? and what are they about? You know, I get told them every six months by the Founders. So, I've learned not to try and predict, Right? Because every future sprint me in the face. But they tell you one by one based on how successful they are Rather than having some big idea about where the future is.

Yeah. I don't know, I'm sorry, I never have good answers for this, this kind of question.

When you think about where YCombinator is going, yeah, is it simply going to get bigger so rather than investing a billion dollars. be investing in five more years, five billion or?

But we don't investing if you like a billion dollars. But the total investment of the value of the companies you've invested in? well, the bigger the better.

Yeah.

Right? I mean we get bigger. We don't try to get bigger. It's just that the number of good start-ups that apply keeps going up and we have to grow to accommodate it. It's like this plant. We sort of plant it by accident really, and it just like naturally grows, you know? So, it's people. If the number of applications keep going up, then we have to keep growing to accommodate them.

You once wrote a piece, maybe, forgotten where, said Microsoft is dead, remember that?

Yeah. I got in a lot of trouble for that at time, but boy that was spot on. That's exactly the kind of things people are now saying as a matter of course in new stories.

And you said this in what, nineteen, when was it, two thousand or something? Yeah.

Two thousand something!

Yeah, two thousand something. Yeah. So what do you think of the Skype purchase?

I don't know what, I don't know.

But what does it say? It says Marc Andreesen and Ben Horowitz are pretty smart guys.
I don't know what they are planning to do with it.

You don't?

No.

Nothing that's come out of Redmond has said anything to you that made you say ah ha, I get it.

Well to be honest.

It's not a Hail Mary for them it's. I don't know what their doing. I don't know. There like M;A guys have a taste for big deals or something, I mean, it's M;A guys have this perverse incentive. It's more prestigious as an M;A guy to buy a company for a lot than a little, right? Really, they should be trying to buy things as cheaply as possible.

They're buyers, right? The should want the price to be low.

Back to the whole bubble thing, when do you think Facebook will have an IPO?

I don't know. Isn't it 2012 or omething like that? I don't know, you probably know as well as I do.

And what's going to happen to Twitter?

I don't know about that either. Like Jack Dorsey is trying to be the CEO both Twitter and Square. It's been historically pretty rare for someone to pull that off. The only one who did anything like that was Steve Jobs.

Yeah.

But those Pixar and Apple are much later stage companies.

Yeah.

So, it 's sort of an unprecedented situation. It would be interesting to watch.

And Cisco?

Cisco? I don't know! IBM.

I'm just taking. Exactly IBM. Who else is dead or something?

No, no, no. But these are like these gigantic companies.

Well, yeah. I know. And you only think of small ones.

Yeah, I'm like down in the trenches with these little. Once a company has 30, 40 employees, it seems like.

It's too big for you.

An international conglomerate to me.

Yeah. When do you cash out?

When the company, when the founders do.

Is that right? When the founders.

Yeah!

When they sell you sell?

Yeah. Well, we couldn't sell 'til they do. Right?

Well, of course. Is angel investing growing faster than venture capital?

I think so, yes. I think so. I think venture capital as a business is actually shrinking. There's been a big.

Start early on it.

Big garbage collection of funds. All the bad ones are dying.

Have you done anything about your interest in painting?

Not really, no. In fact, I haven't really thought about it much lately. It's kind of a bummer.

It is two different parts of my head.

Does all the attention, the cover of magazines, Forbes, hings like that. Does that make it easier or more difficult?

Well, it probably makes more people apply to YCombinator. But that might make it more difficult because it's more applications to read, and if it's an application we only got cause we were in some magazine, then it is probably disproportionately unlikely to be one we'd accept.

But let's assume you wanted more applications, you wanted to see more. would you generate more ideas? More applicants.

Oh, it's totally, it would be totally easy.

I would encourage more single founders to apply.

If I started saying "we love single founders!" instead of saying we discriminate against and you'll never get in unless you're super good.

Yeah.

We'll get ten thousand applications from single founders.


And it's, so it's the so of all the applicants today, you'll choose how many?

About three percent.

Three percent.

Yeah, It's always been between two and a half to three and a half percent.

I think this current summer batch is spot on three percent.

And why do you think you're good at this?

Well, the companies seem to do well.

Well, yeah.

No, but that test how well the companies do.

No no that's the measurement, that's the measurement of why you do well.

Oh, oh you mean what makes us good at it Like when venture capitalists, venture capitalists are often like MBAs, right?

So when they're trying to tell us someone's a good founder, how do they judge?

But we just judge if they seem like us.

Is that right?

Yeah, we're all.

It's like, for us, it's like work all the time, meet a lot of, you know, if they have the same attitude of those qualities that we talked about.

We're just trying to recognize our peers.

Yes.Right ? It is so much easier. You could probably tell if somebody was going to be a good interviewer.

Well, maybe, yes.

Yeah, so we just tell if they seem to be good versions of us.

Well, then why aren't there more people like you then?

You know, I think, it did not used to be cool to be an investor, right? It used to be that investors were like gallery owners, like in the founders more than artists.

Right?

And investors were just like these guys who like ran the infrastructure and were not cool. But now angel investing is getting a little cooler.

How many, how many guys, men or women, could come back to you and say, I told you I was going to be good, and you turned me down.

See, it does happen, we keep track of them actually, in fact when we send out the emails, you know rejecting people, we say tell us if you succeed so we can fix our process.

Oh, then have you learned anything from that?

Oh,yes,yes,yes.

What?

Well, I mean it is always very specific. But we look it, I mean one of the advantages of having so many competitors, is that if we miss somebody, they will fund them. so when one of our competitors has a successful start-up, we go back and look at their application to YCombinator and try to figure out why we didn't accept them.

Right?

That's a pretty smart rule, period. Always be asking yourself, why are we screwing up, and how am I getting better? How do I get better?

So, one of the biggest ones that we missed, was, well I can't mention their name, because then I would be saying rejected them which would be sort of insulting to talk about.

I see. If they were successful I don't think they would worry. They probably carry it as a badge of pride, if in fact they came to you, you turned them down and they went on to be successful.

Yeah but I feel like its hole, anyways, we went back and looked at their application and we have a bunch of us reading the applications, one of our partners is Robert Morris, the professor at MIT, and he's notorious for giving the worst grades on the applications. And we went back and looked like Robert had given them a C, which completely sunk them in the ranking so deeply that no one else has read the application...

giving, so now everyone he gives a bad grade to, one of us checks as well.

Alright, I'm out of time here, two quick questions .

The total value of Y Combinator companies is three billions, not one billion.

Oh yeah, yeah.


Definitely more than one.

And the total value of the investment from you guys is what, 350 million or so?

What, we've invested?


Nothing like it! More like, five.

Five .

Are you listening?

The value that comes to is 3 billion you invested 5 billion and you generally.

For eighteen thousand times three hundred and thirty.

And you generally take about what 6 to 10 percent.

No, it's around six or seven years.

Six or seven?

I'm stretching it by going to ten.

This has been great fun.

Thank you.

For me too.

( speaker name) thank you.

Thank you.

Staying?

I 'm going to sit down with you for a second.

O.k.