Startup School Transcript October 16 2010 Paul Graham - Partner, Y Combinator; Founder, Viaweb Link to video: http://www.justin.tv/startupschool/b/272180509
[applause] Paul Graham: Hey guys! Thank you, thank you, Andy! That was a really good talk. I wasn’t in the audience, I was walking back and forth behind that curtain there, listening to it. It was very distracting, I don’t even remember what I’m going to talk about [laughter] because I couldn’t concentrate on my own talk, because yours was so much more interesting.
What I’m going to talk about today is what is happening in the world of startup funding, which is a big percentage of what a lot of founders spend their time thinking about. That world, quite surprisingly, like, since the last Startup School, that world has just been turned upside down. And as the next door neighbor of the later stage funding guys, I’ve been sort of sitting there on the sidelines, watching this giant battle happen, and trying to figure out who is going to win, and what that’s going to mean for startups. I don’t actually have a prediction yet about who is going to win, but I can tell you what I’ve been seeing so far, and tell you what I’ve been suggesting to founders to do at the moment.
So, after barely changing at all for decades, the startup - oh, and by the way, you don’t have to take any notes, because I wrote this all out in advance, like I sometimes do, and it’s going to be online later, so, and your notes won’t possibly be as well written as my version [laughter], so you can just, you don’t have to type anything. Maybe if we get to Q&A at the end, that would be last, otherwise.
OK, so, at Y Combinator we have observed dramatic changes in the funding environment for startups after demo day, even before demo day. Fortunately for you guys, almost all the changes are for the better, and one of the most notable is super high valuations, so high that there have been reports of angels, investors, meeting in secret [laughter], if inviting Michael Arrington counts as secret [laughter], to grumble about them in sync.
So, I think, though, that the trends we’ve been seeing are not YC specific. I think it’s just in the nature of things, because the startups we fund are so plugged into the value and there’s so many of them, we tend to see trends first. But I think what we’re seeing, everybody is going to be seeing, so I’m going to explain what we’re seeing and what that means.
Let me start by explaining what the startup funding business used to look like, since even that may be news to some of you. There used to be two sharply differentiated types of investors, there were angel investors, who were individual rich guys, like Andy [laughter], who would invest comparatively small - i.e. small by their standards, but not yours - amounts of money in startups. And there were VCs, guys who worked for companies who invested larger, much larger, like million dollar amounts of other people’s money in startups. And it used to be a giant gap between these two. In fact, it was a very inconvenient gap for startups because most of the startups we funded after demo day wanted to raise around three or four hundred thousand dollars, and there was nobody you could raise that amount from conveniently. The angels tended to invest 20 to 50 thousand dollars, right, and if you wanted to get a round of three or four hundred thousand dollars to get you to stitch together all these little investments. It was sort of, like, making a coat out of mouse skins. On the other hand, the VCs were just not interested in investments under a million, so I used to have to tell founders how to let go and talk to investors and pretend they actually needed to raise a million dollars or more than that, when in fact they had nothing they could think of that could possibly cost that much. Well, in fact, they would find things later.
So, now, right smacked ((out)) in the middle, in the middle of the no man’s land, there’s this new thing, boom, the super angels. Super angel is a little bit of a misleading term, some of the super angels themselves don’t like being called super angels. I don’t know who invented it, I don’t know if it will stick. Some of the super angels prefer to be called mini VCs [laughter]. Yeah. That doesn’t sound so great either [laughter], come to think of it. I don’t know, we’ll call them whatever.
The super angels have changed everything because they’re right in the middle of the gap between the angels and VCs. The super angels retain some of the characteristics of the angels, and some of the characteristics of VCs. They’re usually individuals, like the angels were, and in fact, most of the super angels were regular angels investing their own money before they became super angels. What makes you super, apparently, is having limited partners. The thing that super angels share in common with VCs is that they’re investing other people’s money, like the VCs. This makes the VCs particularly afraid of them because they’re not just competing for deals with the VCs, they’re also competing for investors, right. And the VCs are afraid that their LPs, the people who give them the money that they invest, pension funds, and endowments, and so on, will decide that classic VC funds are obsolete, and start giving their money to super angels, or mini VCs, to invest instead. And this is a reasonable worry because this is what a lot of the super angels think too, right, that VCs are obsolete, and they are the replacement.
