Decision Analysis in Venture Capital

so thanks for the introduction Ron a couple things about my time with Ron so I was here as a doctoral student about 20 years ago and there are some fundamental frameworks and lessons that I learned from Ron that have been influential in how I practice venture capital and also frankly how he lived my life so one of those is decision analysis which you'll hear about in a moment another framework that has been just as important and just as influential for me in venture capital is this notion of telling the whole truth so this has really been a theme for me like in my doctoral dissertation I was very proud to have found a little bit of negotiation so my my dissertation was in the role and value of information in negotiation and there's when you're collecting information in negotiation there's a segment where there was I was able to create a framework where there's no incentive to misrepresent your beliefs so I was kind of proud of that and then even more fortunate to be able to write the book with Ron the ethics for the real world for Harvard Business School and when the what Ron mentioned in terms of that introduction so when I first got into venture capital there was a gentleman that was been inventor for about thirty years and he was kind of kept me and took me under his wing and he had about thirty years of experience and after we'd done a number of deals together he started introducing me as here's my colleague Clint he's a venture capitalist that tells the truth and after doing this a number of times I finally took him inside like David what does this say about the industry that that's a remarkable thing for you to mention about me when you introduced me and so so you'll see those two themes throughout this talk so one is the role of the decision analysis framework and also the role of truth if you will and truth in terms of not only what you tell other people but the stories that you tell yourself so when I first started in venture about nine years ago I went to a number of folks that I knew who had funded my prior companies and these are and also folks that I knew in venture capital with a question well how do you do it so I'm kind of new to the industry you know how do you become a good venture capitalist and almost everybody told me some version of you gotta look the entrepreneur on the I yeah I get a feel for the market some people have the magic some people don't I'm like really that that's all you got for me and like well you know I'm kind of thinking about using some data building a few models there's this thing called decision analysis that I think might be really helpful in the industry and they all said clang clang Clint you just don't get it we have a few analysts that run a few numbers and that sort of thing but nobody really pays any attention to them you gotta find the pattern so it's all about pattern matching all right well so I've looked for patterns and venture and turns out there's a lot of them so in venture capital pattern matching often means and this is a caricature but it means what does the good entrepreneur look like well it's a white guy who's dropped out of the computer science department at Stanford or Harvard that's the pot that's the model of a good entrepreneur and you know on the one hand it's a little bit of a joke but on the other hand 2% of venture capital dollars in the US last year went to women only teams 79 percent went to male only teams right so there's some of this pattern matching matching that's just off from my point of view now to give VCS a little bit of credit venture capitals of a really hard business on the decision-making side and here are a few of the challenges from my point of view the first one is uncertainty you know we know what to do with uncertainty but venture capital is extreme in the world of uncertainty just to give you a sense of how extreme there's a group called Cambridge associates that tracks all the venture capital investments and they've mapped that each year identify the top hundred investments in venture capital out of four thousand deals that are done those hundred investments essentially are where a hundred percent of the profits in the industry are so have you step back and think about it if a venture capitalist makes a decision makes an investment they haven't a two and a half percent chance of that being an interesting outcome now you would think with a ninety seven and a half percent failure rate there would be some humility in this industry strangely that's not the case so if you talk to some folks about VCS about okay they're bad decision well what happened and I've had a chance to do this inevitably you get some version of the story of I was just too early I read between the lines I'm just so farsighted in my view of how technology will impact society that I have to hold myself back to keep up with you know the rest of society there's no learning in that but this is the story that people tell themselves feels like how in the world can you get away with telling yourself that story well in the world of venture capital between the time you make your investment and you have an IPO is now nine years and if you're a typical VC you're making one to two bets a year in other words it's an incredibly small number of data points and huge time frames between when you make the decision and you see the outcome lots of room for your ego to show up and you can tell yourself whatever story you like now by the way you know when during an environment with lots of uncertainty and lots of ego these biases essentially don't change very fast they get locked in fortunately there are some things you can do about this and at other ventures we focus a lot on all three of these areas so for uncertainty we got decision analysis which you're all talking about you know in this class and in a lot of ways we look at decision analysis as a way to confront our ignorance with respect and dignity so think about that challenge for a moment so of the of the startup companies that get venture funding it's already a tiny minority that get venture funding two and a half percent of them are going to be interesting outcomes that's a really humbling place to be by the way I was an entrepreneur for a number of years and I sort of I sort of experienced the ninety seven and a half percent so now the question is okay when you're in this realm of such extreme uncertainty how do I have interactions and by the way Soumik who is going to be talking on Thursday as part of ulla ventures how do we have these interactions in a way that respects the entrepreneurs what they're trying to do in the face of all this uncertainty and and from our point of decision analysis is a way to bring dignity and respect into all this so it's not about by the way there's some of these seeds they want to hear the story I'm gonna have a million dollars in revenue at the end of year one and a hundred million dollars in revenue at the end of year three and I can guarantee that like nobody can guarantee that in this business and we're not asking our entrepreneurs to say those kinds of things that they don't believe in order to get our money now on the ego side I think the best way to confront that in our world is simply learning so what do I mean by learning it's like okay well I've got all this uncertainty how do I have a conversation with the entrepreneurs and the other folks that are experts in this in such a way that I could you come up with a good risk return view of the investment and so there's a lot of like the how Ron does his Socratic dialogue it's a lot like what it looks like with my conversations with entrepreneurs so there's certainly the pitch the entrepreneur saying here's my story and what not but a lot of this is really back and forth it's a mutual exploration of where the risk and where the value is biases are a really hard nut to crack in this industry and so hard not not just because of these things but actually the soap the standard social feedback loops and venture are all broken in most worlds if you're a jerk people tend not to want to work with you in venture you can be a jerk and people still come to you all the time wanting your money and so and by the way you know CEOs of big companies face this so so venture is not alone in having broken social feedback but I think I think venture is unique in the fact that they've got broken social feedback on top of all these other challenges and from my point of view the best way to confront that is just diverse points of view so so some of the diversity and hulu ventures for example my partner is my wife and she has a different point of view and let's just say I respect that and she's also Puerto Rican and she was four degrees from Stanford early employee at Google ran the legal department on the board of trustees at Stanford so she's got you know an incredible list of credentials and she shows up in this world not with a decision analytic point of view per se although she appreciates that but as William or one of culture and teens and how do you treat people and empathy with your customer and that sort of thing so a very different point of view and so maccer recently joined us also shows up with a very different point of view so he shows up with a point of view on what's your meaningful purpose and what are the values and the things that you most care about if you're an entrepreneur or you're a VC and how does that enter in this mix probably an appointee that's the best way to confront bias is to bring in a diverse set of points of view alright so that's a little