Khosla Venture's Samir Kaul on building an iconic, durable firm, and the role of non-consensus decision making

Welcome back to another episode of Venture Unlocked, the podcast that takes you behind the scenes of the business of venture capital. I'm your host, Samir Khadji, and in this episode, we have a conversation with Samir Kaul from Khosla Ventures. Founded in 2004, Khosla is one of the largest and most well-known venture capital firms in the world and led by legendary investor and entrepreneur Vinod Khosla. The firm is known for investing in companies that are solving very large and complex problems. The firm currently has over $15 billion in AUM and has invested in companies such as Square, DoorDash, Stripe, OpenAI, and Impossible Foods. During the episode, Samir provided his insights in investing across cycles, the market insanity that we saw pre-2022, and how they approach both building a firm and investing. We recorded this prior to the news that Keith Rabois was rejoining the firm, hence no mention of it during the discussion. Hope you enjoy the conversation and let's get right to it.

Samir Kaul is the CEO and co-founder of Allocate. Allocate and Venture Unlocked are independent of each other. Any statements or references made by Samir or his guests regarding third parties, investments, or securities are solely their views and opinions and are not intended as investment advice or an endorsement of such parties or securities by Samir, his guests, or Allocate. Allocate or its clients may maintain relationships with or investment positions in guests, third parties, or securities mentioned in this podcast. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions.

Samir, thanks for coming on today.

Thank you for having me.

So I want to start way back actually and go through your history of how you got into technology and venture. So maybe we can start with your background working at Flagship and then ultimately going to KV.

I started my career as a scientist and then after a few years of that, I went to business school where I met Noubar, the founder of Flagship, and I joined forces with him back in 2001, 2002. What I basically did with him was a little bit of investing and due diligence, but primarily finding academics with interesting technology and helping them commercialize it. And so we did a number of companies in therapeutics, diagnostics, and in kind of the tools market. Ultimately did one in renewable energy, which led me to Vinod. And Vinod was an investor with both his Kleiner and individual hat in 3 of the companies I started while at Flagship. When he left Kleiner and decided to start his own firm and focus on sustainability and renewable energy, in addition to traditional technology, but he hired He said, let's join forces. We started the firm together and I moved out to California in 2005, and kind of the rest is history.

Maybe we can go to the early days of Khosla, and I'd love to hear from you, what were some of the strategic conversations you and Vinod and David had in terms of the type of firm you wanted to build, both from an investing standpoint, but also the overall operating ethos of the firm?

Well, we always wanted to build a firm that was going to be technology heavy. We always felt like our edge would be our ability to evaluate technology, assess technical risk, prioritize technical risk, and have the, have the appetite for risk that many others wouldn't have. That would be our differentiator as opposed to chasing the next fintech or social media or crypto deal where it affects the highest bid wins. And that works in bull markets like we had for about 10 years. We've lived through a lot of the tougher markets, dot-com, 2008, et cetera. That doesn't hold forever. Technology is a sustainable long-term advantage. That was what we wanted to focus on. We wanted to surround ourselves with investment professionals who were all technical, all had a real curiosity for technology. We wanted to provide industry-leading or top-decile returns, But we wanted to be proud of the companies that we invested in and felt like all companies should be society or mission positive while providing a return. We were going to also, unlike many firms, put a lot of our own money at risk. So we have always been our largest, our own largest LP.

And so that was obviously from the beginning as you, and this was during a time almost 20 years ago where [Speaker:SAMIR Kaul] Now it seems like there's much more investing done in companies that have this intersection between science and technology. Some people call it deep tech. Of course, that term itself is a little bit amorphous, but you've been investing in things like sustainability, clean tech, really from the mid-2000s. When you think about some of the entrepreneurs that you were backing, obviously, in order to build a highly technical defensible company, many of the entrepreneurs are people that are CS engineers, they're scientists. What do you look at when you are evaluating an entrepreneur to know if they are a KV type of entrepreneur and building the type of companies you're looking to back?

