Figma 上涨250%：史上最大IPO定价错配？Meta与Microsoft季度业绩大爆发：拆解

Rory O’Driscoll

The people who said, “Oh, Figma left $3 billion on the table.” The $98 price only happened because the IPO happened at $38. The discussion they were having on the day was, let’s call it the Halligan discussion, of, “Do I go $2 more and exclude Fidelity, or $2 less and take Fidelity?” It’s a good, useful discussion. Had someone walked in and said, “I know this IPO is going to price at $100 a share to open tomorrow morning. Let’s raise at $80,” they wouldn’t have had a book, because no one had bid at that thing. So, that money wasn’t accessible.

Jason Lemkin

Run, Forrest, run. The market’s wide open. The valuations are good. There’s a lot of demand. It’s very seasonal. Timing really matters.

If I were Canva—it’s an amazing company—I’d be lining everything up to go public.

1. The Worst IPO Mis-Pricing Ever: What Really Happened at Figma

We’re just going to start and dive in with the Figma IPO. We saw the most unbelievable mispricing. It went out at $33. It went up to around $145. I’d like to throw this out there to some of the greatest minds in this business: How did we analyze this unbelievable pop? What was it at HubSpot, Brian? There must have been a pop, right?

Brian Halligan

We had a pop.

You priced at $25 and you opened around $30.

Brian Halligan

We priced at $25. I think we opened at $33.

Yeah.

Brian Halligan

It might be worth just talking about what happens behind the scenes on this.

Yeah, everyone’s talking out of their rears on X, aren’t they? They act like they know everything.

Brian Halligan

Yes. I don’t know everything, but I can share what actually happens and how you make that decision and what the pressures are. First of all, you’re making the decision on the price and who the investors are the night before the IPO. At this stage, you have never been as tired in your entire life as you are when you’re making this decision.

You’ve been on the road for the last 2 weeks. You hit 12 countries. You had 6 pitches a day. Your battery is on red. You’re tired, and then the investment bankers sit you down and say, “You’ve got 2 big decisions to make.” One is who the investors are going to be—who are we picking? We were 27x oversubscribed. Figma was 40x oversubscribed. Who are they going to be? Then, what’s the price?

For the whole process, the founders were very well aligned with Morgan Stanley and the investors—perfectly well aligned—until this 1-hour meeting the night before the pricing. All of a sudden, you realize, “Actually, we’re across the table from you.”

The first thing you’re across the table from is they give you the book of all the people who want to buy your stock. Again, 27x more demand than you have supply. In their version of the book, they have a lot of their hedge fund buddies in there with pretty good allocations. We also have some of the quote-unquote long-onlys. We could talk about that, too: Fidelity, T. Rowe Price, Wellington and Capital Group.

What HubSpot wants is just the long-onlys. You want to keep the hedge funds out. Morgan Stanley wants that, too. They want to keep them happy, but they also want their hedge fund buddies in there because they make a lot of money on that. So, that’s a negotiation, trying to get all that—squish the hedge funds down and get the long-onlys up.

Then it’s the price. In HubSpot’s case, it was a very interesting dynamic. We had raised the range throughout the roadshow over the previous 2 weeks. Morgan Stanley—we wanted to go out at $25. Morgan Stanley said we should go out at $24. The reason we had $24 was that Fidelity had told us that they were in at $24 and out at $25.

You really want Fidelity because Fidelity has trillions of dollars, and they could own—you know, they could be a massive buyer. So, then you have a debate in your head: How badly do you want Fidelity? How much do you want to sell for?

We were oversubscribed, and we thought, “I think Fidelity is going to come in anyway. It’s a really good thing. I think they’ll come in. We’re going to stick with $25.” So, we pushed back. We said no, and we priced it at $25. I don’t think there was collusion going on, but definitely Morgan Stanley and Fidelity were on the same side of the table, trying to get us to sell at a lower price. That’s kind of how it happened behind the scenes.

So, there are these micro-conflicts, but they rear their heads at the last minute in particular, right? At the last minute.

Brian Halligan

It’s really just the very last minute. It’s fine, and by the way, there’s a lot of talk around the pricing and how much it pops. Whose fault is it? Morgan Stanley’s fault? The founders make the decision on what the price is, so Dylan decided at the end of the day.

By the way, if you’re Dylan and you’re at Figma, you had a very limited amount of shares that were selling. They were mostly secondary. There were just very few shares to sell and a lot of demand for it. If you’re Dylan, that’s going to drive the price up.

The second thing is, here’s who you really want: You want Fidelity, you want T. Rowe Price, you want Wellington, you want Capital Group. There are about 6 big long-only funds you want in there. They’re pissed because they’re not getting a big allocation. That created the demand environment that was a little tricky for them because they wanted to buy more shares, and that’s what pushes the price way up.

The supply is low and the demand is very high. It’s an interesting little game of chicken. You played chicken with your investment bankers and some of the long-onlys right at, like, 7:00 the night before the IPO.

But there’s no one dumb at Figma. This is the other thing about social media. You think Dylan and his whole finance team and CFO have never heard of these issues before? I mean, everyone’s trying to weigh everything and come out with the optimal outcome, right? No one massively mistakenly underpriced this deal, did they? We’re not talking about rookies here, are we?

Brian Halligan

I don’t think so. I think there are a lot of smart people around the table giving them advice, and they wanted to make their investors happy. Part of it also is you want to have a pop. You want Fidelity, T. Rowe Price and these long-only folks to come in and, right out of the gate, feel good about their investment. You want them to hold for decades and decades and decades. So, there’s a little bit of that. It probably just popped more than he thought.

And that’s the actual sentence, right? Until that, Brian, everything Brian said is correct. The stunning thing is they clearly give the same speech every time, because I’ve been in the room, I think, 5 or 6 times, and they give it. It’s always Fidelity and one of the other 2. It’s like, “You want Fidelity. It’s only another dollar,” right? That whole speech is exactly right, and you want a little pop.