bit of context in terms of I call it the decision-making struggles in venture now before you talk about individual investment decisions I want to give you a little bit of a picture of essentially the decisions that we face in this industry those are the decision hierarchy which you may have seen here in class and before we get to actually making investment decisions which are sort of what you see from venture capital in a public sense tip of the iceberg or bottom of the decision hierarchy first have to figure out your strategy which is ok out of the 500 venture capital venture capital firms in the US how are we going to show up in a way that's interesting and so here you can see is Allu ventures so we're seed stage meaning we're investing very early on enterprise IT as opposed to consumer fun-sized there were a sixty million dollar fund it's interesting it's kind of a decision it's kind of how lucky do you get on fundraising we have to do our own fundraising and you can see some of our other kind of how we think about our strategy now once we've got our strategy the question is how do you put together a portfolio a group of investments that's going to deliver on that strategy and I'm going to talk about that in the second half of my talk and I'm going to start here with what's most visible which is how do you decide whether to invest in a startup company or not so let's talk about investment decisions so here's the loulou investment process so the very first step when we see an entrepreneur and by the way we have about 2000 inbound / outbound interactions with entrepreneurs so if 2000 startups that we touch in one way or another a lot of these can be for just 30 seconds some of them can be for months and we're sorting qualitatively is this company a fit for our strategy so early-stage Enterprise IT silicon valley-based do we like the team is it a big market and we sort qualitatively sort out most of the companies here in step two we do what we call creating a market map so we little to get in a room with the entrepreneurs at a whiteboard and build a picture of the market opportunity so this is their target customer the business model adjacent markets competition how it changes over time and what we're trying to get there is it what makes you excited about this market what makes you worried about this market so get those on the table in steps three and four we quantify everything so you you guys know how all that works on the quantification side but essentially if somebody says this is a billion-dollar market opportunity in five years our question is well buy that to you means somewhere between nine hundred million and 1.1 billion or do you mean somewhere between zero and ten billion because all the action is in the range and so that's our goal up front is to get essentially bound all of these assessments and then we can do a sensitivity analysis and out of the hundred things that you might think about in a venture decision there's only five or six that matter and once we can pinpoint those five or six it matter now we can do a deep dive into those issues and at the end of the day calculated an explicit risk return so this is our process by the way if I just compare this how this compares to most VC processes it we're similar on step one and a lot of VCS I don't know how they get to ultimately making it to decision but then the magic happens so step one looks a lot like ours and they'll even have here are the criteria that we're doing to sword so it looked very organized at front and then a bunch of conversations happen and they make a decision so I'm going to jump into a case study in just a moment but any questions on the context before I jump into the case study yeah to actually see what chances there are the company actually crossing the chasm or getting the you know not getting past it or how do you actually like calculate those probabilities that with the case that it turns out that that's a quite an involved answer to that but yeah but that's that's really the heart of it which is how do you really understand quantify if you all the risks in a believable way all right so let's talk about a case that I mean oh so fie social finance you heard of that company so it was actually a group of folks out of the Stanford Business School that started this back in 2011 and the team over there led by a Mike Cagney essentially looked at student loans and they said this is kind of a strange market so here you have people coming out of the Stanford Business School and Stanford with less than a 1% default rate people coming out of the University of Phoenix with a 20% plus default rate and they're all paying the same interest rates it's like how can that be possible well government regulation is the reason that's possible and so he said well the reason everybody's regulated is because they're all banks so we're gonna show up and give student loans and we're just gonna cherry-pick the best customers the best credit risks and give them a lower rate loan and we're only going to do that for basically there's a hundred schools out there that have a one percent or lower default rate and we can go do that because we're not a bank so that was the initial idea and it came through a group called Stanford Angels and entrepreneurs so if you haven't heard about this I know somebody here from Stanford angels and entrepreneurs and ok we have one member a couple members great so it's a group of alumni just what it sounds like who want to be who are alumni of Stanford and want to be connected into essentially the entrepreneurial ecosystem and so is that we've we found so if I through that group we met them in September of 2011 and and this is what I'm going to show you is the actual analysis we used in evaluating so Phi so I've got permission from Mike the CEO or former CEO over there to use these numbers in a public setting so in terms of the filtering criteria Mike was a very compelling CEO he'd been an executive at Wells Fargo and deeply understood student loan area it was a strategic fit seed stage Enterprise IT it turns out it's a very much an enterprise company in terms of where you get get the funding by the way if you say it's say to a student how would you like a 4% loan or a 6% loan which one do you want that's not a hard consumer sell the hard place is well how do you go get 4% money that's best so that's where the enterprise IT thing came in and then in terms of the market opportunity student loans is a trillion dollar mess in some ways so fits our initial assessment so then our question is okay well what's the chance this company's going to be able to essentially go through those various phases to get to be a successful company so we start looking at an early stage success and by the way the framework I'm going to show you comes from a guy Jeffrey Moore who wrote a book called crossing the chasm so this is back in the 90s he wrote this and to this day I still think it's the best explanation of how a startup needs to change from what being an early customer where essentially you're talking to paying visionary customers who like to have cool technology all the way to a successful company is selling to mass-market customers who can care less about the technology only care about business values so there's a shift that has to happen as companies grow so here you can see some of the dimensions upon which we assess risk and this is just for early stage so usually early stage is defined as you've got somebody that's paying you money so you've been able to build your product you've been able to put together a team get enough money and at the end of the day you're successful because somebody's paid you money for something now a lot of VCS talk about these same dimensions what's different is we put numbers on all these numbers in the sense of what's the chance we think that this company's going to be successful on say like the financial dimension and by the way at the time we invested in so fie they'd already raised most of the two million dollars they were trying to put together further initial rounds so look we were actually gonna top it off so they hundred percent they've raised the money that they've said they needed the team was great hard to imagine a better team the product was interesting in the sense that they had a little bit of risk in their product but this was a company I was like right there they were like they had a deal with Stanford Business to offer this to students and so forth so there was just missing a bit of execution on the product so very high chance of early-stage success however given early-stage success what's the chance now you can cross the chasm and so the distinction here is you're going from visionary customers to pragmatic customers so these are folks that they don't care that you're like the cool new social way to get loans they're just like give me a better rate and make sure that you know there's no risk associated somehow with me transferring my loans to you and so they buy in a very different kind of way and so we look at the market and kind of like okay so the Stanford Business School is willing to take a shot on you oh by the way they're five you are from the Stanford Business School so not you can understand how that might happen but what about you know Harvard Business School or Williams College or these folks that have no connection to you are they really gonna offer you as a basically a student loan program to their students are they gonna endorse you so that was like wow you probably but 70% chance so we could easily imagine these other schools saying nope