Mitchell Green: Why 50% of VCs Should Not Exist

ByteDance is the most advanced AI company in the world. It’s very underappreciated by the Western world.

Today we have Mitchell Green at Lead Edge Capital joining us. In a world of fluff and framework-thinking investors, Mitchell Green is a moneymaker. Mitchell has co-led or led investments in insane companies like Alibaba, Benchling, ByteDance, and Grafana, among many others.

I think 50% of people in the venture business should not actually be in the business. There’s too much money and too many tourists. 50–60% of people in this industry probably add negative value to companies. People who spin out of Anthropic or OpenAI and raise money at $2 billion for a freaking idea—there’s nothing more than an idea on a napkin. To us, that seems like complete lunacy.

There’s going to be a really big downturn. Markets just don’t go up forever, and I think it’s going to happen in the next 10 years. Buying is glamorous. Selling is the job. Don’t count China out. I bet they win the AI world.

Mitchell, dude, it is so good to have you back in the studio. I love doing these with you and Larry; they’re my favorites. You’ve got to have us on together.

Do you know what would be great? If you’re in London together, we should do it over dinner and mic everyone up.

That would be absolutely amazing.

Listen, I want to start with something that’s actually quite disarming for a lot of investors today, which is, bluntly, the SaaS apocalypse. The SaaS stocks are in the shit, and we’re looking at the markets. There are a lot of people questioning whether they’re actually good investors or whether we were just in a buoyant cycle. Is this downturn justified, or is this an overreaction to AI and Anthropic product releases?

We are buyers. We’re buying top SaaS right now. A portion of our funds can be invested in public equities, so we’re buying companies like Procore, Workday, and Appian. We love Clearwater Analytics, but it’s in the process of being taken private, so the stock doesn’t move.

We’re big investors in Toast, which we’ve been buying back. We were early investors in it, sold, and are rebuying. These companies aren’t going anywhere. The incumbents have distribution, data, and balance sheets. It’s a fool’s errand to think all these companies are going away.

That being said, in any period when there are big periods of disruption, there will be new companies that are created. There will be incumbents that thrive and adapt, and there will be some incumbents that blow up.

Help me understand. I love your perspective, but I don’t understand it. Workday is at 6.8% growth. We’re seeing the cannibalization of the seat model, and we’re seeing, bluntly, no impressive use of any agent products within the existing incumbent set.

Workday’s AI business is growing super fast inside it. Keep in mind, they have $10 billion of revenue and about $3 billion of free cash flow. There is a law of large numbers. For all these companies, look, it is not normal for companies to grow like Anthropic or OpenAI have.

Oh, and by the way, Workday does it with serious profits. A company like ByteDance grows at 30% a year with massive profits. Let’s see how these companies would grow if they had profits.

For seat-based pricing at Workday, Workday is more tied to employment and employment growth in the United States. The company is a high-single-digit, low-teens grower. But again, it is a huge business. It’s a $10 billion revenue company.

I actually think what people got wrong, and why a lot of these software stocks have sold off in hindsight, is that if you looked at the end of last year, Street numbers were too high for this year in general. They didn’t show enough deceleration. You’d have a company that grew 20% last year, and the Street thought it would grow 19.5% this year. It’s the law of large numbers. As companies get really big, they decelerate.

Street numbers across software in general were too high. What will happen is that Wall Street analysts and sell-side analysts are like pigeons, like squirrels: the stock goes down, and they say, “Now we’ve got to lower our numbers and take the numbers down.”

A lot of big public hedge funds and public-market investors don’t want to own stocks when numbers need to come down. You want to own them when numbers are about to go up and estimates are about to go up. You’re going to see them take numbers down at a bunch of companies. Give it a quarter or two, and the companies will start to beat numbers. Then they’ll raise the numbers, and the stocks will start to work. But they’ll probably be dead money for a little while.

I was always a big fan of Howard Marks and his investor letters. One of his big things is, “Never try to catch a falling knife.”

I see what you see in terms of the opportunity, but I have no idea where this is going to go. Honestly, a month ago I was looking at Duolingo thinking, “Wow, what a buying opportunity.” The lesson I have on Duolingo is, “Wow, there really is no floor.”

If you don’t have earnings or EBITDA, there is no floor in a lot of these things. What we tell people is, if you want to own them, buy them over a month-long period or buy them on down days.

If you’re an individual and you think you want to own $1 million or $200,000 of some name, buy $50,000 every time it dips and sells off hard. Maybe you’ll never get fully filled, but you won’t catch a bottom.

The funny thing is, if you actually do the long-term analysis of buying just indexes like the Nasdaq or the S&P, it turns out, if you look at very long-term longitudinal data, that if you just buy on a big down day, it’s nearly impossible to try to time it.

We mentioned Workday. We’ve also recently seen Aneel Bhusri, the founder, coming back to Workday. Not specifically about Workday, but I’m unwaveringly negative on companies where the founder is not the CEO and we’re in this AI transformation. Do you share the view that non-founder-led companies are inherently disadvantaged?

I would agree with that partially. I do think there are very good CEOs. You need a growth mindset. I believe there are companies that are run for growth and companies that are run for margins.

Any time you have big technological transformations, you want the entrepreneur or the management team running the company to be focused on growth. They’re growing. By the way, those are oftentimes entrepreneurs.

Another way to think about it is that companies run for margins, earnings, or EBITDA margin are oftentimes heavily levered. I think the biggest opportunity to disrupt incumbents today is in software and tech-enabled services. It can be a manufacturing company; it doesn’t matter. Any company with a bunch of leverage on it—those companies don’t have the cash flow to innovate.

You can look at 1999 and 2000 and see what happened. If we had sat here in 1999, we would have debated whether all the traditional retailers were going to go bust and whether all these e-commerce companies were going to be gigantic.

If you look today at the 10 largest e-commerce companies in the United States, 6 or 7 out of 10 are traditional retailers. They’re companies that were well on their way before 1999: Walmart, Target, Home Depot, and Lowe’s.

However, there were a bunch that went bust, too: Sears, Kmart, Montgomery Ward, and Bed Bath & Beyond. You have to ask yourself why. Most of those companies had huge amounts of leverage, so they couldn’t innovate.

Walmart didn’t have leverage. They were saying, “We’re going all in. We’re going to bet the company on this stuff.” I think you’re going to see the same thing today.