The interesting thing that Brian said at the end was the odd thing here is, instead of getting, let’s call it, the designed 15–20% pop that makes everyone feel good, you ended up with this absurd 250% pop, which is clearly off the charts. I think it’s the largest pop since 1999.

What you’re really saying is you’re trying to engineer this small explosion, and then every once in a while, inadvertently, you create a big explosion and you look like an idiot. So, the question is, are we arguing about whether we need small explosions, which happen all the time—in other words, the classic HubSpot 20% pop? Is that a problem or not? That’s all about a $1 discussion with Fidelity.

Then there’s the separate problem of when you get it massively wrong. What causes that? They’re 2—I think you should talk about them separately, because they’re almost 2 separate issues. Does that make sense?

Brian Halligan

Because there’s 1 small thing I heard nobody talk about on X. At my first startup job, they gave me the job—and this is the only IPO I’ll ever have been through as an employee—of handling the directed shares program. It was my job to hand them only to the employees, not the external ones.

There was a lot of drama because people had to come up with $50,000, $60,000 or $70,000 to buy their stock. This is a lot for folks who had no secondaries. There was no secondary, and everyone in the company basically made $100,000 that day. It was a magical thing.

Now, was it underpriced? But we forget about the employees. They don’t care about that. They don’t even know what dilution is.

Yeah.

Brian Halligan

All they care about is what happens with their stock. It was a magical moment for the employees. Whether it was mispriced is a different issue, but no one’s talked about that transfer. It may be a small issue, but employees may not care.

I guess a core question for me is: Do we think it was fundamentally mispriced, extensively, or is this actually just IPO exuberance on an asset that is good, with a great-quality founder? Is it not dramatically overpriced at the $137 that it hit?

Rory O’Driscoll

Agreed. That’s actually the next truth. You’re exactly right, because the whole 20% pop discussion is an interesting one. Maybe Fidelity, to use Brian’s example, would have bought at $24 or $25, and they probably bought more at $30, right? That’s kind of in the bounded range.

I can tell you right now, the people who said, “Oh, Figma left $3 billion on the table because they could have got $95 a share,” are talking out of their ass. I can guarantee you the order book—I think it was, what did the price say, at $35, $38, whatever it was—there was some at $38, there was some at $39, there was some at $40. No one was putting in an order at $98.

What happened—and this is kind of a bit existential, metaphysical; maybe go with it—is that the $98 price only happened because the IPO happened at $38. It’s exactly what you said, Harry. The discussion they were having on the day was, let’s call it the Halligan discussion, of, “Do I go $2 more and exclude Fidelity, or $2 less and take Fidelity?” It’s a good, useful discussion.

Had someone walked in and said, “I know this IPO is going to price at $100 a share to open tomorrow morning. Let’s raise at $80,” they wouldn’t have had a book, because no one had bid at that thing. So, that money wasn’t accessible. All that happened here was, as you say, I think maybe they could have got a couple of extra bucks, and maybe that might have slightly dampened demand, but really what you’re saying, Harry, is correct.

Rory O'Driscoll

There was some pent-up euphoria on the retail side, and they all rushed in. I can tell you why it's not successful. Going back to Brian's example—and this is a really sad comment, but it's true—if you have a 20% pop and you get Fidelity at $25, they'll buy more at $32.

Let me give you a really sad fact, and I didn't realize this at first. All those big buyers who came in at Figma at $38, before they buy, they have an internal process with a price target to exit. None of those price targets are going to be more than $110 a share.

So, unfortunately, because you want to price it—I won't say underprice it—it's now trading at a price that's probably above the long-term price target of the long-term investors you want. Most of them are selling those shares right now. I used to think, “Oh my God, they'll hold because they know it's a long-term investment.”

But if you're running money as one of these institutions and you bought at $35 and built a business case that says, “We think this will be worth $50 a share in 2 years,” and suddenly it's worth $100 a share in 2 days, at least half those mutual funds have to say, “Should we lose our position?” My guess is they will.

Brian Halligan

People say long-only. They're not long-only. They do sell. They come in and out of HubSpot. I see them coming in. I think they're trying to build a position. They only floated a tiny amount here, and so I think they're like, “Okay, we're in there. We got a toehold. We're going to hold for the long haul. We think this is substantial.”

I doubt they sold. I think it's Harry Stebbings that sold. I think it's hedge funds that sold.

Jason Lemkin

I think they're definitely right on the hedge funds. No one in the hedge fund world is out; it's in.

Brian Halligan

Yeah, I doubt the big long-only—quote-unquote, long-onlys—sold. Can I tell you guys another story about this whole thing?

First of all, on this topic, before HubSpot went public, we did something called a non-deal roadshow with Morgan Stanley. We met all the big investors. All of the big investors we met came in big on our IPO.

The ones we missed, we couldn't convince to come in on the IPO. For us, we missed T. Rowe Price. It's down in Baltimore. We didn't want to go down to Baltimore. And we missed Capital Group in Southern California. It just didn't quite hit. It was a long trip, and we missed both of those in the IPO.

It took a good 3 or 4 years before we got T. Rowe Price and Capital Group in and convinced them this was a good company that was going to be around for the long haul. So you kind of want those guys in there, and you want to make it a strong incentive to get them in right off the bat, or you might not get them for a while.

And meet them in person.

Brian Halligan

Yeah. So the non-deal roadshow is key.

Jason Lemkin

It's also proof of the old rule: You always remember the people you didn't get, no matter how long ago. All of us have been fundraising. You can vividly remember every no. You can vividly remember every, “I didn't go and see them, and I should have.”

Do you think we overestimate the importance of having Fidelity in the IPO?

Brian Halligan

I think it's important. Fidelity and T. Rowe Price own a huge chunk of HubSpot now. Wellington Management owns a very large chunk of HubSpot, and they're pretty stable. They trim on the edges and come in and out.