sorry this is too risky for us to be promoting you to our students and you see a various risks on the product and the team and so forth now what's interesting about this these are all individually pretty high chance of overcoming that risk but you can die on any one of these things so Williams and a bunch of colleges could say yeah we're gonna promote you to our customers and guess what you run out of money you're dead or the team can't get along and the team ends up you know splitting apart oh you're dead so these are all different ways that you can die and collectively ends up being a surprisingly low chance of crossing the chasm once you cross the chasm you typically do it in a niche and so the niche might be these top schools with less than with students with less than a one percent default rate but that's actually a smallish market so now you have to now the question is well how do you get to be a really big mass-market player it's like we're gonna have to figure out what are you gonna do with the other three thousand nine hundred schools that are out there and so and by the way at this point it's a highly regulated environment and you get big enough like was the government going to come in and say I know you're not a bank but you're acting an awful lot like a bank we're going to just apply the banking rules to you which would kill their business model so we put a relatively high risk on the market and collectively ends up being twenty six percent chance of mass-market success and then we've got a way that we think about the mass market share which is a little too involved for today now this is a lot of information but it can be nicely summarized in a decision tree so here are all those numbers from before so 90% chance early stage success given early stage success 45% chance across the chasm and so forth on the market share we break it down into this another book by geoffrey moore called the gorilla game where he talks about how profit pools tend to shake out in different markets where there's like a market leader and a challenger and so forth so we got those distinctions up there 60% chance of failure so even if they get you know a handful of top universities to adopt their program but they can't go beyond that at the end of the day this is effectively a zero for us so this is pretty good by the way sixty percent chance of failure a lot of times we're looking at 80% chance that sort of thing but you know this is kind of the dynamic so that 40% chance of success has to be attractive enough to make up for that 60% chance of failure so now we're talking about well how do you calculate the value of success well so this is that market mapping session that I mentioned before and I won't go into details but essentially you've got at the end of the day this is actually should it's multiple on invested capital I'm looking at and if we know how much we've invested we know our ownership at the time of exit and we know the enterprise exit value we can calculate what our essentially are multiple on our investment is now it turns out a ownership is kind of tricky in the sense that it depends not only on the valuation in terms we can negotiate in this round but a whole bunch of future dilution that might happen so every time you raise another round you hire more people there's more dilution there and interesting we found entrepreneurs are not very good at assessing dilution so we've got benchmark data and we just do that all ourselves whereas when it comes to the market side that's really where the entrepreneur is the expert and so we're relying heavily on the entrepreneur not only to like jointly we create these distinctions but to assess them so once we've done all the assessments we get our sensitivity analysis so these are essentially all those bubbles that were in the influence diagram and you can see here that at the top of the tornado we have mass market share so assuming you get to be a mass market player then the question is okay how do you compete with Wells Fargo Bank of America all these other folks that are in the student loan business and if by the way you have a hard time competing and you have a three percent market share our probability weighted multiple on invested capital is less than 10 by the way 10x is where we want to be if on the other hand you're the market leader and you have a 30 percent share this is a fabulous investment for us so now we're more like in the 50x by the way everything else kept at its base value so this is looking at just a sensitivity around mass market share all by itself and the point of this is not to give you an answer on the decision so much but to help you understand the key drivers of risk and value so this is actually there was you know a bunch more on here this is only looking at the top was it looked like top dozen top 15 but you can see you know you've got two that really stand out and you kind of get down to here and those top four or five variables it's really about ninety-five percent of all the uncertainty in this so those are the ones that we really want to make sure we understand so now you put it all back together and this is if you will the bottom line chart where we bring all of it together so here we have the risk from before and we've pulled out a few select numbers from the analysis in order to help tell the story so in the event you're a market leader two percent shot of that this is about 1.6 billion in revenue it's an enterprise exit value of 9 billion so this would be a good sized IPO our multiple on and that capital is 492 so if we put $1 in we're getting $492 back sounds great but we discount that by the chance that's going to happen so 2 percent times 492 gives a 10-point forex probability weighted multiple on invested capital for that scenario we do the same thing for each scenario add it all up if this bottom line number is 10x or better you're in the investable set for us so we just pause there for a moment because this is kind of the the punchline slide on the investment side what questions do you have yeah yeah so the good question so the goal of this exercise is not to come up with the right answer the goal is clarity of action right so this is this is a way for us to take our gut feelings put them down in a way that we can test them and then once we test them that'll it help inform us so at the end of the day if it's 9.8 this is not that precise of an exercise so that so then it'd be the kind of thing which is do we believe it so that's always the question so here's the thing so at 20x we're like you know there can be a bunch of stuff in here that we're off on and it just doesn't matter because it's so it clears the bar by a lot if it's 9.8 it's like hmm we better really believe everything in here so now we're back to the tornado and we're saying okay let's take those top bars like do we really believe them because if we're off even if we're off on like one of like the high range or the low range that could actually change a nine point eight to a nine or an 11 or that sort of thing so so it's a lot harder for us at the nine point eight sure in short story yeah there's no getting to the market needs by five point two two points still 24 added up with giving 10.2 if you've got nothing soon there's zero problem oh yeah so that's kind of interesting okay in other words that top one becomes kind of great mmm-hmm right I mean that would be wonderful if it happens but it's not a bad investment it didn't it yeah sort of sort of interesting I'm in and and that plays out here because this is was a much better than average set of numbers yeah what's what's interesting if you just take away this top yeah scenario like Ron suggesting in most cases it takes a good investment it makes it not good because like it's okay right here so let's all say this were a 15 X you take away this top scenario and now you're at five likes so that would be kind of more typical so so really really sentence by the way this gets into our filtering so when we're like okay do we go from kind of an initial intro to doing due diligence in a company one thing we look for you does that scenario exist are the entrepreneurs playing in a big enough market opportunity that we could imagine that because if the answer is no we just don't even go through the analysis yeah go ahead is that a time scale for the stages that is taken into account here like when do you determine early stage success and how long before you go past the incubation period to determine the first first gate I guess yeah well so I one of the one of the large decisions we've made or one of the large simplifying decisions we've made is we don't look at timely and and by the way this so that this is I say a lot of people in venture and most people in investments look a lot at timing and even this thing called IRR which is a which is a timing based notion and our notion on this is like it's really hard to predict the timing of innovation waves you know how quickly our smartphone is gonna take off for IOT gonna take off and I don't know shows well IOT is gonna take off and I don't know if it's going to be two years or ten years and so in here basically what we're saying is like with sofa if this happens in five years this is spectacular if it happens in ten years it's still good I mean if we're if we're in this market leader scenario we can hold so five for 10-15 years and we're still get really great investment returns and so so and by the way this is this is because we're doing early stage with really huge potential outcomes we think that allows us to not break our brains if you