There are many things I want to touch on. You said you run a company for growth or you run a company for margins. If we take, like, a Matt and a Zach, he’s running it for growth, which is why free cash flow is in the drain. Now I think it’s valued at something like 1,100× free cash flow, and he’s being pummeled for it.

Is it right to pummel him for it, or is he in the right mindset for growth?

This is the million-dollar question. Is Apple right, or are Google, Microsoft, and Meta right? Apple is spending very little right now, and those other 3 or 4 companies are spending insane amounts of money. Time will tell. We don’t know. It’s probably somewhere in between; that’s probably the right answer.

I would argue, though, that if Mark Zuckerberg deserves to go for it, it’s his business. He built the damn business. I don’t think you really have much to say and tell him, “Don’t bet on the guy.” By the way, he kind of has to, because his competitors are doing the same thing.

Yeah.

So, look, our view is that ByteDance is the most advanced AI company in the world, and it's very underappreciated by the Western world. How much AI they use and how much they're investing in it is underappreciated. Our view is that AI is going to change the world. It's an incredible thing.

Again, we sat here in '99, and the words “social media” didn't show up. We would never have talked about Facebook or anything, right? It's $3 trillion in value now. AI is not going to be about the next call-center company or the next Workday.

I truly believe what we're seeing right now, with people investing in a lot of these AI companies across the board—some of them are going to be gigantic, and a whole bunch of them are going to bust out or bust—but I actually think it's the stuff that's going to start over the next 2 to 5 years. Those are going to be the giant businesses. And I don't even know what it is.

We're going to get outspent by dollars, but do you agree with the “play the game on the field” analogy, or do you actually think there are such shifting sands that the optimal strategy is to be conservative, not invest a ton right now given the transience of the markets, and sit and wait for some form of new equilibrium to emerge?

That's a great question. I think it depends on what business you're in. If you are an early-stage venture fund where returns are made at 100Xs or zeros, you should be investing in stuff now. I think you should always ask, “If I make an investment and it grows for 18 months and it hits my numbers, am I now in the money?”

The problem is, when you can invest at 100 times revenues or something, it grew 18 times. You can grow at some crazy rate for 18 months and you're like, “Well, I'm still lower,” and you're in the money. So, we always like to ask ourselves that.

But if you're trying—we're in the business of trying to make 2 to 5 times our money in 3 to 7 years for a 25% IRR—and we don't drive zeros, we also don't have 20Xs. I think we've had 2 10Xs ever or something like that, but we've only had 1 or 2 zeros ever. So, this environment for us is just weird. It's different.

Are you finding it hard?

It is definitely harder to invest today than it was in 2017, for sure. Although there are different pockets of opportunity. The secondary market for us right now is exploding. In our special-situations business, we just did a deal in a company in a special situation and put $200 million to work, and the company is now raising a round at 2X the price at which we invested. It literally happened a month ago.

I don't understand. When you look at a lot of the growth-equity investments that you and a lot of people have made in recent years, are they not a lot more vulnerable in the new environment we're in? The well-priced company up north in the UK that's doing accounting, whatever.

Look, one of our companies is Grafana Labs. It's a giant business growing crazy fast, and it's benefiting from a lot of this AI spend. Its customers are some of these big AI companies.

But I would say that's a straight-down-the-fairway Silicon Valley deal. It's with Sequoia, and it's like—

When we invested, nobody knew what the company was. It was a bootstrapped, $12 million software company. We and Lightspeed were the first 2 investors. It was bootstrapped. The guy actually built a $12 million company.

How did you find it?

We cold-called him—the CEO.

You cold-called him?

Cold-called. So, we have a team of 18 22- to 24-year-olds pounding the phones, calling companies all day long.

Learned from Insight.

Learned from Insight. Yeah, so, by the way, they just replicated what Summit and T.A. did.

By the way, I just had Jerry Murdock on the show.

Yeah, yeah, he's fantastic.

So, if a company calls you back, it's like, hang up the phone. It's the CEO you call every 2 days for a month—that's who you want to get on the phone.

And look, people are building amazing companies. We have a business down in Florida called PaceMate. It makes cardiac-monitoring software. If you put a pacemaker or a defibrillator in your body, it takes the data off the device. That data is then sent to the manufacturer's website. There are lots of different manufacturers and models. This is a single pane of glass for cardiac clinics.

By the way, it is a 99% gross-dollar-retention business. When we first invested, these guys were at $20 million in revenue. It had only raised $8 million to get there but had only burned $3 million. It was growing at 50% a year. I don't know; it probably did $45 million in revenue last year, a few years after we invested.

We're using AI to our benefit. We have a huge number of people in the call center or in customer service analyzing the data, making sure that all this stuff is working. The company can now continue to grow without adding people. I mean, they don't have to—they're not going to fire all these people—but they can keep the same number of people and make them much more productive.

And I think that's what people are missing. I think there are 2 things. One, this is going to lead to a giant productivity boom. Two, software companies have never been about R&D. This is not semiconductor investing; it's very different.

If you look at your average software company that goes public and you look at cumulative spend since inception, it's usually around 30% R&D. A huge amount of these businesses are about sales and marketing, distribution, customer support, and things like that. AI will help with a lot of those things.

The only way that we might see—I'm sure you read the Citron Research piece. We saw it come out and wipe billions off the stock market, essentially saying that it—

That's the more incredible thing.

What?

That some random person can write a research report. Imagine if we had Twitter in 2008. It's incredible. It's amazing that people are listening to some random research firm versus listening to people like Stan Druckenmiller, Howard Marks, Ken Griffin, Steve Cohen, Marc Benioff, people like Mark Zuckerberg, and Jensen Huang, who's like, “Software is not dead at all.” A random research report can get 25 million views and tank the stock market. It's crazy.

Is this not the ultimate sign that we're seeing the casinoization of public markets?

Yes.

Which actually makes it, by the way, the opportunity for long-term investors.

You buy, as Warren Buffett said, when people are scared. You'll be able to make lots of money.

Or the flip side: you don't want to take part in entirely irrational markets, which are no longer tied to value.

Look, I was working at a hedge fund seeded by Julian Robertson in 2008 and 2009. This is nothing like this. This is amateur hour. This is nothing. This isn't even volatile compared to what was going on back then. The market was whipping—the indexes were whipping up and down, with 8% intraday moves. You'd have 15% swings.