So, are they the replacement? [looks at paper notes] Where in the hell am I talking? Boy, I just ad-libbed a lot there [laughter]. Alright. Ah! So, will they be the replacement? Well, in most businesses they would be, if this were an ordinary product business, if something came along, a sort of faster moving, light weight version of VC, that would be the new VC, right. The thing that may help the VCs, the thing that may allow them to fend after the super angels, to some extent, is the extremely uneven distribution of returns in the startup funding business. Almost all the returns are concentrated on a few big winners. This is something I didn’t understand until I became their next door neighbor and did the math, but basically, the expected value of a startup is the percentage chance that it’s Google [laughter]. So, to the extent that winning depends on absolute returns, the VCs could still beat the super angels if they ha-, if they just get all the big winners, right. The super angels could win almost all the battles, and yet if they miss those few big winners, they could lose the war. Is that going to happen? Probably not, but it’s an interesting scenario to consider.
Because the super angels do a lot more investments per partner, probably five to ten times as many, you get less partner for investment, so the super angels won’t be able to do as much for you as the VCs could. Now, till now, how much is that extra attention worth, how much is, like, the real hands-on attention that VCs give you worth? Well, in some cases it’s worth a negative amount, in some cases it’s worth a positive amount. There is no consensus yet in the general case about how much the extra help the VCs give you is worth. The interesting thing is, now that the super angels have come along, competition is going to put a market price on how much extra help the VCs give you, because you can now get VC sized amounts without the extra help by putting together, you know, this angel round at a giant super angel investments. We’ll see, we’ll see what the market price of VCs’ help is. In some cases, it could be substantial. Until now the claims by VCs about all the value they add, have been sort of, like, claims of that type by the government. They make you feel better, perhaps, but there’s not much you can do about it if you need money on the scale only they can supply. But that’s changing, now we’re going to find out. I think what’s going to happen is the value added by VC funds will d-, very enormously from partner to partner, and so we may have something start to happen in the VC world, like it’s been happening in the world of journalism, where it’s the individuals who start to have brand, and not the companies they work for. At that point why not quit the VC fund and raise your own money and become a super angel, right? It’s so unpredictable what’s going to happen.
One of the big differences between VCs and angels is the amount of your company that they want. VCs, if they can get it, want a third of your company, and I’ll explain why. There’s a critical constraint in the VCs’ model, which is they take a board seat in your company, and partners and VC funds can only be on so many boards at once before they get overloaded. On the order of ten, right. And board seats tend to last about five years, so that means a VC fund can only do about two classic series A deals per partner, per year. And that, that is like the single most important thing to understand about the structure of the VC business, that explains everything. That explains why they take so long to make up their minds, because they can only do two deals a year, right, they’re going to be really careful. And it also explains why they want a large percentage of your company. These guys, a given partner, only has ten board seats at any given time, and he is not going to want to use ((of)) one of his ten board seats for only a small percentage of your company. So, VCs want a giant slug of your company.
The super angels have preserved the very critical quality of angels of not taking board seats, which means they can buy a small percentage of your company, they don’t need a giant slug of it. And that, at least, is good news for founders because, if you talk to most founders raising series A investments now, most of them feel like they’re having to take, they’re having to give up too much of their company at that point. And maybe they’re getting more money than they wanted, as well, so they’re selling too much of their company for too much money, right. Most of them, if you asked them, would rather take half as much money as they raised in return for half as much stock, use that money to make the value of that other half of the stock more valuable, and see what kind of price they can get for that later. That was never an option, the VC said “Take it or leave it, here is the structure of our deal.” And VC funds competed amongst one another, but they never competed by, like, being willing to take less of the company, just by offering higher valuations.