will on timing and still have a robust system but by the way this is another element to your to your question which i think is interesting which is when do we start learning from this so so there are some things that we can see so if they never get out of the gate and never have any customers you can see this very easily and when that happens our notion is well did we miss something so do we just get unlucky by the way so if we make 10 sofa investments we'd expect one in ten of them just to never even make it out of the gate so it could just be it's like oh well you know that was just the one in ten or if there is something in here let's say that the team melts down and by the way we gave the team really high marks here in the early stage it's like boy we missed something and so one of the things that we've built here is a learning system so we can look at what's actually happened versus what we predicted is going to happen along the way and that allows us to if you will update our assessments and hopefully get smarter as we go yeah Ken circumstances where the structure of your funding both altered the probabilities and and capped or reduced the the right-hand column there so it's so it'll be an example that your people are exploring funding arrangements that you don't walk in but come closer delivering 3 X 5 X something like that but they change the probability bands around what you might get which gives you a trade off because you've got a you've got a number there at the end which you know without a probability distribution around it and imagine that we could yes some reasonable degree a priority think about through holding in the tails of that Mister be a what so there's definitely you know investors out there that try and lock in profits if you will but you can if a company goes public you can lock in profits by essentially shorting the company and essentially what that does is to say well if the company's stock goes down I get paid but you essentially lop off the tops of the Bands if you do and for us we're really outlier driven in the sense of you know somebody being in this category makes our entire fund so and our fun we're gonna have sixty investments if we get one that's here everybody's happy we're happy our investors are happy that sort of thing so so we're we're reluctant to do anything that reduces the topside possibilities even if it can you know basically shore up the bottom if you will yeah since you have to trade investments in particular areas in order to increase the odds of getting that market leader do you do multiple bets in a particular area or are you kind of a cross you know so we don't have an explicit policy not to do that I say this is more of a relationship thing so if we were to run across a competitor to safai what we would do is we would actually probably go to sofa and say hey look we're thinking of investing in this company over here would you have a problem with it and especially what's companies kind of get a little bit of success under the belts we found them to be pretty lenient on that hey Bo we're a small investor so we put like five hundred thousand would be a very typical seed stage investment okay so far has raised two billion dollars in capital so anything that Lulu is gonna do is you know fairly noticeable for sulfide and but you know but if but if it was like early on and Mike was like boy you know if you and you know lend up or something like Lending Club I had agreed that we really made concerning to me then we wouldn't do it I guess I miss some more if you picked any other rural interests you what's a IOT yeah how many bets would you feel that you needed to place in that in order to try to get to you know how do you think about that what so that so the portfolio construction is actually it's a somewhat involved topic I'm gonna talk about that in a little more in-depth but but but this is there's there's a lot to answering that yeah yeah beginning how do you decide ownership percentage or ownership stake because this seems to just indicate more of a shouldn't we invest or not rather than how much what percentage should we take yeah so so there's a lot of ways that were different than typical VC out there so a lot of these C's are like I have to have 10% ownership for 20% ownership and they actually build their whole story for their investors around ownership and you're right there's no ownership in this and from our point of view you know 1% of Google works just fine so the ownership is not really Irish our issue is probability weighted multiple and by the way it's built into here so if looks like let's say so if I was gonna double the price at which we paid then essentially all of these multiples get cut in half that's where that's where show up so it's embedded in here and but at the end of the day whether we're buying one percent of a company or ten percent of a company it really just doesn't make any friends it really doesn't make any difference to us so long as we're getting a good multiple in my way there's a I'll talk about this in a little bit in terms of follow-on so a lot of BC's say it's really important to follow on in your winners to maintain your ownership percentage and we'd actually argue that that's a fool's errand yeah you mentioned learning a little bit I think that's very interesting given you know 60 company portfolio has your has your learning indicated the year more or less by it or unbiased or yeah you talk about that like have you learned a lot or is your original process you know I think you're pretty pretty good at well you know so I'm glad I'm glad you asked that so early on but you know the folks that are the experts in the industry said this is a waste of time do it from your gut so we actually did something that our kind of gut tastes but you it's not just flipping a coin right so you call customers and do reference checking and look at their technologies a little bit there's a big long list of things that you do on your new diligence site and we had 20 investments we made our initial portfolio with 64 companies 20 investments we made without this process and some of those because it was you know so obviously was a great there's a great opportunity that we didn't have to go through all this work in this process and 44 we did the process 10 to the 20 companies where we didn't use the process are out of business so a 50 percent failure rate 5 out of the 44 where we did use the process are out of business so you know basically a 5 percent failure rate versus a 50 percent failure rate so one thing that this process at least does for us is that systematically helps us look at all the risk factors and essentially not fall in love with companies so that's one of the problems like you get so excited about an entrepreneur or an area or a that an area that they're in you're investing in the you this confirmation bias shows up right so you're doing all this diligence but you're looking hard for the stuff that confirms what you already want to do and things that don't confirm it you have a tendency to ignore so that's like one on this so as I'd say you know so our initial data ad hoc as it might be suggest that this is a much better way for us to make decisions and then if you say well what about those 44 companies we invested in so how did they play out relative to these probabilities and the short story is they're not failing nearly as fast as we thought you know it's a good problem if you're gonna have a problem on your assessments you know it's a part of our story to ourselves it's been a it's been a real bull market for the last eight or nine years since 2008 it's been a great time to be in the venture business there's just so much money that's piling in you probably heard of the soft Bank fund they've got you know 91 billion dollars that they're investing in late-stage so basically there's all this money to prop up bad companies so so there we think that there are a bunch of companies in our portfolio that are at the end of the day are going to essentially fail but they just haven't done it yet because so much money sloshing around out there and they're good enough to go get that if you look at by the way just so fine so so phi right now is essentially at this level so last year they did about these are public numbers so you can rican find them in the wall street journal' the last year they did about five hundred million dollars in revenue and if i look at how so phi did relative to that so in some ways you know that's a surprise it's already in the kind of the top five percent category and it we're still pretty bullish i think at the end of the day it'll be in this top category what's interesting is i underestimated dilution so we're five hundred million in revenue so you'd think oh we're up 160 X on that we're up 83 X on that investment well you know it's awesome right so that alone makes our portfolio but it's half of what we had estimated back here and one of the reasons is because Mike when you went through this said lots of companies died because their undercapitalized over capitalization you know okay I get deluded a little bit more but I have lots more shots at success so he's raised over two billion dollars and I think that's one of the reasons why he's been successful in the sense that he has not been if you will stingy with how he uses his equity to pull in capital and actually here's a few of those numbers so what happened we invested obviously they've raised a ton of money what's interesting not only their revenue but they're hugely profitable and this is an amazing organization in terms