But that presents the opportunity, by the way. Had you bought in early '09 or late '08 and bought great companies, you could make a ton of money. This is the opportunity to buy stuff on sale. Now, again, you—

Okay, I'm going to push back on you there and say there was no fundamental technology inflection point in 2008 that could render an incumbent set relatively redundant, whereas there is today.

The mainframe business is still a $5.5 billion market. Mainframes came out in 1950. Most banks are run off mainframes. Oracle is a legacy software company. Microsoft is a legacy software company. SAP is a legacy software company. They're some of the biggest software companies in the world. These companies are not going away. I will bet any amount of money on it.

There will be some that will focus on the companies that have 90%, 95%, or 98% gross-dollar retention. There will be new giant companies created, 100%. But most of these incumbents will not disappear. Some of them will innovate and become exponentially bigger. Some of them will grow 5% to 10% a year. There are a ton of software companies that have been around 20 or 30 years that are still growing.

I actually think the biggest disruption you're going to see isn't in manufacturing, isn't in health care. Think about the companies that can figure out how to get drugs to market much faster than anybody else. I think AI could potentially solve huge parts of cancer. Dementia could be solved because you can run drug trials faster.

Think about manufacturing. If AI and robotics can come about, if you have 2 competing companies that make cars, one's levered and one's not levered, the one that's not levered is probably going to be able to invest a lot of resources. It's going to be hugely beneficial for them.

Totally got that. Going back to what we were just talking about—the casinoization and how crazy it is...

You mentioned productivity increases.

Yeah.

Are you worried that we’ll see productivity increases but, with that, fewer and fewer consumers having jobs and a weakening of consumer wallets?

Not really. There were a couple of million switchboard operators in 1980. There have been lots of jobs that have been lost over the years. Think about the number of retailers over the years that have gone bust. People innovate.

It’s funny—I was just talking to somebody at one of the world’s largest banks, a very senior person at one of the world’s largest banks, this week in London. His point was, “Look, we have hundreds of thousands of people in back and middle office. We’ve trained them for 5 to 25 or 30 years, right? We’re not getting rid of them all. We’re going to retrain them.”

And by the way, the people that don’t want to be retrained will be told, “Okay, fine, go work for the government then,” because they can do old-school jobs there. A lot of these companies will retrain people and have them do different things.

It’s remarkable. Throughout history, there’s been lots of technological disruption over the last 100 years, and people find new things. If everybody’s worried that everybody’s going to lose their job, you shouldn’t invest in any of these companies, because it would be a complete disaster.

I guess I’d tell people that if you’re worried about China invading Taiwan, you really shouldn’t worry about your ByteDance position, because you’re going to have a lot bigger things to worry about. At some point, by the way, the government would get very involved. If all of a sudden all these jobs start to disappear and you have 10, 20, 30% unemployment, it’s not happening.

First of all, people always think this change comes faster than it does. Most big companies that are financially regulated—you can’t even go on to Claude or ChatGPT. You can’t even get on the system to do work. And by the way, we’re still 5 years in.

It’s literally like people woke up a month and a half ago and were like, “Oh my gosh, all these companies are going to go away and unemployment’s going to 30%.” It’s nonsense. It’s silly.

We’re going to sit here in 10 years and it’s going to be an amazing time to invest. You’re going to have a bunch of amazing new companies being created. You’ll have some legacy companies that went to zero—not all of them, or anywhere near all of them. You’re still going to go to retailers. You’re still going to get your hair cut.

I just think it’s going to be an amazing time to invest. And by the way, during periods of volatility—and the casino is a part of the stock market—you want to buy fundamentally good businesses on multiples of earnings.

Do you worry that the world is just being memed? That the world is being casinoized? Is the world just a casino, with Polymarket as its replica?

It’s totally insane. Look, I mean, I’ve joked for years that social media is the demise of society. Jonathan Haidt has done amazing research on this stuff. It needs to be regulated. It will be regulated.

It’s incredible how fast information moves, right? Just imagine if Twitter had been around in 2000 and had been big in 2008 and 2009. It’s absolutely incredible.

The makeup of the stock market is different today than it was 20 years ago. Passive ETFs are much bigger. The retail stuff comes and goes, like retail trading.

You just have to buy good businesses when they’re on sale. Buy good businesses at multiples of fundamental earnings. If you don’t have earnings, there is no floor. But if you have earnings and free cash flow, then you have something to work with.

Another thing that people reason a lot of these internet stocks are still not cheap is because the stock-based comp in a lot of these companies is totally nuts. The amount of equity compensation and dilution for shareholders is very high. I’m surprised more people don’t talk about it.

Why are we not talking about it? What do we not know that we should know? Are we in a new norm for the Snaps and OpenAIs of the world to have unreasonably high SBC and no one question it?

Some of the big public-market investors have been questioning it for a while. It’s surprising that more people don’t talk about it. Just look at how much stock-option dilution there is in a bunch of these big Silicon Valley companies. It’s not as bad outside of Silicon Valley, but the dilution is real. I’m surprised more people don’t talk about it.

What should happen? If you’re a Snap holder, Evan is running a gifting program right now.

Yeah, I don’t know. We’re not activist shareholders. I’m not an activist at all.

I have a lot of respect for entrepreneurs like Larry Ellison, who effectively did a levered recapitalization of Oracle. He basically said, “I have all this free cash flow. I’m going to borrow debt and buy back an enormous amount of stock.” What did he do in the process? He didn’t sell any of his own. He just kept making the company go down, not up.

People forget that in companies, market cap equals the number of shares times the price of the shares. Companies that respect that and have discipline on that are powerful.

We’ve seen so many well-known names being in the dumps, and a lot of people are questioning, “Are the CEOs and the management teams buying when they’re in the dumps?” The ServiceNow CEO bought $3 million worth.

I think Salesforce just came out and said they’re going to buy $50 billion of stock, or some crazy number.

Was that from the earnings report last night?

No, I agree with you. By the way, companies should be buying back stock. Those that aren’t buying back stock, you should question and ask why.

Okay, but ServiceNow’s CEO bought $3 million worth, which is less than his car collection.

Yeah. I don’t know if the company is buying a lot of stock back. I don’t know. But you should also ask how much stock the CEO of ServiceNow already owns. Does he own $300 million of stock? That should tell you something.

The companies where the founders are buying, or the companies are buying huge amounts of stock back, would make us more bullish on that company versus another company. One hundred percent.