Well, the super angels, because they don’t have the constraint of taking board seats, they can take a small percentage of your company. You can now decide, like, exactly the percentage you want to sell in return for exactly how much money. So that is a big change, and we found probably many, maybe most, of the startups coming out of Y Combinator, that could have raised series A rounds in the past, are now raising large angel rounds of maybe 800.000 to 1.200.000, with mostly super angels as investors, and regular angels, and interestingly enough, VCs, but I’ll get to that later.
The advantage of raising an angel round over a VC round is not just that it’s, it doesn’t involve, you can sort of adjust the amount of dilution. Also, another big advantage is that it’s quicker, and most importantly, you get feedback as it’s progressing. If you’re raising an angel round, especially now that convertible notes have become the norm, you go investor by investor and get money from, and so you start with the committed ones, they nice guys who say “Yes, I’m absolutely in.” and work your way outwards towards the assholes who give you lines like “Come back to me to fill out the round.” right, which is like “I’ll invest if lots of other people want to.” You work your way out, and at each point you kind of know how you’re doing, the round isn’t this all or nothing thing, whereas when you raise a series A round from a VC fund, you have to go through a whole series of meetings, usually taking weeks, if not months, although the VCs have been getting a lot faster lately. And at the end you have this meeting, usually on a Monday, a full partner meeting, where you finally present to all the partners, and they say yes or no. Well, the scary thing for founders is they could still say no at the end of this whole process, right. An angel round you’re sort of raising it piecemeal as you go, you know how it’s going. A VC series A round, you get to that final partner meeting and the on average across firms, there’s a 25% chance at the end of that you’re going to get nothing. At some firms it’s as high as 50%. So, that’s kind of stressful, right. And so a lot of startups are starting to raise these rounds where they stitch together hundred thousand dollars investments and end up raising, you know, a total of six or seven hundred thousand dollars.
However, the VCs have a weapon that they can use against the super angels, and they have started to use it, the VCs who started to make angel sized investments themselves. Just in the last year there have been an immense number of VCs who have all come to me as if they were like the first person to tell me this thing “We are now making hundred thousand dollar investment, you know, pass it on.” [laughter]. So they are, I’m here to tell you. So, when startups raise these big angel rounds, I mean, they’re angel rounds in structure, but they have VCs in them, so we’ve seen deals where there were like four or five different VC funds in an angel round, alongside the super angels, investing, you know, hundred thousand dollars a piece.
When VCs are in angel rounds they can do things that the super angels don’t like. VCs are price insensitive in angel rounds, partly because they’re price insensitive in general, they want to put lots of money to work, and partly because they don’t care about the angel round still. They think of a way to recruit you, to get a series A round later. This is a way to get close to you. So, they don’t care about the valuation, so they can blow up the valuations for the angels and super angels in the angel round. Which is great for you, right, you can play these parties off against one another and get these super high valuations. And I think that’s a lot of the reason that valuations out of demo day were so enormously high. I was thinking about the valuations out of demo day this year, and thinking “Geez, are we having another bubble?” And after thinking about it, I think not, fortunately, it’s just this weird dynamic in the funding business that’s causing valuations to be so high, and it should work for you guys too, if you want to go raise money.
So, the super angels definitely care about the valuations, they’re pissed when the valuations get high, and that actually mystified me. A bunch of super angels sort of bragged, apparently, about turning down startups who they’d funded after demo day because the valuation was too high. And I’m thinking to myself, now, wait a minute! Don’t these guys know, like, rule number one of the startup funding business, which is the fact I mentioned earlier, that all the returns are concentrated on a few big winners? Don’t they know that they can, at most, get 2x better returns by getting a lower price, whereas they might get 100x better returns by picking the right startups? You know, how could they be so stupid? Well, these guys aren’t stupid, so I thought more about this and I realized there’s an explanation why super angels care about valuations. The flip word. They, I believe, at least some of them, are hoping to invest in companies that get bought quickly, that is the only rational explanation for why they would care about valuation. If you’re looking for the next Google, you don’t care if you invest in a valuation of 20 million, right, it’s going to be worth 200 billion. 20 million is not an expensive valuation. But if you’re looking for a company that’s going to get bought for 30 million, 20 million is a really high valuation, you’re only going to get 1.5x, you might as well buy Apple stock, and just skip this whole startup thing [laughter].