Most people think consumers own a lot of the stock. The retail investors own a lot of the stock. Very little of HubSpot is owned by mom-and-pop investors. It's like 90% big institutional investors.

Brian, you never gave me access to the IPO. I would have been—

Brian Halligan

For the record, Harry, I think you were 12 at the time. I'm not even sure you could legally own stock, big guy.

Dude, there's no way I could have bought. I was totally underage. The one thing that I do think is, if we're Dylan Field now at Figma, are we thinking, “You know what? I wish I'd done a direct listing”? Does this make direct listings much more attractive, and would this have solved the problems that we've seen?

Rory O'Driscoll

Again, I'm going to push, because when you sent out the questions, I did more reading on this than I had in a while. One thing I didn't realize is the SEC has amended the rule, so you can now raise capital on a direct listing. So it is possible, right? I think Amplitude did it.

But—and this is the big but—if you did a direct listing, and this is why I said it gets kind of weird, I don't think you would have direct-listed at $100 a share. It might have been $39 or $40, right?

The weird kind of thing is, I've just—how do you put it?—come to think that it may well be that these random, weird pops every once in a while are just a natural phenomenon, like earthquakes. They just randomly happen, right?

2. Is Canva Next? Why Founders Should "Run, Forrest, Run" to the NASDAQ

Because it's not like, if you did a direct listing, people would have said, “Yeah, I think I should pay 8 times run-rate revenues for this puppy.” It would have been the same analysis at 15 times or 20 times run-rate revenues: $35, maybe $38. You might have captured the 2 or 3 extra dollars that you left on the table, in return for getting Fidelity.

And then you would have had the Brian CEO comment of, “Is Fidelity worth 3 extra bucks?” But what you would not have done is place the stock at the lofty level it is now.

So it may well be that mega-pops are a natural, intermittent consequence of the IPO process, and that a direct listing doesn't solve that part of the problem. I'm sorry—was that as clear? It's kind of weird. What you're basically saying is it's almost like a psychological phenomenon, and I'll give you an example in venture of the same thing. We're seeing it now.

I get it, but it's not, though, because you've got Circle, you've got CoreWeave, and you've now actually had 3 in the space of 2 to 3 months.

Rory O'Driscoll

I think timing in the market matters, and you've had a similar thing in 1999. I think it's a once-in-a-while phenomenon when conditions are adjusting from one more pessimistic stage, which we were in in April. Remember how bearish we were in April? It's only July or August, right?

When people are adjusting, maybe we just don't adjust quickly enough, and the euphoria comes in after the stock gets priced and it's trading. There's a small number of highly speculative assets, and there's a lot of appetite for those assets, so they all just rush in.

It's the venture equivalent of what we're seeing now. When someone with a really big brand name does a round, 6 weeks later many of these companies are doing a follow-on round. Nothing's changed. Everyone bid 120, the brand name won, and 6 weeks later there's a bunch of people doing it at 350. Why? The brand name is in. This is kind of the public-market equivalent of FOMO. That's how FOMO manifests.

Brian Halligan

I think the direct listing is like you're a little bit scared you're not going to get those long-onlys. You're not going to properly market it to them, and 7% to the bankers, in the grand scheme of things, is not that much. I think it's risky unless you're Google or Facebook.

Just to pile on what Rory said, timing really matters. HubSpot went out 3 months before a kind of sister company of ours, Zendesk, went out, and it was just a shaky week when they went out. They didn't get the big long-onlys, and they had a good story, but they didn't get them and they never really got them. They always had a lot of hedge funds in there, and they always had a lot of individuals in there.

Our timing was pretty good—not great, but pretty good. Figma's timing was obviously really good.

Jason Lemkin

But that's super interesting, Harry. Zendesk is kind of a tough tale, right? I don't think Mikkel Svane wanted to sell, right? He turned it down at whatever $20 billion and had to sell for—I mean, $10 billion. These are still good numbers, but he was a reluctant seller, right?

And maybe—and to Brian, I hadn't realized this—maybe part of it was never getting that buffer in, assuming they held, right? He had never gotten that buffer in at the IPO, right? And then you're just—I mean, I'm just watching everything. It's horrible to deal with these activist investors, right? It is. It's a terrible experience as a—

I'm going to unpack this because it won't be obvious to everyone listening what happened there. I had never heard that point before, Brian, and I remember both IPOs as well. What you're basically saying is you guys and Zendesk, I want to say, priced in 2013 or 2014-ish—I can't remember, right?

Right, and then what you're saying is, because of the way your deal came together, you got the long-onlys day one. Zendesk didn't. They priced 3 months early, and the VCs had to put money in the round. It was so tough.

And what you're saying is, 5 or 6 years later, they still didn't have quite as strong an investor base. When they hit an issue and they had activist pressure, it's probably not the only thing that caused them to have to sell, but at the margin, you had a stronger investor base the whole way through.

That's your argument for giving the book and getting Fidelity. Part of it, I would argue, is most of that stronger cap table was—it was more luck than skill. The Mikkel story, the Zendesk story, and the HubSpot story just rhymed. But our timing was better. We had a better cap table. Having a good cap table is underrated.

Brian Halligan

That's why you should take money from me, Jason, and Rory. That's what we tell founders.

By the way, while we're talking about this, just to give the inside baseball on the whole thing.

Brian Halligan

So that night, you’re exhausted. You’ve never been tireder. If my co-founder is really introverted, he has negative energy because you spend so much time with humans, and then that night you have a big IPO party with all your friends. You’re so tired, and the last thing you want to do is go to a party. Inevitably, somebody gets wasted at the party. It’s not the founders. Somebody gets wasted, and it’s a big thing.

Then the next morning, the inside baseball is that we did New York. The next morning, you have a big dinner, and then you’re up on that platform. There’s a huge discussion about who’s on the platform and who didn’t make the platform.

Who’s deciding who’s on the platform? That’s another thing Dylan was dealing with, which is super irritating and probably dealt with a couple of weeks before. Did the VCs make it onto the platform, for example? That’s going to be very important to the VCs, to be up there, right?