of the business model and they're doing three billion dollars in refinancings every quarter now the other thing that I missed in here by the way was well I didn't really miss it it we just didn't model it but it from the very beginning Mike was like hey I've come from Wells Fargo there's two places where banks make money high net worth individuals and small businesses we're going to go after student loans as a way to get to the future high net worth individuals and future small business owners out there so that's ultimately where the big money is all right let me switch gears here and talk a little bit about portfolio construction so that's the individual investment decision but there's in some ways I'd say it's mixed out there in venture some people do a high-quality job some people do a lousy quality job and you know nobody quite does it in this portfolio constructions actually in some ways it's even worse in venture than the individual investment decisions so portfolio construction that's this piece in here so by portfolio construction how many investments are you gonna have in your portfolio how much do you have in reserve so how much do you invest up front versus hold money for follow on investments in the same couple by the way industry averages for every $1 you put upfront you hold three dollars in revert and reserve for a future and then what's the mix so this is the how many companies in what areas and that sort of thing so back to our notion of ego the standard approach out there and this is you hear this a lot when these things are raising money for themselves is invest in me because I can pick winners you know look at my past you know I was an early employee at Facebook or I used to work at Sequoia I've got this great track record of picking winners because I can pick winners I'm gonna have this concentrated portfolio of ten to twenty companies by the way I can make winners not just pick winners so these I'm gonna eliminate by I have to be on the board because I'm gonna really help out and win those winners emerge I'm gonna have a big pool here to double down on those winners sounds pretty good right I mean like a lot of confidence yet somebody get you know what's going on well we would say that all three of these things are a fool's errand first would say no one picks winners in this industry not John Doerr not Mike Moore it's not doug Leonean the famous names if you actually look at their track record okay John Doerr so he did Google he did Amazon you know amazing stuff right okay now he's got he's also got 30 investments in the clean tech world that have all gone to zip all right so so John Doerr his portfolio looks a lot like and it actually it's about twice as good as industry average but still you know a lot more failures in success and by the way if you can't pick winners okay now you're faced with you know this two and a half percent number or some really low of uncertainty numbers what does that imply about your portfolio we'd argue you have to have a large portfolio and oh by the way the reason you have a large portfolio is could even a couple of outlier successes that make everything the only way to get the outlier success is to put your money upfront when you can buy a lot of the ownership if you're doubling down on winter is conceptual you're putting money into late-stage venture with late stage venture returns so let's look at this from a couple of look at some data in the industry so we like data this is data that comes from essentially Burgess which is an accounting system for private equity and venture capital so the nice thing about this it's not like biased in terms of who's in and who's out if you use their accounting system they've collected this data over 40 years if you look at early stage venture the mean of the industry over 40 years 22 percent by the way best performing asset class by far of any major asset class the median of the industry 5.6 percent now this is a little weird actually this is the only major asset class so think public equities real estate distressed dead etc etc etc the only major asset class where there's a huge difference between the mean and the median what do you think what's going on there yeah great but really successful startups that make like way more than expected yeah right so there's sort of disproportionate returns to the successes matter of fact if you look at the underlying probability distribution over returns early-stage venture is described but what they call the power law so let's wanted to think it kind of looks like this where you have a huge amount of failures and then you have this very very long tail and out on the end of the tail you have Facebook and Google that are returning a thousand or 10,000 times people's money and those small number of huge returns pull up the mean even though the median looks pretty normal Oriol looks a lot worse so if you look at what are some of the lessons from this well one of the lessons is if you're playing early-stage venture you want to be the mean you don't want to be the median so one way to get to be the mean imagine you if you invested in every single venture investment done you'd have the mean by definition so if you can do an index fund a venture early-stage venture that would be an awesome investment product well it turns out that does that doesn't work so I mean there isn't at that product if you make one investment in venture odds are you're going to be closer to the median so now by the way late stage VC has a little bit of this dynamic but not nearly as much but now late stage VC looks about half the returns on the mean side in other words if I'm doing follow-on rounds my follow our rounds are have that as a profile my early stage rounds have this as a profile so this would argue for put more of your money early stage in a bigger portfolio that's a top-down look if we do a bottoms-up look so this is what I mentioned before the two and a half percent of venture deals generate essentially all the profit in the industry this is cambridge data there's a group called hoarsely bridge which is a fund of funds that invest in a bunch of other venture funds and we go to a top quality fund so a sequoia a Greylock a benchmark excel turns out sixty percent of their profits come from four and a half percent of their capital investments so round numbers they're about twice as good as the average VC in terms of their ability to predict outliers and when they will pedal the pet ourselves the back on the back a little bit here and say we've got you know we do good sourcing we've got our process so let's say that uh lou has on average a four and a half percent chance of an outlier every time we write a check so we're in the same category as the top-tier VCS now we've got the data we need to answer the question how big does our portfolio need to be to feel confident we're going to give our investors high quality returns the blue line here is a top-tier VC the red line is the average VC and this shows your chance of an outlier well remember if you have an outlier everybody's happy if you don't have an outlier you're really going to struggle as a function of your portfolio size so at 50 we have roughly a 90% chance of an outlier now this is kind of a scary number so when I talk to our investors about this you know essentially what this means is we could do everything right and there's still a 10% chance we're gonna have mediocre results because we just got unlucky now you know if we can get up here to 70 or 80 you know we're maybe close to 95% but you just never to a hundred percent in this industry we now compare what we're doing with the typical VC story which is I'm really good I can pick winter so I'm gonna have a concentrated portfolio so I'm down here at ten investments and let's say I'm the average VC look at round numbers it's 20 percent shot at having an outlier in my portfolio and then you say well you know why would anybody do this and by the way you talk to most VCS and they're doing this especially in my world some of the micro VC world and typically what they'll say is well look at my last fund I had huge returns it's like okay yeah did you get lucky or were you good and how would you know the difference all right so by the way if there are a hundred early-stage venture funds following this concentrated portfolio twenty out of those hundreds are gonna look like geniuses because they're gonna have a concentrated portfolio with an outlier that's going to give them huge returns so really the goal here is not just one fund with great returns but it's a series of funds with great returns so if I've got ten investments I've got and by the way let's give these folks to benefit of the doubt so let's say that there are a top-tier VC there just to have a concentrated portfolio 37% chance of one fund with an outlier but this is not a one fund business you do one fund and a couple years later you go raise the next fund and you keep going the idea that you're gonna have three funds in a row with outliers 5% shot so it's just you know doing the probability math so at 50 you know the 90% goes to 73 and so from our goal when we want to get it as close to a hundred as we can and the reason we're not doing 100 right now is because there are some qualitative factors first of all how about sourcing and what's your ability to sort and by the way we source at Stanford there's fabulous entrepreneurs coming out of the Stanford community so no problems in terms of sourcing a hundred deals but now in terms of selecting our selection process is a little expensive so to run through that