What you want is someone that can take feedback well. We tend to be pretty direct, we're demanding, so you want someone that is going to run through walls to get the company successful, is nimble on their feet. The only thing you can guarantee in life and startups is that the initial plan will not be the final business plan. They need to be great recruiters. The team you build is the company you build, and they have to be maniacal about retiring risk. And so you want to always lay out what are the binary risks to a company working or not, and as quickly as possible, retire those risks. You want to fail fast if you're going to fail. The worst companies are the ones that linger.

You said something that I want to double-click on because I do think it goes to the overall ethos that you've created. So on the website, and I think I've heard you talk about, I've heard Vinod talk about this, is that the concept of venture capital in terms of company building, so not just investing checks but helping entrepreneurs build these long-term companies that are built to last, requires a level of brutal honesty versus hypocritical politics. And I, I think over the last decade, decade and a half, what we've seen is this concept of founder-friendly being you do whatever the entrepreneur wants, you, you act as a cheerleader And from what I've seen with CoSlide, it is much more around, let's be shareholder-aligned, let's build a company. That's going to require some very tough conversation. How do you balance that level of tough conversation with this level of empathy that people are building really hard things and inherently within a startup, you're going to go through these ups and downs?

People feel like you can't be founder-friendly and tough. I think, I actually think it can't be the opposite. You know, you're not doing the founder a favor by letting them make mistakes or turning a blind eye when they're doing things that are wrong. You know, the investors on WeWork's board weren't founder-friendly to Adam Neumann. He got thrown out of his own fucking company because they allowed him to do a lot of silly things like build a sauna in his office or throw these crazy parties on the weekend or sell his own real estate. Like, kind of, you know, inside baseball stuff, but that didn't do the— that wasn't being founder-friendly. Company's not bankrupt. So just because they cheered on the person, that's not founder-friendly. That's hurting the founder. We're not tough on founders because we get our jollies by being tough. You know, we're genuinely trying to help the founder build a great company. You know, I have 3 kids. You know, I let them know when they're making mistakes, and it's not because I want to be mean to them. It's because I love them. And you have to be honest and tough to make them better. Why would we get paid fees and carry if our— if all we did was invest in a company, let the CEO, he or she, do whatever the hell they wanted and have no input? Then we're like a hedge fund. We're just buying stock and selling stock. That's not what we do, and that's not what we get paid to do. Now, we're pretty open about it because entrepreneurs should know what they're getting into when they work with us. If they do— if they want cheerleaders, there's plenty of firms they can go to if they want someone that's going to be there Saturday night at midnight or jump on a red-eye to go help recruit someone or close some sales deal. They'll come to us.

So when you think about this, and I completely agree with everything you said, I think during the peak time of ZIRP, everyone felt like in order to win deals, you had to be a certain way with founders to be able to get that high NPS score. I think that the more mature founders, the ones that have been around the block, actually like somebody that is going to act as a real sounding board, is gonna challenge certain hypotheses and help them build businesses. You talk about this on the website, obviously you let founders know what they're getting into when they take an investment from Khosla, but oftentimes you have an entire board around with other VCs that might have a different way of interfacing. So how does this manifest actually with the day-to-day with an entrepreneur, whether you're talking to them one-on-one or even within the dynamics of a boardroom when you have other VCs that might simply act as cheerleaders or simply lack the experience that you have in providing credible feedback to the entrepreneur?