I would suspect that over the next year, as earnings come through, you will see more things like Salesforce putting in place big buyback programs to start buying back stock here. I think you’ll see it.

We spoke about ByteDance a little bit.

Mm-hmm.

Everyone for years has been saying, “The ByteDance discount—it’s so cheap.” ByteDance is insane because there’s the China discount, and ByteDance is the China discount on the China discount. That sounds great, but it’s only good for you as an investor if that discount chasm shrinks. What’s it going to take for the discount chasm to shrink?

I think it already is, to some degree. Private-market implied valuation multiples should be determined by public-market investments. You should argue that a software company today should be getting done in the private markets cheaper than in the public markets. It definitely doesn’t always occur like that.

Alibaba and Tencent are 2 giant Chinese companies that should represent roughly how ByteDance should trade. I haven’t looked at them in the last couple of weeks or month. They were trading at mid-teens earnings multiples—not EBITDA, not revenue, earnings multiples—and they don’t really grow.

This is a business that grows fast, 25 to 30% or more a year, generates a tremendous amount of earnings, and you can put an earnings multiple on this company and get to a very, very big number based on fundamental earnings.

I think Facebook is trading at 25 to 30 times earnings right now. Put that multiple on it. But let me ask: maybe that’s too high because it’s a Western company. So then put the China-company multiples on it.

I think it is possible to see, in a few years, that this company is doing $70 billion, $80 billion, or $100 billion of earnings in the next 5 years.

Yeah, but in a deglobalized Trump world, I’m just making the alternate argument here to understand: in a deglobalized Trump world, it’s not going to list in the U.S., is it?

Zero percent chance it’ll list in the U.S. No, I mean, there’s a chance—I have no clue—but no, it’ll list in Hong Kong. That’s where a lot of these Chinese companies list.

People have been talking for years about how they’re going to delist all these Chinese companies—Baidu, Ctrip, Alibaba. They never did.

Keep in mind, Alibaba’s stock has roughly doubled off the lows over the last year. Sentiment on China today is a lot better than it was 18 months ago, when we were buying ByteDance stock. We were buying it at prices below $200. We thought the risk-adjusted reward, given the earnings power, was incredible.

What was cool?

Don't count China out. I bet they win the AI war. I bet they win it. That might not be right, but look, the great thing in China is you can build a nuclear power plant in a couple of years. You can build power plants, no problem. In the U.S., we're going to run into major issues around power.

Why do you think they win the AI war?

Because of the power, resources, consumption, the number of PhDs, and how much they value science and technology. I think power is always a real advantage. And look, there are things that could totally change it. Nobody's really talking about quantum. We're not quantum experts at all, but is that something that could make these things exponentially more efficient?

How does that realization change how you invest? I completely hear you, and I agree. I think we still dramatically underestimate the capability of China, or just choose not to think about it or push it to one side. But if that is the realization, how does that impact your go-forward mindset on investing?

Well, we own a lot of ByteDance. But it's not winner-take-all. It's not that ByteDance wins and Google and Facebook lose.

A year ago, when I was probably on the show—I don't know if we talked about Google—everybody thought Google was going to lose. “Google's dead. It's done. Nobody's going to search it.” Have you seen the stock in the last year? The stock has doubled. Now it's going to win everything. Now it's going to beat OpenAI and all these other companies. No, they're both going to be fine.

The biggest question for us about these LLM model companies is whether they can ever turn a real profit. I just don't know the answer, and I don't think anybody really does right now.

Again, I think another thing you're going to see in the United States, as it relates back to power—and you really haven't seen much of it yet—is local communities getting really upset. You're the small local town in Iowa, or the small town in Kansas, Ohio, or wherever in Virginia. They built this giant data center. They employed all your people. They employed a ton of people locally to build it. They then built it, and now it sits there and has 50 local people who work there, while your local power prices have tripled. Is it polluting the environment? It's this big, ugly building.

I think you're going to see real local pushback, and I don't think it's only in the United States. I think it's probably Europe as well. These people are not better off today than they were 20 years ago, and these things are in their backyard while people in Silicon Valley and on the coasts are making tens of billions or hundreds of billions of dollars off of them. I think there's going to be real pushback, and there needs to be regulation around it.

Do you not think climate is a luxury problem? We were also worried about it in the last 3 to 5 years, and now no one gives a shit about it.

We should be worried about it.

Climate investing is in the drains.

We should be worried about it. It's probably important. We've always struggled with how to invest there because of how capital-inefficient a lot of those businesses are. China has some advantages in that respect, whether it's the giant solar farms they can build or whether they can just do things that we can't do.

If you look at the internet, the biggest innovation in the internet over the last decade or 15 years has usually come out of China. If you actually want to look at where e-commerce and social media are going, go look at China. It was not a surprise to me when DeepSeek came out, but don't underestimate Chinese creativeness and individuality to figure out how to reverse-engineer and engineer things in much cheaper ways than Americans can.

Can I ask you, going back—not specifically to the ByteDance play—but I really want your advice on the sell side. When you look at a company like ByteDance, there are many opportunities to sell in a lot of these names. Not taking ByteDance specifically, how do you think about it? We're 3x up on where we are, we've been in it for 4 years, so let's take some chips off the table. How do you think about sizing positions over time, and have you got any big lessons or advice for me on that?

Buying is glamorous; selling is the job. Constantly re-underwrite. That is actually what it really is. We're trying to make 2 to 5x in 3 to 7 years. If you put that on a 25% IRR curve, put it into a fund, and make it 2 to 2.5x net to the fund, that's what we're trying to do. That's what we tell our investors.

We're constantly re-underwriting and saying, “Okay, if we were going to sell a bunch of ByteDance today at $550 billion, which is where it's reported that General Atlantic is selling a bunch—and there are other people—we've been offered higher than that.” I think it's always: What is the probability it can double? At ByteDance, we look at what the fundamental earnings are, and we're like, “Okay, this is doubling, no problem.”

Now, if somebody came to us today and said, “Hey, I'll offer you $1.3 trillion,” we'd sell a bunch. It's not that I don't think the company will do $100 billion of earnings in the next 5 years; I think it will. At 20 times, that's worth $2 trillion. But there's a risk. What does it trade out at on a multiple basis? What would we be on our total investment at that price?

Sure, and how far ahead are you paying for growth? At what point is that a really valuable moment to go, “Yes, you're paying 4 years out”?