So why do the super angels, why are the super angels trying to bu-, invest in companies that will get bought quickly? Well, because if the value, depending on the value of quickly, if it’s quick enough, it could actually be very profitable. And I’ll tell you, I didn’t even real-, even if I’ve been, like, in this business, watching these guys close up for months, years, I didn’t really figure this out until I wrote this talk. If you-, if they can get, you know, if they can get a 10x return, just 10x, but in one year, that’s actually amazing performance, because what matters in investing is not the multiple you get, it’s the multiple you get per year, the rate of return, as the call it in the business. And if you can get 10x in one year, that’s like, that’s a way better return than you could ever get with the old model of investing in a company that you hope will go public, you know, in the amount of time that it takes to go public. Think about it, if you invest in a company that ends up going public, but it takes six years to do it, which is pretty typical, then to match the returns of that super angel who gets 10x in one year, you have to get 10^6x, you have to get one million x, and like, Google, like, the big winner of all times did not come within two orders of magnitude of that.
So, buying companies that are going to get bought a year later at a 10x higher valuation, if it works out, could actually be a big win for the super angels. And I think that’s what they’re thinking too. If they haven’t thought about it, they should definitely think about it, because it could be very good for them. And I think that’s why they care about valuations. And if my theory is, and, like, no one ever admits that they want to sell companies that they invest in, so I can’t ask them this, but either they’re stupid or this is the explanation of what they’re doing [laughter], so let’s give them the benefit of the doubt. But what that means for you is, if you’re dealing with super angels, it’s going to be significantly different from VCs because they’re going to be tougher about valuations, but much more accommodating if you want to sell early. Like, for VCs, still, selling a company for $30 million is considered a failure, right, a grim failure. If you do the math of VC funds, the average exit for the companies they fund in a series A round I believe has to be around 80 to 90 million for their numbers to work out right to generate the returns they need for their LPs. So, 30 million is a disaster for a VC. That usually means that they gave up on you.
OK. Who’s going to win? The super angels or the VCs? I think the answer to that is some of each. They’re each going to become more like the other, so the super angels are going to start investing larger amounts, the VCs are going to figure out ways to invest more smaller amounts faster, they’ll either be on-, they’ll figure out ways to be on more boards, or they’ll figure out ways to do investments where they don’t have to take board seats. I think ten years from now the VCs and super angels will be hard to distinguish from one another, and there will probably be survivors from each part.
What does that mean for you? Well, one thing it means is that the high valuations that the startups are currently getting by effectively playing off VCs and super angels against one another may not last. If the VCs become like super angels, and that the economic model that I just mentioned actually works, then VCs are going to start to get miserly about valuations too, right. The good news is, for the first time ever, the slowness of VCs in now working in your favor. This change will only happen at the rate the VC business evolves, which in glacial [laughter]. OK.
The short term forecast is more competition between investors, which is good news for you. The super angels will try to undermine the VCs by moving faster, and the VCs will try to undermine the super angels by blowing up the valuation, which for you produces the perfect combination: rounds that close quickly at high valuations [laughter]. What could be better? But, remember, to get that combination your startup will have to appeal to both groups. It’ll have to seem acquirable, and it will also have to seem that it could one day go public. Most companies that seem they could one day go public also seem acquirable, so that’s not a problem. In fact, we have Andrew Mason speaking later [laughter]. ((More jokes)) there [laughter]. But it’s not the other way around, definitely. So, remember, if you want to play this trick of getting high valuations by appealing to VCs too, you have to seem like you could go public. I can give you advice about that. There is a danger… [looks at his wrist watch] There-, Wow! [looks towards Jessica Livingston]
Jessica Livingston: You got like five to ten minutes.