Well, you know, in a broader board setting, it's hard to give that kind of feedback and be effective. So I tend to save a lot of the toughest criticism one-on-one at the boardroom? You know, I'm not going to shy away from it, but that— the dynamics there are different, and you want to be careful. A lot of other VCs like to position us as the bad guys, so they'll like, you know, stare at their shoes when there's something obvious that needs to be discussed at the boardroom. Look, the good CEOs want to be told. Look, the very good CEOs could probably sell all their companies. The reason why they take money is they want to help. They want help and they want a sounding board. They want people to give them advice. And it's— and if you look at the CEOs in our portfolio, the Max Levchins, Sam Altmans, the Jack Dorsey's, you know, these people have chosen— no secret how we are to work with. They've chosen to work with us multiple times, time and time again, because they appreciate the feedback and they're strong enough to know which things to listen to, which things to discount. And they want to be told because they want to get better. That's pretty clear. And in boardrooms, I often find myself defending the CEO more than anybody else. You know, hey, look, let the person— let he or she spend money on R&D. It's okay. They missed their sales number. You know, the goals are ambitious. The prize is still there. Those kind of things. Like, I usually end up being the biggest champion. The irony is, right, In a more public setting, I tend to be more of the champion. Look at the OpenAI drama, right? Who was tweeting more in support of Sam Altman than Vinodh? So in a public setting, you know, we tend to be the biggest supporters of our CEOs and they know we have their back. But one-on-one is when I say my toughest criticism because there's no, you don't need to, we're not looking for an audience.

Yeah, I'm glad you brought up some of those examples. And as we think about support of entrepreneurs, Vinod, for example, has been very publicly supportive of Samir Kaul through the entire OpenAI ordeal, which lasted about a week actually. Thinking, I guess, maybe beyond that for a second, you've been around for 20 years now in this firm, and the venture market during that time has changed, and there have been a lot of people that have done well, because of interest rates and the tailwinds that we saw. And the notion of founder-friendly, of course, as you mentioned, changed over time. One question I have, I guess, for you, just kind of zooming out, is how have you seen the evolution of venture during the time that Khosla has existed, and how has it potentially impacted how you work with founders and what you do in terms of creating these competitive edges?

Well, any dipshitter could have made money the last 10 years with zero interest rates and So there are a lot of people who are riding high. And now I think it was evident, I think our LP spoke, we raised the largest fund we've ever raised in a time where both VCs are condensing or going out of business because I think people love the venture asset class. They see that this is a great time to invest because prices are finally coming to reality in most cases. There's still some sectors who I think are pretty expensive. Pretty frothy. And we've been around the block. We've been there through dot-com, through 2008, through COVID, through all these things that we've seen really tough days. We've got to navigate it. A lot of these newer firms and even partners who are newer partners at established firms, I'm seeing it in boardrooms, are, you know, like deer with their headlights because they don't know. They've not seen an inability to raise money. They've not seen the need to do 30%, 40% rifts and layoffs. You know, they've not seen debt holders coming after them for breaking covenants. For a few years in there, we weren't that popular with some entrepreneurs because we were still trying to enforce discipline, not have crazy expenditures and stay relatively out of the, you know, not over PR because we knew that eventually, frankly, it took longer for the doomsday to come than I thought it would. But it stayed. And now those CEOs are spending a lot more time with us because they know that we have the experience. It's not to say that there are a lot of other VCs that are, that are out there that have had that experience. The guys at Andreessen or Sequoia, General Catalyst and other, many others have been through all these ups and downs also. But, you know, we're definitely in the minority now. Probably a dozen or so groups that have leadership who have been through tough times. A lot of the newer firms, I think, are going to struggle because they don't— they're— it's just, it's hard the first time. The first crisis I experienced in this business was terrifying. Now you could be a little more even-keeled. You don't want to overpay, you don't want to overcut, you don't want to sell everything under the sun. You got to be just— and you need to prioritize. You only have— you have a fixed closed-end fund, so you only have X number of dollars as reserves. Every company is not going to get reserves. Many companies that probably deserve a shot just aren't going to get there. So now the trick is, how do you sell companies that don't meet the bar for getting further investment? You know, and how do you pick the companies that are going to get more money and optimize those investments? And that's another thing, is that you need to be realistic about the right valuations. The more experienced VCs don't care about what they're— about some artificial vanity mark on your books. They want the company to be marked for success. So there's upside. So employees get upside. And that's another thing. A lot of VCs want to go out and try to raise other funds and they're scared to mark down their investments. And that's, again, very short-sighted.