hope it's not it's not hugely expensive it takes us a couple of days but we can always still only do that a limited amount with some extent so we can do it all online so then that'll help let's move down this way and then support it so if we've made even with 50 investments we've got three of us in the firm how do we support them in a quality way and there's a bunch of stuff we've done in the sense we say we're going to be your best partner from seed to series a and when Sequoia comes in and leads your series a we're gonna roll off the board you're in great hands call us if you need us but so are our notions we want to we want to show up where we have comparative advantage so this story by the way of a large portfolio probably 10% of the folks in venture capital believe this and follow it so 90% are in that category and by the way a lot of the folks that we talk to that are potentially investing in us so endowments and fund-to-funds and so forth look at that look at us like we have holes in our heads when we say we're gonna have a large portfolio now there's one other point that I'll put on here in the portfolio construction which is followings and if only 10% of the VCS out there believe this we're talking more like 2% of the VCS believe our story on the doubling down so almost everybody in the industry believes that the key thing to do is double down on your winners by god when you've got that uber in there you know you want to keep that ownership percentage up and I we had one fun to fund that came to us and they're like Clint you clearly don't understand portfolio construction so we're gonna we're gonna build your portfolio model for you I was like awesome like Joey show me how it's done and so they got some more information they came back with their portfolio model I know this is based on their portfolio model and they said so this is you know you're gonna be a pro-rata through the series seed and so you got your 50 million dollar fun you're gonna put two hundred sixty thousand dollars in a seed and you know for your very best companies you want to maintain your ownership percentage so that means two hundred sixty thousand in the a and then the B and the C and I say well C with that C if you look at like where your returns from that that 1.3 million dollars is generating a lot of returns so that's why you should be you know doing your pro rata maintaining your ownership percentage you know it's very interesting so I took that same model and I just shoved all of the money into the seed round across the winners and the losers so it happens in the seed round then is we end up buying twelve and a half percent of each company as opposed to three point three percent but then the fund of funds like all but you're going to get diluted because you're not invest following on in your series a a B and C right yeah we are getting diluted and so here you can see the dilution and at the end of the day we end up with twice the ownership by putting it all up front this is twice the ownership in the best companies the ones that have survived and at the end of the day the fund multiple it's like everything being identical we're just investing the money earlier is twice as good as it is compared to doing the follow-ons know it what's interesting about it so I send it back the the model to them and I said so using your model but just looking at an alternative way to invest the capital it shows you're gonna be twice as good and the response it was yeah we believe in being aggressive upfront - okay but yeah so you know it's this is sort of interesting this is like so counter to the wisdom of the industry that you know even when people like literally see it in front of them with their model and their assumptions it doesn't register yeah you mean in terms of whether we nothing whether to invest or not or probability weighted multiple head yes so it turns out it's a little more complicated than this in the sense that we don't necessarily put all the money at the series seed so we get to a Series A and we do the exact same analysis we did for our initial investments and so we basically have an opportunity cost notion so every dollar we're going to invest has to compete with every other place we could make that investment and if it turns out that that series a investment is better than 10x then we're investors and if it's not better than ten by weight we think about a third of the time when our companies get to follow on Series A it'll be in that 10x or better category so I I just did it like this to sort of like take the extreme the corner condition to kind of highlight the differences yeah hang on like initial stages how often do you do investment companies are already vc-backed so you can you know not use so many resources to support the company and have a large portfolio so it's I say just descriptively by the way so so we're in fun too right now and we've made 27 investments in fun 264 and fund 127 in fun too and I am on three boards and about to join two more and my partner's on three boards so you add that up so five eight so it eight boards out of call it and we're making a cremeux so eight boards out of 30 call it so that means there's 22 companies where we're not on the board and we're definitely relying on other people to our co investors to be essentially the governance and watch out for our interest and be helpful and so forth and and and seed stage actually is really nicely constructed for that in the sense that most of these seed investment seed rounds are like a million to two million dollars and there's probably two or three or sometimes even four venture firms like us that are writing five hundred thousand dollar checks in that so you know and because this is a I mean there's a community here you know so I'm on your board I want to I'm on a board where you've made an investment you're on a board where I've made an investment we trust each other and so it works out very well in terms of economies of scale if you will and what's interesting is it does not work that way at all at the series a so the series a now where the the big guys come in they want a hundred percent of that for themselves so it's very rare to see Sequoia and Greylock both in a deal because they both have billions of dollars to invest and just by the way and they all know that the big money's in the beginning so they want to put all their money in the beginning so they've got a very different model on that stuff but short story is yeah we are we're looking to you know collaborate with others to on the support side making a large number of small investments early on is that there's so much more of a upfront cost there's upfront cost for each investment that you make now I'm curious about how you do do like lawyers fees and where else do you encapsulate that in your model or how I mean it seems like then you'd need a larger team to go and make so many small investments how does that work well so so it's certainly true that you know each investment has expenses associated with it and mostly on the time side and and I'd say you know by the way we struggle with that sometimes internally so my my partner Miriam really likes to build relationships with folks and get deeply involved in that sort of thing and and so she sort of feels the pinch on this sometimes more than I do so I was a consultant my prior life so I like to kind of like do the quick strike go in and help you figure out your pricing and now I am out and you have to do the hard work so but I'd say you know doing follow-ons is also a lot of work so it's not like it's uh because like when we do follow ones it's it's not as much work because we have a framework already and so we can update the framework so I mean if I ask you a question again maybe a look so it seems like you're getting at a slightly different issue more of a logical extreme and just say we're gonna invest a hundred dollars in every early-stage company yeah so so you know by the way that would be a really smart strategy if you could have access all right so the problem is by the way so so this is a world where we're batting four and a half percent and so if we start lowering our standard so we're starting investing companies that have a one percent chance of an outlier success we need a much bigger portfolio so it's kind of a balance between sort of keeping the the risk-return high and then enough volume and then you know so if I could figure out a way to put $100 in every company has gotten a venture investment from a reputable venture firm like I said I think that would be an awesome product and it's like that would be a really old correlation ventures by the way that's trying to do this in the marketplace and their biggest challenge is essentially access so on the very best deals like on our very best deals it's actually competitive so a company might be raising a million dollars and there's five million dollars it wants to come in and so one of our challenges is we have to essentially you know tell the entrepreneur of story why it's valuable to have Alou there as opposed to these other three male micro VC firms that would rather have our spot so so there's there's some bets actually those are the balances that lead us to where we are all right so what's next for us so from our point of view we've kind of feel like the one-eyed man in the land of the blind you know it's a pretty good place to be but I'm still operating with one eye if you will there's a whole lot that we think we can do next to improve what we're doing so one of the things that we're doing is we're spending some time line