Brian Halligan

Yeah. You don’t want to take your picture staring at the big drape of Figma outside Wall Street. That’s pretty embarrassing. That went on X.

Okay.

Brian Halligan

And just a pro tip to the listeners who are eventually going to go public: we’re on the platform, it opens, and everyone on the floor is up there looking at you on the platform. Most companies just kind of sit there like, “Yeah, yeah. Yeah, yeah, yeah. Clap.” Everyone down below is like, “Oh, they’re so boring.”

3. CRV Shrinks, Benchmark’s Bet, and the Future of Venture Strategy

You have to have a plan to do something interesting when you’re up there to get the floor excited and to get the press excited. That’s one thing. Then the market opens, and in my head I thought, “Well, we’re trading. That’s it.” But what happens in Figma, what happens in HubSpot, all these things, is that it takes a while for that first price to settle in.

You and your executive team, and maybe your VCs, are all sitting there extremely awkwardly on the floor of the New York Stock Exchange, looking at a million monitors. You’re not trading yet, and it takes a couple of hours for the darn price to settle in. It took Figma about 6 hours, and it finally settled in. It was like, “Okay, okay. It’s $33. Great.”

I remember that moment because my co-founder, likely Dharmesh, who’s a prince, had the stock app. He showed it to me and said, “Brian, look, look, look. We’re worth $1 billion.” I remember saying, “Take a screenshot. We’ll never see that again.” We were so excited. That’s the behind-the-scenes on the floor.

The other pro tip is that you can do NASDAQ or the New York Stock Exchange. They’re virtually exactly the same, except with the New York Stock Exchange, you get to ring the bell. That’s why we picked the New York Stock Exchange.

That’s really sweet. We did one in Cboe, where you get to ring it, but in an empty TV room. It was kind of very soulless. You’re pretending to be excited.

But no, you’re right. The whole point of the dinner—remember, the dinner comes right after the pricing committee. My mental model of the dinner is that the bankers have just screwed you over for a buck a share, and in return they buy you a very expensive dinner and liquor you up so you forget.

It’s totally discordant because you’ve literally come, as Brian said, from a very angry meeting where these people who’ve been your friends for 2 weeks on the road, carried your bags, and done everything for you suddenly start picking your pocket and telling you that they have to give money to their friends. It’s controversial, you argue, and then at the end you walk away and go back, and then they all ply you with liquor, right?

Remember, this is the information asymmetry. 1 day later, you drive out of town, and they bring in the next group of their best friends. The next lamb is led into the slaughter, baby, right? That’s the deal.

But it worked. I mean, where else in the world are you going to get a couple of billion dollars for 2 weeks’ work?

Brian Halligan

Oh, in Silicon Valley, you get it every day. That’s happening now.

Well, this is true. This is.

Brian, can I ask you one meta question on this? I think I know the answer. When you look at Bill Gurley’s criticism of all of this, one is all the money left on the table, right? That’s the math. But I think underlying that also is dilution, right?

Most early-stage investors, as investors—not as founders, because I’ll ask my question—dilution does creep up on you as a founder who’s gone through this journey, right? Did you ever sweat any of this dilution, the IPO dilution, the post-IPO dilution? Did it even come into your calculation, or did you just not care?

Brian Halligan

My net worth went from X to 100x. I was just happy to be there.

4. Why CEO Compensation is Broken

That sentence is why this process is so hard to change. The profound truth is, as Brian said, you’re doing this once, and it’s the most important thing in your life. All these guys are doing it every day, and they know so much more than you. It’s a really hard process to reform. It’s one of those ahas.

I do want to stick on that. You said your net worth went from X to 100x. I think an element that we chatted and texted about before, Brian, that not enough people are talking about is actually the package that’s in place for Dylan, which is obviously this kind of $2 billion moonshot grant, kind of Elon Musk-style.

Brian, we were saying that we don’t talk enough about CEO comp. Why don’t we start with you, Brian, on this, given that you’re the best person here to speak about it? How do you think about this, and how did you analyze that?

Brian Halligan

Yeah, I think CEO comp is pretty broken at the moment. There are 2 things that I think are pretty broken about it. The first is that everyone really relies heavily on RSUs.

When I grew up in the industry—I hate to be that guy, being like, “Back in the old days”—it was mostly ISOs. It was options until 2006, when regulations changed and the expensing of that changed. So the world kind of moved to RSUs.

It creates sort of risk-averse behavior in the CEO. Cash comp goes up and down a little bit, let’s say, but with an ISO, you’re swinging for the fences. You have a strong incentive to swing, and so it’s really had a dampening effect on the risk-seeking behavior of a CEO that I think more companies should want. It’s kind of pervasive across the industry.

I don’t like this RSU comp thing. That’s the first problem I see with all this stuff. The other problem with comp is that almost every company looks at CEO comp, and the way it works behind the scenes is that HubSpot has a compensation committee. Everyone’s got a compensation committee.

HubSpot wants to pay the CEO, let’s say, at the 75th percentile of what her peers make. We look at 20 different peers of similar-size companies and peg her at that 75th percentile, which in her case is $20 million—a lot of money.

Now, if you did that for Dylan, who would be in our comp group with a similar market cap to HubSpot, he’d make $20 million a year. But if you think about it, that’s less. That’s like 3% of Dylan’s own personal net worth. It doesn’t move the needle. It doesn’t matter at all to him.

You have to get creative. I actually like what they did with his comp. They use PSUs very heavily, not RSUs. I like the idea of not pegging your comp to your peers, but you kind of have to peg the comp to the net worth.

It’s the same thing with Elon Musk. If you paid Elon Musk what Mary Barra makes—$29 million a year—do you think Elon cares about $29 million? You have to kind of comp it to the CEO’s net worth as opposed to just the peers. So that’s what I like about this.