I completely agree with that. And We certainly saw that in the 2010s where because of the capital abundance in the market and the number of people willing to mark up companies, especially larger crossover funds, funds were able to mark up deals and as a result, were able to use those marks to raise subsequent funds and bigger funds. And of course now everything's unwound. The two areas that seem somewhat insulated from the reset that we've seen over the last two years are seed stage financing and also anything that relates to AI. And of course, the discussion generally is around generative AI and the potential impact to enterprises and consumers in the future. And since you've been an investor, you were an early investor in OpenAI, have tracked AI as well as some of the last technology supercycles, whether it be mobile, cloud, the internet, wanted to get your take on how you compare the potential of the AI supercycle relative to some of the tech supercycles of the past?

Well, it'll be bigger than anything that's been done before, but there's not like an AI company. Any company, even pharmaceutical companies or companies that you wouldn't think would incorporate, wouldn't come to the tip of your tongue when you think about AI. AI is going to become like software or the internet or mobile or the iPhone. Unless a company has incorporated AI And whatever they do, supply chain, manufacturing, etc., etc., they will be left behind. That is clear. But, you know, if someone says, oh, I'm doing an AI company, that's probably just marketing bullshit. You should run away. If you've got someone that isn't incorporating AI in their short to near-term plan, I would run away. It's just going to be a pervasive technology across all verticals. The valuations in AI right now are super high. For the most part. There are some that are warranted and justified, but some are very, very high and are following on. If OpenAI is worth $86 billion, we're worth 10% of that, 5% of that. Those are always kind of a fool's rush, right? Like a fool's errand to do analysis like that. But that's how a lot of companies are getting valued right now. You know, we've seen this in dot-com, we saw crypto, we saw clean tech. You have the version 1.0 where there's a lot of exuberance and then a lot of disappointment. And then in version 2.0, it's much more robust. People learn their lessons. You started in '99, so you'll know that the eToys, Pets.com, you know, volume, volume, volume, despite the fact that every transaction was profitable. Was all the craze. And then people left dot-com for that. I remember 2002, I graduated school, people are like, wow, the internet, there's no more tech investing, tech investing is dead, blah, blah, blah. And then you had Facebook, Uber, Airbnb, Stripe, Square, DoorDash, you name it, on and on. Yeah, kind of the most, uh, probably the most richest, most lucrative part of venture capital was in Tech 2.0. You're seeing that, you know, you mentioned seed rounds are inflated because it's not that much money and not that much. There's still a lot of— and seed checks are in the funnel. It's more— there are a lot of people who can write a $50,000 check. The other area that we're seeing a lot of lift in valuation is in the sustainability area. I lived through 1.0. I have all the scars on my back. You had some successes. Lonzatec and Tesla and Quantscape, but there were definitely some successes. But for the most part, it was at least 80, 90% failed. Now you've got version 2.0 and people have learned a lot of the lessons. Don't be dependent on subsidies, be clever on capital allocation, focus on the founder, don't hire executives from British Petroleum or Shell or Exxon. We've learned a lot of that, incorporated that, really excited about that area. But my bet is that, you know, AI will go through the same thing. I think it'll be a quicker cycle. But, you know, then you'll have a bunch of companies that fail. You're already starting to see some companies stumble. We'll see more of that. And then you'll see a second wave come that will be very powerful. So aside from OpenAI, we're the first VC investor. Most of our investing in AI companies are What's next? What's next after LLM? What are some specific applications? Maybe it's in healthcare, maybe it's in financial services, maybe it's in legal services. Where can AI play a real role in more shorter or longer term and start investing in those areas right now? And stayed away from kind of the multi-billion dollar valuation mega growth route. It would really, with very few exceptions, you stayed away from those rounds and passed on a lot of those companies. A lot of them are pretty good companies, good teams, but just We think there's better ways to deploy capital.