how do we automate some of this so right now this is performance art so I get up in front of the whiteboard and I do my thing or so much does this thing in front of the whiteboard but that's a very limited resource and the question is can we take this and centrally technologies it so we've got this benchmark data there's a set of patterns that we find out there is their way and the extreme would be imagine self-service venture-capital so if we've really figured all this stuff out make it available to folks online and say if you want money from from Oulu so fill out form and oh by the way when you fill out the form there's got to be validation for it so if you say you've got a customer that's great upload the customer contract alright so there's a proof points behind all this stuff and I mean but to me that's a pretty exciting place to go in the sense of like right now it's two relatively restrictive set of folks that we can interact with we only have so much time and so forth and you know we have two thousand potential entrepreneurs we could talk to each year and we have two hundred first meetings but in those 1,800 entrepreneurs I'm not talking to I'm sure there's fabulous opportunities and we just don't have the resources to evaluate them so if we could make this self-service and out of those 1,800 the ten that are really interesting could sort of rise to the surface I think that'd be a very exciting place to be and the other thing we're working on is omics has this working on if you will our meaningful purpose so think about entrepreneurs one of the exciting things about being in this industry is entrepreneurs come to us with a huge amount of purpose and when we're really attracted this by the way this is my values map that Soumik did with me and really what really drives me and so we've got from an emotional energy point of view it's leading by example on values like Rhonda mentioned that I had that company that wasn't going well I had raised some angel funding and I returned all the money and for me that was really important because you know I I had essentially an agreement in my mind with the investors and then the world changed on me and I couldn't I couldn't live up to my side of the bargain and so I wanted to make that whole and I forgot a few other places where that to me has really been sort of signature events in my life if you will unstoppable energy so what gives you energy in my mind this is really possibilities so in that sense venture capital is a spectacular business for me it's like I'm always looking at like the next possibility and I get a lot of energy from those but look this made me a lousy entrepreneur so so I spent 15 years before I figured this out and and it turns out there's this early part of entrepreneurship that I just love it's the conceptualization of you know what's the problem that customers have and are they willing to pay for it and how do I get product market fit and convinced those first couple of customers to buy my product and that was just super fun and then you got to this point and I got I had four companies I got to this point was where the word repeatable would show up and it turns out I'm bad at repeatable and so I was always kind of stranger how can I always get these companies up to a million or a few million dollars in revenue and then nothing happens like I said you know I finally got figured it out and then from an intellectual point of view you know what really drives me right now is decision analysis and venture so this was sort of I would call it hypothesis nine years ago and I'm fortunate to have had enough success in this to feel like I've now got a platform for doing more so enough success me when I first doing actually investing we were just doing angel investing so I was out of our own pocketbooks and now we've got a few legit VC so we've got outside investors we've got you know real capital and we've also got a group of folks that are forming around this both VCS and other folks in the industry that are essentially intrigued with this notion of decision analysis and venture and so now I've got a set of colleagues if you will to play this off so by the way so GV which is Google Ventures venture team is one of those folks that is kind of a thought partner and all this correlation ventures and a few other folks so so it's starting to become more like data in venture and what are the implications for a decision making you know you know knock on wood this will become wave at the moment it's a little bitty thing but hopefully this is but if I were step back for a second and say you know and I half believe this half want this so where is venture going so think 10 20 years from now well if you look at essentially where public equities where twenty years ago it was a bunch of smart guys in the back room making decisions it looked just like venture does today and in public equities you can't find anybody doing public equity investing like that anymore because it turns out data and models are such a better way to do this it's completely dominated the field of public equity investing and I'd argue that venture capital will could follow the same route where you know the thing about data and logic and models it's like it doesn't take that much in terms of wherewithal to be able to use those to your advantage so it's not like you need like this magic crystal ball in order to figure things out it's not like magic and maybe you have it maybe you don't you know if you're smart and hard-working and have some insights you know you can do some credible stuff in this industry all right there you have it look watch what questions do you have wait a few minutes well yeah swinger failures is this crossing like the 1 to 10 million annual revenue catherine or you just speak to a little bit more about those earlier also yeah so I'd say there what's up like a couple of the categories of failures so I do one company outcome software which was essentially we're going after risky recurrent decisions in big enterprise so risky in that they screwed it up recurrent in that we could build a model and structure to sort of help them on a consistent basis so like one of our customers with Eli Lilly and the question is well which products do you launch in which countries US Europe Japan it's easy but when is product mean literally like a portfolio of like thousands of SKUs it's easy huge markets you sell everything what about Peru or Czechoslovakia and so we had the system that we are we're the very first internet-based system Lille or intranet based system and we would pull data out of there ASAP system on my cost data and then we'd have a market manager saying crew making estimates on sales and volume and we've combined all this together and what used to take six to eight weeks would literally take 30 minutes however long it would take you to type in your sales and volume estimate you could immediately get you know the full analysis that we showed here completely sensitivity analysis and whatnot so it's hugely successful work for a bunch of big banks this would have been 2007 so we were no no sorry this was 1997 when we started this and so by 2000 we were geniuses and in 2001 we weren't very smarter anymore and you know if I look back at it's like so that had that at the fundamental flaw of we had to get two yeses inside each Enterprise so typically the finance department would own the analysis so they would have to say yes but then there'd be a business unit manager that was making the decision if you will with finance approval to launch a product and so we have to get both groups to say yes and it turns out if you're selling into an enterprise to get two yeses from two different silos is a nightmare and if you look at actually the big successes in enterprise they're almost always we're selling to the VP of HR or the VP of manufacturing or the VP of Finance we're not selling to multiple of them so that was I call it a strategic mean was both unlucky in terms of timing but it was a strategic mistake as well yeah and also there's a bunch of so I'd say our most successful company that has the most promise where I just got tired of it was a group called decision quality international so we were selling training materials to big enterprises we were as how to make and cause high quality decisions and it turns out if you go in there and to enterprise and say I'm gonna teach you how to make better decisions it's like most people are like well you'd help make me better how do you think I got to be an executive all right but I don't need that help but then we said well and cause well how about your colleagues tell me about the decisions they make so I go oh my god are they as awful so that was our hook and we got some really big organizations to train thousands of people in this and and the great thing about training is a fabulous business model so if you show up in front of like this it's like the economics are okay but if an organization says you know what we want to do this for all 5,000 of our managers and you're way too expensive Clint to show up in person it's like okay so what do you want me to do so you train the trainer and then you license them the materials and so that actually had a lot of promise but I got to the point where you know if I have to get up in front of a group of executives and do two days of decision training again just shoot me yeah it was like you know after like 40 or 50 times this is like I said I'm just not good at repeatable any on some respect that's that was the one with you know that could have been an interesting business