I broadly agree, especially on the process side, and you’re right on the PSU. I knew you were going to come in on that, so I’ll just save you the trouble.

A PSU, basically, is an RSU that’s like an option but with a zero strike price, so it’s guaranteed money. Brian’s right on all the negatives there. PSUs—performance stock units—have evolved to effectively make the RSU more like an option.

What you’re saying is that it’s guaranteed money, but only if something happens. The typical thing that people are pegging it to in the public markets, even though I don’t agree—and I’ll come back to it—is stock price. Right?

In other words, instead of saying, “Here’s 10,000 shares no matter what,” it’s, “Here’s 10,000 shares, but you only get them if the stock price is $40.” What you’ve bizarrely done is recreate options, because they got regulated out of existence in 2006. Now you’ve effectively recreated them. The PSU is making an RSU more like a stock option.

Brian Halligan

I like that.

I know. In general, I do too. But watch this. The problem with the Dylan comp package—I’m going to say it didn’t work—is that if you look at it, the problem with stock-price triggers for compensation is that it sounds rational, but they put this in place before the IPO. If you read the triggers, they’ve already achieved them.

The weird thing—and I’ll tell you why they do that in a second—is that all these triggers effectively said that you’ve got a whole bunch of price targets up to $118 a share, I think. When they were making those triggers, literally 2 months ago, they were like, “Yeah, this is going to be great for the next 3 years, but totally out of Dylan’s control just because of the way things have priced.” He’s made all the triggers already.

He still has to vest over 7 years, so it’s not like he takes all the money and runs, right? But the performance element vanished very quickly because of the pop. The aha for me is that I prefer performance-based to not performance-based whenever you can.

I would have preferred, even for a public company, to do them on tangible goals like revenue and operating income and all that over multiple years.

Rory O’Driscoll

And I see, Brian, the problem, and the reason you don't end up doing that is because, A, you have to disclose them, and, B, things change. When you change comp at a public company, ISS and all the whiny babies give you a whole lot of shit.

So what happens—and I've been in the room—is you say to yourself, “I would love to pay Brian Halligan for 35% revenue growth and 35% EPS growth for the next 7 years, and we'd pay him $1 billion.” But if we put that on the table and then circumstances change, we have to disclose it. Then all the analysts will start saying, “Oh my God, they think they can make 35%.” So if they only do 30% growth, Brian missed.

It just becomes problematic, and what they do in the end is say, “Screw it, we'll just do stock-price targets,” which are better than nothing, as Brian says, because it's more swing for the fences. Sometimes you have this weird thing: If you look at all these packages that were put in place, all the ones that were put in place in 2018 worked and gave the shareholders what they wanted and gave the employee—the CEO—what they wanted because the stock price went up.

All the ones that were put in place in 2021 are stranded because all those price targets, like the Airbnb price targets, ain't ever going to happen now. So it's an imperfect mechanism, as comp often is, but directionally the right approach.

Brian Halligan

I'm not in violent disagreement with what you're saying. At HubSpot, the way HubSpot does it is on net-new ARR.

Nice.

Brian Halligan

And an earnings—like, a floor earnings number. And that's where we ended up, by the way. None of this is perfect. It's like, what's the least bad you can do? I kind of like the way we ended up on it.

That is really nice. I'm curious: Do you have to disclose that? Do you get angsty about disclosing that because you're kind of hinting at what you think you can do?

Brian Halligan

Yes, but people can figure out net-new ARR, and they can figure out what your bottom line is. What we used to do was net promoter score and stuff like that, and that gets really tricky, having to disclose all that.

Rory O’Driscoll

I really like it. Very few comp committees do what Brian does. I think that's so much better than just stock price.

Jason Lemkin

Every growth round that I've seen in my little portfolio—all of them—has had moonshot packages. They're going in earlier, and they're becoming a standard part of how many growth funds win deals. They go in and say, “Yeah, I'll do the deal at $1 billion, but I'm going to give Harry another 7% of the company,” and they're always at least 10x. There's a quid pro quo.

It's not 10x; it's 10x from what I'm paying as a growth investor. Fine, I'll do Clay at $3 billion, but if you hit $30 billion, you guys both share another 10% of the company. So I think this is getting institutionalized earlier and earlier as valuations go up.

It may not matter what any of us think, because the growth guys are adding this to the standard term sheet. The flamiest version is where you literally say to the CEO, “I'll reimburse you in options for the dilution you're taking on in the round.” This is a little more high-class than that because it's saying at least you have to achieve first. But I will say—

Rory O’Driscoll

Usually you get more. What I've seen is, instead of a growth fund saying, “I'll give you another 2% back”—forget that—“I'll give you 7% or 8%. I'll give you a massive package, but I've got to make my 10x,” right? It's their version of the Elon package, right?

Jason Lemkin

And I think what happens there is—my prediction is—and again, I'm often disagreed with on this on comp committees because I've done some of these where I've tried to make them based on tangible targets—what will happen when you do the stock-price ones is, if the company's doing really well but the stock price isn't achieved, the CEO will be sitting down 2 years from now and asking to waive some of the criteria. I can just see the movie now, right? And, you know, so be it, right?

Rory O’Driscoll

Maybe I'll bet you nickels to dollars that it doesn't happen, because founders are just signing up. I don't know what, Brian, you see at Sequoia, but founders are signing up for crazy stuff these days. They're signing up for massive stuff, and I don't think a lot of grouchy VCs are going to waive it. I just don't think it's going to happen.

Jason Lemkin

I hear your point, Rory; I just don't think it's 2024 anymore.

Brian, you are an amazing coach to founders. I speak to Pat, Andrew, and Reid a lot, and they say that founders love your coaching, mentorship, and advice when they have big growth rounds like these with performance-based incentives, like we're talking about. How do you advise them? Is it what you thought you'd see now on the other side?

Brian Halligan

I like all this stuff for the founders. When I came through, it was pretty rare to do that. The other thing I like is founders taking a little bit of money off the table on the way.