Oh, yes, correct. What we like to say—and I think this has actually gotten people in a lot of trouble, too—is, “Are we in the money 18 months out?” That's what we think. With a reasonable multiple, are we in the money 18 months out?

Can you unpack that? What do you really mean by that?

If we invest today and revenues are $20 million, with a reasonable model—

Yeah.

At 18 months out, are we in the money? With a reasonable multiple, not 50 times revenue.

But do you buy companies like the pacemaker company? I didn't want to pick on them so much. $20 million to $45 million, which is great, phenomenal, fantastic. But who's going to buy that company?

I mean, there's a line of strategics that would buy it. There are private equity firms that would buy it.

At a good multiple?

Yeah, we'd make great money. By the way, we forgot—what did I pay for it? A good investment and a good company are 2 very fundamentally different things, too. You're trying to get the union of both of them. There are a lot of A+ companies at D- prices. Nor do I want to buy a D- company at an A+ price.

Do you not want to make that investment? I'm always told by most people on the show that I've never made money with a good deal.

I would strongly disagree with that statement. There are a lot of people that invested in good companies—great companies—in 2020 and 2021 at really stupid prices and didn't make money. It's the intersection of both. You're not trying to buy a D asset at an A+ price. That's like zero.

By the way, we also buy A+ companies. The great thing is, can you do structured secondaries, or can you do really unique things like some of the stuff Larry will do and we'll do? Can you buy an A+ company or an A company at a B-/C+ price? Those are incredible deals.

Can you get in cheap by buying out old LPs from an old fund that need liquidity? By the way, I joke that we're 1 NVIDIA earnings miss away from a recession. If you get that, a bunch of old funds are going to be like—you know, they have horrible DPIs. You'll have all these LPs that are like, “You know what? I'll sell an interest in my 2013 fund or this 2015 fund and be able to buy stuff super cheap.”

But look, there are also different beliefs. There are a lot of people in the venture business who are just like, “I will pay anything because I'm trying to get the power law and I'm trying to find the next Google.”

Would you not advise me, though, that when you have a world where upside is relatively uncapped and you have trillion-dollar companies, or at least many more hundred-billion-dollar companies, I should be so much more elastic on my pre-billion-dollar entry price, paying $300 million or $600 million?

Well, because most companies don't become that. It just depends what you tell your LPs. By the way, we tell our LPs: Make 2 to 5x in 3 to 7 years, rinse and repeat; generate 2 to 2.25x, 2 to 2.5x net to funds with 29% IRR. It turns out that if you can do that over a 20-year period of time, you are the best of the best.

Dude, I'm very good friends with Jason Lemkin from Shasta, who, on a show with me very recently, said, “Oh, fuck this picking-winners-early-stage stuff.”

I just want to do an Anthropic SPV. It’s much easier.

I thought it was easy to do this stuff. I think 50% of people in the venture business should not actually be in the business. There are 50% too many VCs. Maybe more. There might be 70%. Actually, it’s not only VCs; it’s private equity and alternative assets.

What makes you say that?

There’s too much money, and there are too many tourists and people that don’t show investing discipline on price. Go talk to the greatest venture investors on the planet, and they’ll be like, “Price matters at the end of the day.” Again, let’s see what all these companies valued in the trillions of dollars get out at the end of the day. There aren’t that many of them, to be clear.

Look at how much earnings companies like Facebook, Google, Microsoft, Amazon, and Nvidia generate—profits. I think there are a lot of companies that are way ahead of themselves in terms of valuations and what their profit numbers will be at the end of the day.

Do you think we will have more or less money in venture in 3 years’ time?

Probably less. I don’t know. At some point—I don’t know if it’s 3 years, 5 years, or 7 years for sure—people are going to wake up in 2030 or 2032 and realize, “Oh my God, they’re still all in this stuff from 2012 and 2015.” If you weren’t selling, when are you going to sell?

That’s actually the best advice I would give to young fund managers and people starting funds: liquidity windows open and close, and when they are open, take advantage of them. You should be selling, even if they’re winners. Sell 20%, sell 30%, sell 5%, and continue to get money back. Again, your job is to return money.

Companies are bought, not sold.

Yeah, I guess marks are opinions and DPI is math. No, no, “companies are bought, not sold”—that’s not true. But you don’t have to sell the whole company. If there’s a round being done in a company that you’re invested in, especially if you’re a small, new fund, you can go to the founder and be like, “Look, founder, you’ve taken some money off the table. I really need to sell some stock here. I won’t be in business in 5 years if I can’t get some liquidity back to our company.”

By the way, you can do the math. If you build the next giant company and you sell some at $500 million or $1 billion, who cares? You still own 80% or 90% of the whole thing. But LPs want money back, and the people that are going to stay in business 10, 15, or 20 years from now are people that will continue to give money back to their investors.

Are you seeing LP sentiment change today around what they care about?

Absolutely. I think you’ve already started to see it. We’ve always cared about DPI, but we’ve always been really disciplined: 2–5x, 3–7 years—did it, move on. Probably a third of our deals have been secondary sales, so people have accused us of being traders. That’s fine. I guess I’m a trader, but you know what? I gave money back to my investors.

The investor is my client. I have 2 clients: entrepreneurs and investors. Without investors, I don’t have any money and I don’t have a business. I think people need to remember who pays the bills.

I do think that investors are very, very, very focused on DPI now. It is possible with a small, early-stage fund as a new, relative upcomer in your first couple of funds to generate amazing DPI. It’s a game you can play. I don’t know if you’ve ever had Fabrice Grinda here, or Jose Marin from After Labs. Those guys have been LPs of ours for 10 or 15 years, and they’re friends. Those guys have played the game extremely well. They understand that not every company goes to the moon: take some chips off the table, give it back to your investors, rinse and repeat.

I think the advantage that a lot of these small funds have, which most people don’t consider, is they’re able to sell so much more easily without really disturbing the asset pool. If you’re Fabrice Grinda, with respect to him, he can sell very easily and it’s not a big problem. If you’re Sequoia—

Correct.

—it’s hard.

It’s like negative signaling. 100%. First of all, that’s why I’m telling younger funds: you can say to the entrepreneur, “Hey, you’re selling some stock. I really need to sell some stock here. Why? Because if I don’t return money to my investors, we won’t be talking in 3 years because I’ll need to find a new job, because I won’t have my second fund.”

Do you worry about being a trader and that not being an attractive investor to founders?