One of the things I do want to unpack a little bit more is something you alluded to. So thinking back in '97, '98, '99, actually companies like Pets.com or even Webvan, at the time the infrastructure, the adoption wasn't there to sustain those companies. The distribution was there, of course. Webvan, ultimately now we have Instacart, right? Where—

It's good to bring that up because we invested a million bucks in Instacart Instacart for 10% of the company. And the reason why Aporva had struggled to raise money at a better price was everyone kept breaking up Webvan. It was still early days of the iPhone, but remember, you and I are old enough. Webvan was, you got home, you went into your dial-up AOL modem, you had a shitty user interface, and you ordered groceries that took forever to get there. Instacart, you get in your Uber on the way home, you press a few buttons, that the ship's there by the time you get home. Very different. But you had to make that technical leap. Back then when Instacart started, people were still scared to put their credit card online. You had to make the leap that everyone was gonna have an iPhone, broadband was gonna be more pervasive and faster, people would be very comfortable doing all commerce on their phone, and that people would figure out the delivery infrastructure, 1099 workers, all that stuff would all work and people would do that. It's obvious now, but back then it wasn't, or we wouldn't have been able to own such a big chunk of the company for a small amount of money.

Of course, that's the fundamental nature of being a successful VC, of seeing in the future and not being tethered to what has happened in the past, understanding that biases that may exist from past failures simply may not provide the map to the future territory. So Great example. But maybe going back to AI for a second, in particular, going back to the hypothesis of when you invested in OpenAI. At that time, it wasn't really in the public eye, generative AI. There was no ChatGPT-3 people were using. But what did you see as you looked into the future that got you so excited about the opportunity?

It was pretty simple. Fido and the firm has been pushing AI and thinking about what AI could do for a long time. So that's one. So anything— so probably very early days we're looking at that area. And then second, more important than an idea or even the technology, at least to me, is the quality of the entrepreneur. And look, Sam Altman's a once-in-a-generation entrepreneur. There's not many people like Sam. There aren't many. He's super smart, very mission-driven, hardworking, recruits well, raises money well. You know, we would invest in anything Sam was doing. I mean, case in point, $50 million, our single biggest check, first check ever into a company that was actually structured as a nonprofit. There's less than 5 people on the planet that we would've done that for. Samir Kaul happened to be one of them. So it really wasn't more complicated than that.

Now, now that you mentioned things that you're looking at in the future, so it's the next generation of AI, the next generation of sustainability startups. How do you know? Because ultimately you want to invest before things become evident. But not too far in advance, i.e., the Webvan, the Pets.com. What are the markers of timing when you're looking at things to invest in companies that are building for this future you believe in, but not something that might be 10 years away?

[Speaker:SAMIR_KAUL] So I don't think 10 years is too far. We incubated QuantumScape 12 years ago, and they're a public company. You can read their reports on what their schedule is. I don't think that's too far. I think Webvan was early, but they spent too much money. But if the prize is big enough, I mean, we're investing in nuclear fusion, which won't have a real commercial reactor till the full 2030s. If the prize is big enough, the timeline is fine to be 10 years, 15 years, 20 years, but you need to spend against that timeline. So you can't spend as if you're going to be a commercial entity in 2 years if it's going to take you 10 years. Which should be doing this. And, and remember, that's another reason why we like these— some of these companies is that if the market is huge and the technology advantage is big enough, you can wait. You don't need to rush to market while you're developing the technology. The burn rate, which is the most critical thing, is very controlled. $250,000 per year, fully loaded, per engineer. That's not going to change. That's what your burn rate is. What's unpredictable is when you go to market. So how many salespeople do you need? What's the quota? How long does it take them to spin up a salesperson? What's the sales cycle? What's the cost of customer acquisition? What's the retention rate? Those are all only knowable once you're in market. So me personally, I'd rather wait 2 years to enter a market with a clearly superior product. Versus a product that's just slightly better, where I'm taking all of that unknowable risk, because I'll take the knowable risk of what the burn rate is going to be. And so if something is delayed, which it almost always is, I know, okay, I have 100 engineers working on it. That's $25 million. When I raise money, when I look at my balance sheet and say, what's the minimum threshold I can hit before I need to put more money on the balance sheet? It's— I know what the costs are. Where it starts to get scary is, you know, like people say, oh, well, you know, clean tech, it's super capital intensive. I mean, how much money have enterprise software companies raised? I mean, look at the history of Salesforce, Facebook, companies have raised— Snowflake— billions of dollars for markets that aren't nearly as big as electricity or electric vehicles or, you know, these things. The difference is you have this false sense of comfort that you don't need to raise that much money to get to $1 million or $5 million. And that's true. But your upside is also much lower for a CRM company or HR tech company. And most likely you're going to get acquired, which is fine. But the TAM for nuclear fusion or plant-based foods or electric vehicles are just massive. And so you can afford to wait until you got it right.