For us, when we did our round, Sequoia came to us and said, “We'd like to buy some of your shares.” In retrospect, it was a horrible financial decision for me, but it was good at the time. I sold, I forget, a couple million dollars' worth of HubSpot shares. My co-founder and a few other people did.

What I liked about that was that Salesforce came knocking and wanted to acquire HubSpot. It stiffens my backbone a little bit. So it's good for the founder, and it's good for the VC.

I think it gets a little wobbly when it's a $50 million secondary in the Series B, but in general, I like what's going on in venture. I think the valuation is very high right now across venture, so we'll see how this thing plays out. I think it's a little bubbly right now, but I generally like the trend that's going on in terms of secondaries happening for founders, and I like these PSU-type rounds.

There's a founder I work with right now who's a terrific founder who just took 4 years to get it going, and now it's ripping. He'd been massively diluted, and so I'm like, “Well, let's figure out a way to give you a nice, big grant.”

Rory O’Driscoll

I go back to—yes, I agree with all that, and I do think, though, that valuation is the imperfect metric. And Jason, I think you're wrong. I think investors and grouchy VCs will recut because CEOs are smarter than us at comp.

What I've learned is this: I have empathy, but they are running their company and they obsess about it day and night. Let me tell you, if you've got a CEO who doubles revenue for the next 2 years and the only reason he's not making his extra 3% is because we overpaid 2 years ago and now revenue multiples are normalized, and therefore he's not getting his 5x, he's going to come into the comp committee and say, “I have nailed running this company. We have 3x'd revenue. We're operating-income positive. Give me my damn shares.”

And I'll sit there going, “I knew we should have done revenue and operating income at HubSpot from day 1.” What we're doing right now is we're taking what Bob Bryan correctly calls high-end valuations, and then we're 10xing them, and we're basing comp on that. Brutal. We're basing comp on a chimera. It's never going to happen.

Look at all those 2021 value-multiple disclosed ones. Most of them are like, “What were we thinking? We thought that we'd go from $200 billion to $2 trillion.” Maybe not.

If we think about going from $200 billion to $2 trillion, there's going to be a boardroom that's going to be thinking, “Huh, should we take some action now?” If you are Canva, are you looking at this going, “Forrest Gump, run—let's head to Nasdaq”? How do you think about the impact of this on Canva and subsequent companies' willingness to go out?

Jason Lemkin

Yeah, run, Forrest, run. The market's wide open. The valuations are good. There's a lot of demand. It's very seasonal and oddly seasonal, and timing really matters. If I were Canva—it's an amazing company—I would be lining up, lining everything up, to go public.

But the founders have already pledged to give away the majority of their stock. It's not about money for them. The company's profitable.

Jason Lemkin

They're going to give it away to good causes, and they want those causes to get as much as possible.

Yes.

Jason Lemkin

I'm not saying I know the answer. I just think it's more complicated than someone who needs the money. All the early-stage investors have had a chance to trade at tens of billions, right? And the founders—there's just a lot of liquidity already there. The company's massively profitable. I just can't—I'm not smart enough to predict how those factors stand together, right?

This clearly isn't a Musk empire that's being built in Canva, right? It's very different, right?

Rory O’Driscoll

I think you have—I've been thinking about this. You have all the idiosyncratic personal things. You have people who say, “I don't want to go public for a long time,” and you have people who need to go public early, right? All those are idiosyncratic.

But if you zoom out 1 level, I just had this conversation with an LP. They're asking, “When does it open?” In the end, price clears all markets. For the longest time, the money in the private round was cheaper and less hassle for the last 3 years than the public market, so no surprise we did more private.

Let me see: 5x revenues with a bunch of people in New York busting my balls versus 10x revenue, and I never get to talk to these growth-stage guys except once a year. I'm doing option B.

Right now you have a situation where maybe, at the IPO price for Figma—20, 10, 15 times—I should be 18 times revenues. You kind of go, “I can get that privately.” You can't get 80 times revenue privately. So if you're now looking at where things are trading today, I think at the margin those prices are higher than the private things.

Jason Lemkin

So, stepping back from the idiosyncratic stuff at the highest level, the cheap money is now in the public markets. Brian's right: you'd be an idiot not to go for it. If you need to raise money in the next 2 years, now would be a really good freaking time.

5. The Case for Going Public: VCs Are a Bigger Pain Than Public Markets

Brian Halligan

Can I just build on that? In Silicon Valley, I think Stripe is really beating the drum on this: why the heck would we ever go public? There's that sentiment out there because there's so much private capital and you can do secondaries.

I don't have a violent disagreement, but I think people are just nervous about what's on the other side. My take on it was, when we were a private company, we had a bunch of quirky, slightly misaligned venture capital investors who were definitely in our shorts. Then we flipped to public, got rid of those VCs, and had a bunch of quirky, slightly misaligned public investors who were less in our shorts.

It's actually better in a lot of ways than being private. There's so much written about what happened with Zendesk or Autodesk, or these really bad things that happen. It's pretty rare, and I have found public investors to be pretty rational if you paint them a picture of what will happen over a long period of time.

If you're pretty conservative with your numbers, they're rational and they'll stick with you. I think they're underrated, and I think people think it's something scary over there. It's not as scary as people think.

I have a lot of founders who say, “I'm terrified of the activist investors. I'm terrified about what happened to Jeff at Twilio. This is my company. This is my life. I don't want to have that happen to me.” What would you say to that founder?

Brian Halligan

It's pretty rare what happened to Jeff. Everyone talks about Jeff, and everyone talks about Zendesk, but it's pretty rare. That company was having some issues.

Rory O'Driscoll

I agree. I agree about it, and I am a “we should go public” person. I think it's a little overwrought. Yes, some of these issues do happen, but overall, I think the activist VC—because we were one of them—

Jason Lemkin

VCs are a much bigger pain in the ass than public investors.

Brian Halligan

We are much agreed.