Nope, because I believe if you help founders and do what you say you’re going to do. Now, a lot of investors don’t do that either. I think there are 50–60% of people in this industry that actually probably add negative value to companies.

The simplest lesson I think entrepreneurs—anybody—can learn, and I learned it early in life, is just: if you say you’re going to do something, actually do it. It’s the number of people that promise, overpromise, and underdeliver. Be the reverse: underpromise, overdeliver.

If you’ve been really helpful to an entrepreneur, helped them recruit, helped them with customers, and you sell 20% of your holdings, who cares? Keep helping them. We have companies we’ve sold 100% of that we still help drive customers to. It’s great. By the way, “Great, you helped me make 5 times my money. Nobody else on the cap table is liquid. We sold all of our stock. I’ll keep helping you. Please.”

What are the most common ways you see investors provide negative value to companies, for founders listening, that they should watch out for?

Burn money at all costs. Recruiting. They act like they know how to run the business. I’ve never run a company in my life. Neither have 98% of venture investors or private equity investors. Get people around the table who have done what the entrepreneur is trying to do.

If you’re a $20 million AI company or a $20 million software company, find entrepreneurs around the table. Help the founder recruit people that have built businesses from $20 million to $200 million, get them around the table, and get out of the way. It is truly the advice that I think entrepreneurs need.

I just think there are too many knuckleheads. The worst is somebody who went to Stanford Business School and worked for 18 months at a startup, and now comes in as a venture investor and now they’re experts, right? I just think it’s about being humble and not acting like you know, because most of these people—myself included—have never actually run a company.

By the way, if you want to start a venture fund or growth equity fund, I’ve done that. I can give people advice on that. But if you’re trying to figure out how to build out a great sales team, go talk to great sales leaders and get advice from them.

I actually think that’s the best way that VCs and private equity people can help founders and entrepreneurs: connect them with people who have done it before and help them recruit. I truly think that’s why Sequoia, Benchmark, and Index are great, because they help founders and entrepreneurs recruit amazing talent. People want to work for those funds’ portfolio companies.

Dude, I just said I’m a switchboard. I honestly just feel like the women in the 1960s, connecting people all day, because I have no idea how to do that. Speak to my friend C-3PO.

100%.

Great. But I have no idea how to do that.

And I think that’s what the best entrepreneurs actually want.

100%. Uh-huh. Here’s my mid-view from seeing a load of companies: no, don’t want that. You sit there like, “Oh, a negative-value investor is like, ‘Hey, burn money, burn money.’” They say “burn money” because they need to see growth, and they’re looking at markets today going, “For me to get my next round of funding, the game’s changed.”

Yep.

It’s no longer triple-triple, double-double. It’s, “I need you to go from zero to, honestly, 30 or 40 for this to be interesting.”

Well, that’s because that’s the price that they paid in the deal. If you paid these asinine prices on the way in, that’s the result.

By the way, what’s really interesting to me in the market right now—and we don’t even see any of these deals because it’s not what we do—is people who spin out of Anthropic or OpenAI and raise money at $2 billion or $1 billion for a freaking idea. There’s nothing more than an idea and a napkin. To us, that seems complete lunacy.

What I actually want to know is, have any of those ever actually worked? Has anybody ever raised at some crazy price initially, just on an idea and a napkin—like billions of dollars or $500 million—and have any of these things actually ever worked? I don’t know.

Anthropic?

I don’t know. When did Dustin Moskovitz invest? I think it was pretty small.

I was like, “Billions.”

No, no, I think the original seed deal was much lower than that. I don’t know is the short answer, but I thought when Dustin and some of these guys invested, the seed was much lower. But I don’t know.

The astonishing thing is Dustin will definitely make more money from his Anthropic investment than he will from Asana.

Definitely from Asana. Definitely from Asana.

I don't want to be horrible. Facebook, though, I'm not sure.

Yeah, yeah, not sure about that.

We were talking about burning money and wanting growth. Is it the triple, triple, double, double? There are so many SaaS founders who ping me every day, and they're like, “I've been told for 10 years: triple, triple, double, double, and then we'll be in a good place to raise our next round.” I'm in that third year where I've doubled, and I've now gone from $10 million to $20 million, and no investor wants me.

I think it's about that. Well, call us. We think those things are interesting because we can get them at potentially good prices. That's where it matters. What is your gross dollar retention?

It's the most important number in tech companies: gross dollar retention. What I mean by gross dollar retention—because everybody wants to quote net—is gross dollar retention. You ended 2024 with $20 million in revenue. What did you end 2025 with from just those same customers? No upsells, just downsells. You want 90%.

Anything less than 80%, we won't touch. 90% is great. 95% is incredible. You're looking for 90% gross as good, 95% as great, and 98% as amazing. By the way, the reason why there's so much deadwood inventory and all these living-dead companies is that there are so many companies with 60%, 70%, or 80% gross dollar retention. Good luck.

The problem isn't when you're at $10 million in revenue; it's when you get to $150 million in revenue and have 70% gross dollar retention. You're just churning through customers. That's why a company with 95% gross dollar retention can grow really fast and not burn much money, because it's not spending money on sales and marketing to fill up the bucket.

If you were Thoma Bravo and you've got your Coupas and your Anaplans, where the companies are growing at best in the mid-teens, what happens to this generation of growth-equity and private-equity investors in tech?

I think growth equity and buyouts are very different. If you want to go to large-cap, people like Hellman & Friedman and Permira are probably slightly more growth-oriented. There are probably other firms that are more margin-focused. I think it's probably a function of how much debt they have on their companies.

To be honest, I have not spent tons of time studying the financials of Coupa Software or Anaplan and things like that. I know that all these companies drive EBITDA margins from 5% to 40%. The question is how they're doing it, which I actually don't know. I would hope that they've done it mainly through cutting really inefficient go-to-market, sales and marketing, and G&A. I would hope they haven't taken engineering headcount down to 20. I suspect they have not, but that would worry me.

By the way, these people are really smart. If those companies were bought with no debt, I'm sure they would be investing hugely in AI. They probably already are. But, for me, that's why—as I said at the beginning—we worry about any company facing any big technological disruption that is levered with a lot of debt, regardless of whether it's a software company, a manufacturing company, an accounting-services firm, or an industrial-services firm. With debt, you're hamstrung in how much you can do because you have massive interest payments to pay.