Right. And that's one of the things that people have struggled with when they're thinking about these type of companies, which, yes, there's a lot of technical risk, but the TAMs are big, the commercial application is big.

If they can solve and de-risk on the technical side, you have a natural moat because the amount of money and the time, you know, your natural moat, it's not like someone coming quickly and like Clubhouse or Cameo or some of these things where they get a lot of hype, they get a lot of usage, and then next day people say, "I'm kind of bored of it." BeReal. I mean, these companies, they get super hypey, super usage, and then the usage falls off versus look at a Tesla or these kind of bigger markets. Once the company's there, people stuck with it.

Let's shift a little bit to the investing side. And in particular, when you look at these long-term trends that you have high conviction AI is a great example, and I think most people would agree that it at least has the potential to exceed some of the big tech supercycles of the past. And yet when that happens, you have a lot of momentum, you have a lot of dollars flow in, and ultimately things like valuations creep up to a point where they lack fundamental metrics in terms of what usually would make sense for a company generating either none or some small amount of revenue. And you have to reconcile that, of course, with your own investing thesis and what you look for from an ownership standpoint, the valuations you think allow for enough upside. How do you navigate that? And are there times where you make exceptions to your normal parameters of investing?

We made an exception with Replit. We invested at a big valuation at Replit with Amjad because Amjad's an exceptional entrepreneur and we believe in what he's doing. What he's building is critical. We like his go-to-market experience. Codecademy makes a lot of sense for his initial go-to-market. We fell in love with him as an entrepreneur. There are a dozen or so other companies, similar valuation, similar traction that we passed on because we didn't have the same level of conviction.

So it's a level of discipline but not rigidity and an understanding of when you need to stretch the boundaries for the right individual. So it does sound that it's very much based on the individual and the founder and the opportunity.

How big is the upside? Big enough to make an exception. And that's one of the things about us, as you know, Samir, is that we don't have a veto in our shop. So if any of the 4 MDs want to do an investment, they can do it. And there's a real— but look, if someone makes a lot of bad investments, they get fired. We're not scared to do that. But this business is all about power law, which means you have to shoot for asymmetric upside. To get asymmetric upside at this stage, we invest in very early. There's no fucking way you're going to get consensus. There's no way. If you're looking for consensus, you're going to, by definition, take away beta. If you take away beta, you take away asymmetric upside. OpenAI should return that fund 3 or 4 times over just by itself. You know, but it was a $50 million check into a nonprofit entity.

What for?

And the reason why the only VCs in there is, right, if other firms that have consensus, there's no way they would have done that deal. That would have made some jackass would have said no, you know, pretending to be super smart. But I don't pretend, you know, we don't pretend that any of us has a perfect lens on these things. So we vigorously— and it allows us to really vigorously debate companies and we can say, look, And we rate companies. We don't just make it qualitative. We score every company on a scale of 1 to 4, where at any other firm, a 4 would be a veto. But it gives you something to think about. I can't tell you the number of people we interview who are like, oh man, I would've done Square, Stripe, and Airbnb, but I couldn't do it. I couldn't get it through my partnership. I'm like, great. I'd love to be 6 foot tall too. Run like Usain Bolt. It's just nonsense. So I could bring in the next square, everyone could be a 4, and I could choose to pass and then it becomes square. That's all on me. I didn't have the courage and conviction in my investment. So it's really important, I think, in a business where you're rewarded by taking asymmetric risk to not be bogged down by consensus that isn't going to happen.