Let's add that to the B-roll.

Jason Lemkin

Yes, I think VCs are a much bigger pain in the ass than the typical public investor, and slightly less of a pain in the ass than the public activist investor, right? I remember saying that one. Believe me, I'm pretty sure on that one, right? I'm going to defend us there, Brian.

Most times, the public guys are benign. Though I will say, in the last 5 years on the venture side, definitely the benign content has ramped up. The marketing in the last 10 years on the venture growth side has very much been, “We are benign. We are more benign than the public investors.” I think that might be misleading. That might not be true. That might be reliance on preference.

I think you're right that it's a healthy trend to realize that you can go public, you get this liquid stock, and it's not as terrifying as you think. You have liquidity every day, not just once a year by appointment only. I'm a big “IPOs should happen” person, and I think that—

Brian Halligan

Here's what's underrated about the IPO: it's very stressful. You're exhausted, but the day you go public is going to be one of the top 2 or 3 days of your life. It is an amazing day. You'll go back to your company, and 2 days later you'll have a party with your company. You will cry, you will laugh, you will hug. You worked so hard. There's something about that that is really, really special that people—

Jason Lemkin

Having said all that, I've only lived through it on the employee side, when it was great, right? The day after that was weird, going back to your desk because the world's changed. The next day is really weird.

But if I could IPO at $30B or sell my company for $30B in cash, I'd much rather run the company—don't get me wrong—rather than sell, right? But if it were just a financial decision, I'd rather have the $30B on one day than wait a decade for it to drip and dribble out, right? So there is a conflict there as a founder.

Selling is tough for 98% of founders. I'm sure if you ask Sequoia, only 2% of founders say it was the greatest experience of my life selling my company. But, man, getting it all at once if it's the same, even with a net present value, it's just a weird trade-off between the two: the journey and the economics. It's complicated.

I just want to butt in there, Jason, because you said there about getting it all at once versus drips and drabs. I'm super naive here, Brian. You said about it going from, like, 1× to 100× when you do go public. Is it drips and drabs over years? How does that—

Brian Halligan

Is it dripping and drabbing? If you can look at the way I do it, I sell the same number of shares every month since we went public. The reason I do that is I don't want people to think I'm signaling something. If I make a big buy or a big sell, it's going to signal something. It's funny: investors will look at it.

So I just have the same exact number of shares I sell every month. It's on autopilot, and it's drips and drabs. You're definitely right.

Do you know how much you have left?

Brian Halligan

Yeah, I've got plenty left because I got more shares. Here's the thing that I didn't understand as a founder: you get more shares as time goes on. I actually didn't know that when I started hustling. I thought the pie would get smaller. Actually, the pie was much smaller, and then it started growing a little bit.

Jason Lemkin

One of the things we didn't warn you about joining the podcast, Brian, is that every once in a while Harry basically asks you to disclose your net worth. You have to remember, it's like in a prison cell: you don't have to answer. You can plead the Fifth on anything, right?

Brian Halligan

No, Harry, I will not disclose my loss rate on deals or term sheets, nor my net worth. Just refuse. Point blank.

You handled that very well, Brian. Rory, what was your multiple on HubSpot?

Rory O'Driscoll

I'm not going to disclose that. It was extraordinarily good, and I'm extraordinarily grateful to Brian for allowing us to do the Series C, especially when I had whiffed the Series B and my colleague Rob, and then Stacey, took over and did the Series C and corrected my dumb decision. We were extraordinarily lucky to invest in them at something like a $70M pre-money valuation on a company doing $10M. Thank you, God, from the bottom of my heart every time I go to my little house. Thank you. Thank you, Brian.

For the backstory on Scale's investment—

Brian Halligan

It was the Series C, and it was bad timing. That timing matters more than I ever would have thought or studied in school.

It was in the throes of the recession back in 2009. Dharmesh and I got off the plane from Boston, landed on Sand Hill Road, and we were like, “We've got this. We're going to go up and down.” It was 20 meetings, and we went to all 20 meetings and got back on the plane. We were like, “We got no offers.”

Up and down Sand Hill Road, every big-name firm—every household name—said no. We went back and forth. We had nothing. So we were about to do an inside round at a slight uptick to our Series B. At the very last minute, Rory and Rob Theis had us in. We pitched them, and they gave us a term sheet at $66M.

So actually, the thanks goes from me to you, Rory, because you marked up our deal. I appreciate you.

Rory O'Driscoll

Synergy. We call it synergy.

Absolutely. No, yes, I remember that was in '09. We did HubSpot, Box, DocuSign, and RingCentral. The time to buy is when everyone else is not buying, especially when—let's put it out there—you're just little old Scale and you're not Sequoia, who came in after.

The world might have to end, and that's obviously a little tough for everyone, but it'll be really great when companies like HubSpot are grateful to get my term sheet seat at $66M pre-money. Just to remind everyone, at that point they were doing somewhere between $7M and $10M, doubling year over year.

Other than that, it was a tough decision. So grateful. Thank you very much. My multiple was excellent, Harry, so pound sand. It would have been even better if we had had the mechanism then to hold for longer after the IPO. I profoundly wish we'd had that as well.

Jason Lemkin

The smart money on HubSpot was in the public markets, since we went from $1B to $25B in a relatively short period of time. That was the play with HubSpot, and I think Sequoia is really smart. I suspect that was one of the reasons why they said, “Let's hold these companies after they go public.” It's backfired in some cases, but I think it's going to work over the long haul.

You couldn't hold, Rory. You weren't allowed to in your LP.

Rory O'Driscoll

It's a long story. We had a single LP. We had to distribute. It's a long story. Move on.

Don't you love the way I asked Rory for the multiple? He's like, “No, no, no, moving on, moving on.” And then Brian's like, “Let me tell you how he missed it in the B,” and you're like, “Oh, I prefer the multiple question. Let's go back.”