Paul Graham: Oh, I do? OK. There is a danger of having VCs in your angel rounds, so-called signaling risk. Signaling risk means, because everybody knows the VCs are doing it in the hope of investing more in you later, if they don’t invest in you, that sends the signal that you suck. So, how much of a worry is that? The answer is it depends on how far along you’re going to be after you’ve spent that money. If after taking this investment you’re going to get-, you’re going to be at the point where you get this graph of users or money that goes like this [hand moving upwards], no one is going to worry about what signal your investors are sending by investing more or not, right, because your own results are going to speak louder than that. Whereas if at the end of spending this money you’re still going to be a bunch of smart guys with a really promising idea, if the investors don’t invest more in the next round, maybe that’s a bad signal, right, but my gut feel is that this is-, that this whole signaling risk thing it just smells like one of those problems that founders worry about that’s not a real problem.
What I found is like a good startup, the only thing that can kill a good startup is itself. Like, competitors don’t kill you, you kill yourself, and I think this is going to be true for, like, investors, too. Investors can’t kill you, if you’re good. So this whole signaling risk thing, by default, I wouldn’t worry about it. We’ll see though, it’s all new. One thing that startups we funded have been doing to mitigate this risk though, is to take not too much from any given VC, so it doesn’t look to people who are looking at the cap table later, who’s invested in you, it doesn’t look like there’s any one VC who’s sort of de facto lead, right, and everyone would be looking to them for the signal. So, you could do that, it probably doesn’t matter that much.
The good news is more and more of you are going to be in this position to turn down money, if necessary, because, finally, after decades of competition that could best be described as intra-mural, the VC world is having real competition, the kind that comes from outsiders. And what that means is, the next couple of years are going to be a great time to raise money, hooray! So, the exciting thing about that is, not just for you individually, that you’re going to be able to have an easier time raising money, but also, just generally, there are going to be a lot more startups in the next couple of years. It’s going to be like the Internet bubble, but hopefully without the bursting at the end [laughter]. Alright, so, that is my talk. Do I have time for Q&A?
Jessica Livingston: ((You can maybe)) take two questions.
Paul Graham: Two questions! I only get two questions. Alright. Yes, yell it and I’ll repeat it.
Audience: So, I am the CEO of (( )) ((Santa Monica)). I was wondering what your thought about revenue royalty was.
Paul Graham: My thought about revenue royalty.
Audience: Yeah, you give a company like 100K and then, over the next year, you (( )) revenue and…
Paul Graham: Oh, oh, oh! Will companies start to pay dividends? I don’t think so because even Microsoft only barely started to pay dividends and they’re practically, well, they’re an established company [laughter] so, I can’t see it happening. There’s such structural obstacles to that, I doubt it. Peter!
Peter: Do you think that (( ))
Paul Graham: Will the fight between the VCs and angels lead to participating preferred re-emerging? No, because participating preferred is bad for founders, and the fight between investors is making founders able to get more of what they want. Yes.
Audience: So, do you think that’s sustainable for the super angels to rely on smaller, quicker exits for the return, or do you think it’s a temporary thing that’s being caused by a couple of larger companies that have a lot of cash and need a higher value?
Paul Graham: So will it be sustainable for the super angels to rely on quick exists? Nobody knows yet, that’s the fascinating thing. The super angels themselves don’t know, they’re thinking “Well, it looks like a good bet, let’s try it.” And the VCs are thinking “Geez, this thing could kill us, or maybe not.” But no one knows, no one knows, the next couple of years are going to show us. How much time do I have? Am I done?
Jessica Livingston: (( ))
Paul Graham: Alright, I’m done. Thank you very much, guys!
[applause]