Everyone knows, I mean, I think you and I have been around long enough where you've seen these firms where there's also politicking going on at the GP level. It's like, you vote for my deal, I'll vote for your deal.

Oh, and then the person is asleep at the switch and not helpful. For our firm, David Swen, Vinod, and I all add different value. And if I didn't think they added value, why would I give them any of my carry? And if I'm giving them my carry, they better be awake to the presentation and help and make it, make it happen. But they've already kind of— if you've horse traded in the morning before your Monday meeting, hey, vote for mine, I'll vote for yours, then you're going to be on your iPhone screwing around. I don't know what you're doing. And then also the entrepreneur gets hosed because the entrepreneur doesn't get the benefit of the tough questions. It doesn't get the benefit of the whole help of the partnership. You know, better or worse, if you work with KB, you're working with all of us. Nobody has deal-specific, Gary. You know what, vigorous debate and people have different opinions, but once they make an investment, everyone is working for that company. It's not, you know, we've had people here who's not lasted who would rather be right. And when they didn't like a company, they'd root for it to fail versus be successful. And we weed that crap out. We don't have silos. Oftentimes multiple of us work on a company. Sometimes we rotate in and out. We have no deal attribution, as you know, because you've been an LP. If you don't give deal attribution, you don't— then it doesn't become political, like, oh, that's my deal, I get to work on it. All right, your people are fighting for credit on who got the referral. None of that happens here.

You know, it's such probably a cultural shift for people that you might bring in from other organizations that are used to more or consensus. How do you get people comfortable? Because it's fine for you or David or Sven or Vinod to make a non-consensus decision, do a deal that you have, even if you're getting vigorous debate or contra reasons not to do a deal.

[Speaker:JOSH_FARLEY] But they're not an FD then. FD has to sponsor the deal. So they have to work with one of us that hopefully they— Sven got promoted, is our most recent FD and he got promoted, but he worked with us as a partner where he couldn't do a deal without one of us giving him air cover. And that's really important. One of the main reasons— there are a lot of reasons why people don't work out here and we're very quick on the draw that if someone doesn't work out, we part ways in a friendly way, but part ways because you're only as good as the worst investor at the table, right? So what happens— this is another insight— is in a firm where there isn't whether it's consensus-driven, the quality of due diligence also becomes worse. Because if you're, if you're an investor and you fall in love with an entrepreneur, you're now selling the investment to your partner. And what that means is that you're looking for all the positive attributes and not the negative. When I fall in love with an entrepreneur, I want to do the deal. I'll bring it to my partners. I know I could do it right. So I'm not selling my partner. I'd say, here's why I'm excited. Here are the risks I've identified. Here's why I think these risks are manageable with X amount of money. What do you guys think? You know, what am I missing? Help me out. Help me identify risks I haven't identified. Introduce me to someone that would say, hey, that risk really isn't achievable with this round. You need more money for that. Or introduce me to someone that's done a similar type of— whether it be customer acquisition or solving some material problem or something at another company. So you're going to people with help as opposed to a sales pitch.

Right.

And that would show. So my due diligence is finding where is this company going to fail? How can I figure that out as quickly and cheaply as possible? And then is that worth the risk? Is the reward worth running that by my partners, seeing if I'm thinking about it the right way as opposed to giving them a sales job where I'm not whether intentionally or nefariously or not, just human nature is such that you probably aren't highlighting the negative things that people are saying. You could also— the other firms, you're fighting for allocation. There's a fund that's X size divided by X number of partners. You're all fighting for dollars. So if you— and if you may not really like the other person because they kind of shit on your previous deal, you're gonna look for a reason to do it. So that person's been I just put on the bravest face and was the honest, hey, here's where I could— I remember one of our most successful investments is a public company called Garden Health. We made nearly $1 billion on it. And when I was doing the diligence, one of the world's most famous cancer biologists shit all over the company that this ain't— this is nonsense. These people are nonsense. And I still remember in my investment memo, I put all the reasons why I was excited. And I said, oh, by the way, this particular person, who now is a huge fan of the field. And it's kind of comedy now when I see him. He's a big, big name in the cancer biology field. He was super negative. I was insane for doing this. I wrote up exactly what he wrote and I highlighted it in the memo. I didn't hide it in the corner. I said, I'm still a one. I'm going to do this investment, but I want everyone to know this guy who knows more apparently than everyone else, and wasn't shy about telling me how smart he was, said this, and I've highlighted it. So if this thing fails, you can't come to me and say, oh, well, what about this? Put it right there. But you would never do that at a firm where I had to get consensus. You would bury that.