Do you not worry that, given the casinoization of public markets and the volatility that ensues from Figma riding up to where it is and then ending up in the gutter, and Atlassian posting accelerating numbers while being down 76%—

You buy a company like Atlassian. That would be my argument.

Okay, fantastic. I agree. I think it might be amazing. Also, a founder-like company. Going back to my point, if I'm Canva, I'm looking at this and going, “There's no freaking way I'm going out.” If I'm Stripe, I'm going, “Thank you. I feel very vindicated in my decision to stay private.”

Does this meme-ified stock market not just make—

Good for LPs. What are you talking about?

—the liquidity problem worse?

Correct. So, what will change it? When LPs go to the biggest venture funds in the world and private-equity funds and say, “We're not investing in your next fund until you get liquidity in these names.” That's the reality of it.

I agree with founders, by the way. There is something to be said for a company going public. I have friends with a bunch of guys who run big public companies.

You just had EquipmentShare.

Yeah, good one. By the way, they went public, and their stock has actually done quite well. It's a derivative play. It was our derivative play on AI because they built data centers, right? You have to get a shovel and a dump truck.

But again, that was a company in 2020 and 2021 that everybody thought was a tech company. It's not a tech company. It's an equipment-rentals business run by a couple of awesome founders who are studs and killers. We were able to partner with them to buy a bunch of secondary from people who were desperate to sell, and it's been an awesome investment. That's how you go into that business. You buy a ton of secondary.

You bought a ton of secondary from people who wanted to sell.

Yeah, they were like, “Oh, this is a tech company.” No, no, this is an equipment-rentals business that uses some technology. It will drive higher EBITDA margins and better utilization rates.

By the way, they went public because they wanted to get their brand well-known. There are a bunch of companies out there, like the Veevas of the world, Datadog, and CrowdStrike. Why do those companies go public? Because big enterprise companies are like, “Hey, all right, how big are you? Are you survivable?” You have big public companies, like Microsoft, telling people, “This company is a fraud.” You're like, “That's not a business. It's a $30 billion company.”

Being public helps, I think, with credibility. On the flip side, for the entrepreneur, I understand why they don't go public. They have $1 billion or $2 billion in cash. For us, it's really about finding entrepreneurs that we are very confident will be public companies. We talk to them up front: “Listen, if your plan is to try to stay private forever, you're just not for us. We don't want to invest in you.” Not because we don't respect you, but because we need to get out of things at some point.

Why, given the fluidity of secondary markets, if you spoke to a Collison who was like, “Yeah, we absolutely want to stay private forever,” and you were like, “Great, but there are very liquid secondary markets for that”?

Yeah, that's true. There are now. Our view is that we want to back entrepreneurs who, if we think the company has a very good shot at going public, want to hear from the entrepreneur, “I want to go public.”

By the way, you know how many more public companies there would be right now if these companies didn't have $600 million of cash on the balance sheet or $1 billion of cash? If they only had $30 million, they'd all be public companies. The amount of money that's available for these companies is incredible.

Now, again, who knows? I have legitimately no clue if Stripe is trying to buy PayPal. But if they're trying to buy PayPal, they would probably rather be a public company right now to do it, unless they've got some sovereign wealth fund that's going to write them a $40 billion check.

So, just help me understand: why would Stripe being public help them buy PayPal?

How are they going to pay for it now? If Stripe was worth $100 billion—and I don't even know what the market cap of—

$150 billion for Stripe.

For Stripe. If Stripe was worth $150 billion and PayPal was worth $50 billion, they'd have a liquid, publicly traded stock that they could then merge together. It wouldn't be a cash offer; the shareholders of PayPal would get liquid stock and sell the stock. That's an all-stock merger that happens all the time with public companies. PayPal shareholders are not taking private stock in Stripe. That's not going to happen. Public companies don't do all-cash offers.

Now, could they go public in a reverse merger? Maybe that's the way. Maybe you reverse-merge into PayPal. Again—

There is no freaking way the shareholders would ever do that.

The reality is, I think the only way a deal like that would happen is through some sovereign wealth fund. Private-equity funds don't have enough money to write $40 billion checks into companies. So, if Stripe was really trying to buy PayPal—again, who knows if it's even true—if Stripe were a public company right now with a $150 billion market cap, it would probably be less now because the stock market has gone down or whatever.

Say it was $100 billion, they would be able to do a public-to-public merger. They could do one. It gives them less flexibility to do big, really big M&A, obviously. They have access to a lot of cash, so they can go buy some $1 billion company.

Do you think there is a massive disconnect between publics and privates when you look at a $150 billion Stripe and a $45 billion Adyen and the same, similar-ish transaction volume?

I don't know. Yes, we see rounds get done.

You know, Wix is at $4.5 billion and Ramp is at $9 billion.

Yes, correct. Exactly what you'd think. It's kind of silly. Do these private-market investors look at the public markets?

Should venture investors have more flexible mandates that allow them to adjust to asset classes where there is the most opportunity at a given time?

I think most—look, it surprises me that more growth and private-equity investors can't do public companies inside their funds. I like these guys, actually, but TCV can, I think ICONIQ can, and I think GA probably can. So I guess these people probably can.

It surprises me that more people don't. Early-stage venture is extremely different from going to buy Atlassian now, or Workday, Salesforce, Toast, or something like that. But what surprises me is that more funds don't do this. If you were an early investor in Toast and you're fully out of it, you love the company, and the stock's down 60%, why not go buy it again? It surprises me that more people don't do that.

Totally get that. Going back to the equity sell-off, you said about that size of IPO, a lot of people say the $3 billion to $10 billion IPO range is just—you’re too small for anyone to care. It's not meaningful in terms of size to the actual markets. Is that bullshit, or is that fair?

In a world where ETFs and passive funds have become huge and the number of fundamental investors is shrinking, I don't think you want to be a $2 billion company. Ideally, you don't want to be a $2 billion to $3 billion company, but I don't think there's any reason you can't be a $7 billion to $10 billion company, or a $5 billion to $10 billion company. I think those are legitimate.

When you get below $1 billion or $2 billion, it's kind of just a pain in the ass to a public-company investor. But you can do it. Who says you can't do it? Actually, we wish probably more would. Why? Because they'd be great small-cap stocks, and you could hold them for 10 years and make a ton of money.

Look at AppFolio. The thing was, I think, $1 billion or $2 billion—I think it was less. I think it was around a $700 million market cap when it went public. I don't know what it's done in the last month or 2, but it was like a 10X.