if I just would have had would have been wired differently yeah there's an analysis process did you apply to choose enterprise IT category yeah you know so it's a bit of a trial and error so when we first started doing venture like nine years ago well so I might so my expertise is in the enterprise so I'd sold enterprise software and that sort of thing so it was easy to sort of gravitate towards that and we did a bunch of consumer stuff we did a bunch of healthcare stuff and it turns out that consumer is really hard to model at least given my level of expertise because there's this thing called cool that's hard to figure out so so some things come along and they're just cool and it's like boom you've got this monstrous adoption but the problem is I had a hard time picking out cool before it actually pumped by the way might my 18 year old daughter can spot cool from a mile away so if she were part of this firm we be doing consumer stuff but now if you go on the enterprise I think of the enterprise better faster cheaper so somebody can basically have an initial pilot customer in and we think we've got a fighting chance of figuring out whether it's that is something that's representative of the needs of the industry and they can have it sell through distribution channel so I think we've got a fighting chance at figuring it out in a way that in consumer we just could never figure out how we have sort of any advantage versus the other investor because as soon as something explodes in consumer then it's easy for everybody to see that oh yeah I want to be an investor yeah there's limitations the types of markets where things might be more difficult to apply soap examinee know what you have to compute all the probabilities for because that would've been a shortage maybe saw more blue sky type application or like I said a consumer application we just you don't know what you don't know yet so I guess what I'd say is the more uncertainty there is more value there is process so that's what we're looking for and oh by the way there's a lot of uncertainty and reinsert into consumer so I think consumer is a great candidate for this but we would just need to make some investments in figuring out the right distinctions and and yeah I mean you know I mean I could probably go do that means we've done some consumer so what does consumer looks like this it's driven more by an engagement than body so if you have a small number of users that are using it like you know an hour a day okay well now I can imagine the million people using an hour day versus if you had say a thousand people they would use it you know a couple of minutes a month all right well I mean you might have the same total power is in your system but one has got a really nice foundation and the other one doesn't and then you say well what are the patterns so let's I mean it's like I could imagine doing consumer with this it's just I haven't done it yet wait so I'd say we're we're it doesn't it doesn't work well actually it doesn't work well when nobody has any expertise alright so so if you can't come up with the distinctions that are important to think about you know that's the blind leading the blind what about is it more big mocha Java that's a biotech I would think availables do do it stage 1 stage 2 stage is probability thing yeah maybe it's even more applicable well so it turns out so actually so a lot of what we do here was inspired by what's happened in the pharmaceutical industry and actually oil and gas so in pharmaceutical R&D decision analysis or some version of it is essentially best practice and by the way in my prior life I spent 15 years helping to make it best practices at a few of these places same thing with oil and gas where you're poking a hole on the ground you got all this data and you know it's like one in ten one and twenty are interesting so so anytime there's lots of uncertainty this stuff shows up as being very important model effect or influence if it's a marketplace business where you have two sided players and in the ecosystem so so I guess you know okay we pull it up here but I just say so so so these are things I think the models very well designed to handle in the sense that you have to get critical mass on both sides of the marketplace in order for it to hit credit and then the question is what's what does critical mass really mean so if I look at sort of like the early stage success and crossing the chasm so crossing the chasm of the marketplace is what is critical mass because you kind of get to a marketplace where once you hit critical mass then the thing has a life of its own and until you get the critical mass that's pushing a rock uphill so somebody's asking about diversity in what we've seen so we have diversity as one of our investment theses say there's a lot of VC is basically okay no where do you think the opportunity is out there and so for us it's big data its enterprise cloud but diversity is actually one of our investment theses so if you get back to ok if 2 percent of the venture capital dollars are going after women teams it's kind of like there's systemic bias here that actually creates opportunity for us because we're gonna sort of systematically look at that group and other underrepresented groups and so we explicitly have you know you might go out and you're like in Big Data go find the interesting data companies we go out and find the interesting diverse companies as well and at the end of the day we have about a third of our CEOs are women if you look at like underrepresented minorities about 9% of our CEOs are but nine percent underrepresented minority CEOs compared to less than one percent in the industry and in terms of women CEOs it's depending on the numbers you believe it's like five to eight percent women CEOs and we're thirty percent and if you actually look at our outcomes so two of our best outcomes so far are actually two Stanford companies so crux which was done by actually one of Ron Howard's PhD students is a his family emigrated from Mexico his grandfather was actually an illegal immigrant his parents none of them went to college did five brothers and sisters all went to Harvard undergrad and then Stan Tom came to Stanford here he did a PhD he sold his company to Salesforce for a billion dollars so if you look at Latino entrepreneurs that's probably one of the biggest exits in the last five maybe ten years there's another company at blue river so was a gentleman from Peru into the GSB executive education program sold his company for three hundred million dollars to John Deere so if you look at me now I would argue that if you look at our result and like where the big results are coming from you know it's a it's a much more diverse set of entrepreneurs and and it you know but from Stanford right so okay so oh yeah so let's pill diversity in a box you know it's interesting so so my my partner Miriam is really into diversity and we've got some other folks that like the Kapoor Capital and some other folks out there that are really focused on diversity and we have sort of an interesting debate in some ways where like Courtney is Cal Kapur capital so Mitch Kapoor inventor of Lotus and actually original early found early investor in uber Sun incredibly well and his whole notion is and Freitas they want to raise the floor and so a lot of their investments are about raising the floor and Miriam's point of view it's like you know that's great love to have people out there raising the floor we need somebody to raise the ceiling to and that's where she wants to focus on raising the ceiling so I mean Canada diverse entrepreneurs coming out of Stanford it's a great place to you know work on the ceiling all right so I think we're out of time maybe last question investment strategy can do better this gives more volume yeah you know so I'd say we need a critical massive volume to get an outlier but once we've got an outlier I'd say you know so so the volume really drives do you have an outlier and if you have an outlier then the follow-on strategy maximizes the impact of that elevator that maximizes the returns from that where you know what number is it like are you investing in cookie what does it take sure thing well I mean so it's you know there's not there's not a number as much as there's a relationship between between oh you know actually that so there is this questions like so when is a follow-on spill way to ask ask this is when did to follow on a smart strategy and it turns out there is a there is a couple scenarios where follow on investing in a smart strategy so one is if you're terrible at selecting right so imagine you do a really lousy job selecting so you make 50 investments only one of them is successful well yeah you're gonna be better off holding on to your money and investing in that one successful investment so the other time that follow-ons make a lot of sense is if you're lousy at negotiating terms so let's say you go in and basically you pay a really high price and in fact that's exactly the same price that the series a folks invest in so you've taken all this risk and gotten none of the benefit financially so basically if you're not very good at seed stage investing don't do it yeah yeah I can share those slides is there a way to share the slides I'm happy to share it if we have a way to do it all right well thank you for coming [Applause]

Channel: Stanford Decisions and Ethics Center

Views: 49,437

Keywords: vc, da, decision analysis, venture capital, stanford university

Id: Wi3PiZsIfBU

Length: 76min 37sec (4597 seconds)

Published: Wed May 16 2018