It's a great story, and you're absolutely right. It would be buried because it would make it dead on arrival in a consensus firm. Speaking about building a firm around this ethos where people are willing to put their neck out there and make decisions on deals where they have high conviction despite what others may think around the partnership. You also have to get the type of people that have that type of DNA that can do that, and it's not easy to do. Are there things that you do as part of the interview process for potential Khosla principals or other investing staff that allows you to determine What kind of person is going to be a successful Khosla investor?

It's just super hard to find people. You're right. It's very hard to get sort of partner-level people once they've been indoctrinated in another shop. We're so unique to come over to our shop. Where we've had, like, Swen was actually joined as a technical expert and then tried his hand at investing and done great. The next-gen people actually have worked in our portfolio companies. Us. We got to seeing them for years and years. In the interview process, you do as many references, you spend as much time as you can. And usually I'll have my 3 or 4 favorite CEOs spend time with that person and get me a read on would they want them helping them? Do they think they know me pretty well? Do I— do they think I'd get along? And you're still wrong at least 30, 40, 50% of the time, even doing all that.

Yeah, it is a difficult process. So I want to maybe end with a kind of look-back question. Now you've been in venture for a quarter of a century now almost. Khosla, from the founding days to where it is now, 20 years. If there was a single piece of advice you'd give yourself from 20 years ago when it comes to investing, running a firm, what would it be?

You can only lose 1x your money and you can make an infinite amount of money. So if a company is going down, be polite, be generous, be empathetic, but don't waste a lot of time trying to get a 70 cents on the dollar. Take that time, take that money, and put it into the winners because the power law always wins. And early on in my career, I fretted so much about losing money. That was just a waste of time and effort. If I'd taken that time, I could have maybe been in a company that was an 8x, 10x. That's much more effective use of time than getting 40 cents versus 10 cents on the dollar.

It's hard for a lot of people to actually understand that. And I've looked at a lot of data on highly successful funds as well as funds that haven't done that well. And there was a high correlation of funds that did really well in terms of cash-on-cash returns, IRR, where the loss rates were actually really high. And in those cases, it was almost this power law where you had one or two companies really drive the preponderance returns, but it didn't correlate to the actual loss rates.

That's our fun. It's our fun. Because you don't take— if you don't fail a lot, you're not taking enough risk. You're not going to get that upside. What we get paid to do, especially now, you know, with interest rates where they are now, risk-free rate, you could make 6, 7, 8% pretty easily. You know, that means we need to make 25 to 30% net of fees and carry to be worth the risk. If you can't, you can't do that without at least one or two companies providing a multiple of the fund return. It doesn't matter if you do a 2x fund, but bat 75% or bat 25% and give us a 5x fund. It's very clear what people want.

This has been a great conversation. I really enjoyed the insights. Congrats on the new set of funds and being able to deploy Dry Powder in a time where I think we're in this great intersection where we've seen the macro dislocation with what I think is going to be the next supercycle of technology with things like AI. So again, thanks for coming on.

Thanks, Samir, for your time. Really enjoyed it.

Thanks so much for listening to another episode of Venture Unlocked. We really hope you enjoyed it. To learn more about Samir or Coastal Ventures, be sure to go to the Venture Unlocked Substack at ventureunlock.substack.com, where you'll find detailed notes of the show and a listing of the past episodes, along with other commentary I have on venture capital. You'll also find us on Apple or Spotify, where you can subscribe to get all of the latest shows as soon as they're released.