And then you got Shopify. Correct? It was a $2 billion public company. I mean—

And by the way, that's what you would hope for. You want to find those next companies.

Is there anything where you have to be a certain size in terms of public markets? If you apply that to funds, we see $15 billion Andreessen and $10 billion Thrive.

It's insane. These funds are way too big. Just do the fund math: You clearly have to have the next OpenAI or the next Google, effectively, or the math doesn't work, I don't think. I've even heard people say you don't have to have 1 of them; you have to have 2 or 3 of them.

The amount of money being thrown at some of these funds and the size of the funds is astonishing to me. I hope they prove me wrong, because it's good for me, too. If these funds get so big and these companies get so big, then they can buy a bunch of our companies like that.

I respect people like Benchmark or Index. I think they're just wrong, but Index can raise as much money as it wants.

Yeah, and it's actually rather a big $1.5 billion, or whatever it is.

It's tiny compared to these other funds. By the way, those guys are the best in the world. I think it just gets really hard. It's crazy.

Why? Let me push back on you. There's $9 billion in Thrive's growth fund. If they aren't able to put $2 billion or $3 billion into Cursor or Databricks, you can break them up.

You just have to underwrite $150 billion companies. In steady state, companies trade at 10 times earnings. That's historically what they trade at when they don't really grow. Now, we can argue: Is it 12? Is it 8?

Are you investing in something that's worth $100 billion? You're effectively saying that, to make it double with dilution, it's probably $250 billion. You're making the bet that it's going to do $25 billion of earnings. There aren't that many companies that do $25 billion of earnings. It's freaking hard.

That doesn't mean you can't make a lot of money between then and steady state. A lot of it is growing 50% a year. But again, I would say there's too much money chasing too few things.

AI is going to create a bunch of great new companies. These guys probably will go find them. If anybody's going to be on the bandwagon to find the next giant company—the social-media equivalent in the next 5 years—it will probably be 1 of these funds. I can see the argument as well.

Everyone says that you need to be big or boutique. Does that mean you agree that you die in the middle?

No, I don't think so. I wouldn't bet against people like Benchmark or Index at all. I think you can stay nimble. Those funds will invest earlier, and their percentage ownership along the way will just get diluted over time.

If anybody's going to find the next OpenAI or Anthropic, I would give Index or Benchmark just as much probability as Thrive or anybody else.

You seem very focused on praising Index, which I love.

No, I—

No, no, no, no. I love Danny, and Danny's a good friend. What specifically about Index is impressing you so much?

Just look at the returns in the funds. It's math. Look at their DPIs. They are amazing investors. There are a lot of people who get lucky with 1 or 2 funds, but if you can do this over a 20- or 30-year period and consistently put up world-class returns, you're doing something very unique.

Does king-making exist?

King-making is when you get Sequoia and Andreessen, and then Benchmark, and then other top firms. The names are so good and the money is so much that you make the winner in the market. No one else wants to go in because there's the Sequoia-funded company, which has got Iconiq and then got Andreessen and everything else.

What you're seeing now is that there's so much money, there'll be multiple players in that space. But Anthropic—I give the guys at Menlo Ventures a huge amount of credit. I give Matt Murphy, who's a buddy of mine, a huge amount of credit. When they did Anthropic, it was not obvious. It was like, “Are you nuts?” Look, I mean, he's right.

Let's do a quick fire, my friend. What have you changed your mind on in the last 12 months?

I think AI is going to be even bigger than we thought it would be. It's going to change the world in so many more areas that we're not even thinking about.

What's the single most memorable first founder meeting you've had?

Nat Friedman, founder of Xamarin. We whiteboarded how to save money with Starwood hotel points and Delta airline points. He's just a great, humble, normal guy, and he already runs, obviously, AI at Facebook.

You can invest in 1 seed fund, 1 Series A fund, and 1 growth fund. Which do you choose?

A Series A fund would probably be Benchmark. For a growth fund, it would be myself, because that's what I invest in. I have a huge amount of respect for the guys at Iconiq, though. We don't really compete against them because they're a Silicon Valley company; we're outside Silicon Valley.

For a seed fund, I don't know. We don't have much experience with them, but I think the Founders Fund guys have returns that are totally nuts. That would be another really good fund as well.

What's been the hardest decision you've made in your career?

The hardest decision has been to hire people who have the same view on how to generate returns, so that you're a cohesive group. I think that's the hardest thing.

Because yours isn't the sexy way.

It's not the sexy way, correct. The hardest decision was: Should we have, in 2017 and 2018, like in 2016, paid when everybody was paying 5 to 7 times revenue for software companies, and Iconiq came in with everyone paying 10 times and just winning the best deals at those prices? We were like, “We know these are the best companies.”

Should we do these deals? We did not. We were wrong.

What was the biggest miss, and how did that change your mindset?

Procore—we were in it a tiny amount. We did the deal with Bessemer and put a tiny amount of money in, and we were in a position to lead the next round that Iconiq did. Then, 2 rounds later, we sold to them. Again, this decision to sell actually wasn't that big of a deal because it wasn't that big of a position. The mistake was not actually investing.

We also got $2 million in the Shopify IPO because we knew the founders, and that would have returned our second fund 2X had we just not sold the stock. We didn't have to do anything.

What investor do you most respect who does not get the spotlight?

Probably my partner, Neema. I just think he's insanely disciplined. His loss ratios—I don't think he's ever lost money. Maybe he's lost money in 1 deal ever. Again, he's never had 10Xs or 20Xs either. Probably my partner, Neema.

Final one: What are you most excited for when you think about the next 10 years?

Actually, what I'm most excited about is that there's going to be a really bad downturn. It's different from 1999 and 2000, but there's going to be a really big downturn. Markets just don't go up forever. Economies just don't go up forever. I think there are a lot of policies in the government and the world right now that might end really badly, and I think it's going to happen in the next 10 years.

That'll be the best time ever to invest. Combined with the productivity booms you're going to have with AI, it's like you avoid the Gen 1 AI companies, just like if you had avoided the Internet 1.0 companies. Then think about all the internet companies that were started from 2003 to 2006. I think the same thing could happen with AI, which again could benefit an Andreessen or a Thrive, who are raising these new giant funds. They could potentially invest them during those periods as well.

Always have money to play the game.

Correct. If you don't have money, you're out of the game.

Dude, this has been such a pleasure. Thank you so much.

Thanks for having me on.