Jason Lemkin

No, I'll give Rory credit. As great a company as HubSpot was, it clearly wasn't a consensus bet at the time, objectively.

Brian Halligan

Marketo was the consensus bet.

Rory O'Driscoll

Marketo was raising at bigger valuations. They were always raising 3×.

Jason Lemkin

Well, they were more enterprise, Brian. That's the play. You’ve got to go more enterprise. Yeah.

6. Can AI Even Work for SMBs? Why No One’s Cracked the Code (Yet)

Marketo. It's Marketo, correct? Yes—Marketo. And it is actually worth riffing on that because we have the king of SMB on this call with Brian. I mean, you're right: the interesting thing about HubSpot was the way they built a $30 billion market-cap business in SMB, starting with sub-50-person employee companies. And you're right, Jason, it was very counter-consensus wisdom at the time.

Brian, any thoughts on that? What was different about it? Did you ever want to go upmarket?

Brian Halligan

The reason we bet on SMB was that I had spent my entire life doing the soul-crushing exercises of selling to CIOs. It is just soul-crushing work. I didn't want to do it. And I also felt, at least back when we started, that the internet disproportionately benefited small relative to large. Your success was much more about the width of your brain than the width of your wallet, and so we had sort of a play on that.

We also looked at the consumer business. Actually, the P&L is a shitty way to look at these businesses. Let's look at CAC and LTV. We kind of convinced ourselves that it worked. One of the interesting things about HubSpot SMB is that the other SMB company that's done amazingly well, even better than HubSpot, is Shopify. They're both outside of consensus land in Silicon Valley.

I think there's a big echo chamber in Silicon Valley and a big negative bias toward SMB, but you can make it work in SMB. HubSpot and Shopify have shown it. Block has shown it.

Jason Lemkin

Monday.com as well—massively anti-consensus. Totally agree.

Rory O'Driscoll

Yes. I think they're moving to enterprise.

Yes, they are, in a way that you guys didn't. Would you have made the choice today, Brian, in AI? The reason I ask is that to make a lot of great AI products work, you need training, you need forward-deployed engineers, and you need daily—even if it's SMB, every day someone's got to be your AI orchestrator. If you were doing all that today, would you go a little bit more mid-market? Because these SMBs don't have time to train their AIs.

Brian Halligan

Yeah. In fact, here's one of the interesting things about my life: all these founders come to me and say, “How did it happen with HubSpot? How did you do marketing?” We talk about our Website Grader, content marketing, and inbound marketing. “How did you take on Salesforce?” I kind of came up with freemium, and we learned a lot. We innovated a bit, and a lot of what we learned just wouldn't work today. You’ve got to keep innovating. You’ve got to keep turning that over.

I'm careful to give people advice and tell them what we did at HubSpot because it worked at the time. It was really innovative at the time. But a lot of what we did on go-to-market, a lot of the freemium stuff, a lot of the PLG stuff, a lot of the interesting culture stuff, and our SMB play—some of it works, some of it doesn't.

Some of it works, some of it doesn't, I think. Right?

Brian Halligan

Some of it works, some doesn't. Yeah.

Do we have Jason? I'm just literally—this is me exploring, not knowing the answer. What are the great SMB AI products? Is there a great SMB AI product out there today where you go, “That…”? Well, Lovable's a stupid question, Rory. I mean, you could argue that what you seem to imply is that it's harder to do enterprise-kind-of-SMB-level apps in AI because of the need for training data. Then I'm trying to think: what are the mass-adopted enterprise apps? Obviously, you're the Lovable and Replit guys, so there are some. I don't know others.

Even the exceptions sometimes prove the rule. I mean, I'll tell you, I learned a lot of AI from Brian's AI. I copied it and made it better, right? When Brian built his clone, he and I were talking in the very beginning. He's like, “Well…” I'm like, “Brian, it's pretty good.” I had some fun, cathartic conversations with Brian's AI, and I had some fun ones. I'm like, “It's pretty good. Mine's better, but only because it's trained on more data. Mine is better.”

Jason Lemkin

But I didn't get it. I asked Brian, “Why is yours so good?” He said, “I spent a lot of time training it.” This was 6 months ago. That's a lot of time in AI, and I didn't get it, right? But now, when I look at every true AI company, it's hard to train it.

This is a quandary for SMB investments, right? What I'm looking for in SMB companies and AI is: how do you self-train? How do you solve the unsolvable issue? How do you solve the fact that it can take 6 months to roll out a Palantir-grade deployment? How do you do that in 60 seconds? Any founders that crack that code, I want to invest this hour, this second.

I'm pushing every startup I work with that's SMB to be more AI, and sometimes they push back on this, right? But you’ve got to do it. How do you train? How do you train the training?

Rory O'Driscoll

The stuff I see, Jason, is that there's a ton of prosumer stuff that's working. Lovable's working, Gamma is working. So many prosumer things seem to be working, and then enterprise is working. You're right, I haven't seen much in between.

Jason Lemkin

But there's no training in the—I mean, we use Gamma too. We love Gamma, but you don't train Gamma. You do train Gamma a little bit, don't get me wrong, but you sort of accidentally train Gamma by uploading your templates and your things, right?

Replit and Lovable—I mean, I'm Mr. Vibe Coder, right? Those are AI under the hood. The training is weird, though. It's AI under the hood. I just think this AI B2B SMB is something that hasn't been cracked, to Rory's point, and I think we should all just rush all our capital into that.

Listen, if you can train an AI, it's easy. HubSpot just put out this report on AI with SMBs, right? It said something like 80% of folks have an AI team to do this. Even with what they call SMB and VSB, 80% have a team. SMBs don't have a team, right?

The owner doesn't. The owner of a restaurant—one restaurant that Brian and I are invested in—they don't have an AI team, right? At Brian and Jason's sandwich shop, there's no AI team.

As I say, I opened the door to a non-prepared topic, so I'm winging it, but my sense is two comments. One is it may well be that it takes a year or 2 longer, because if you think about even the HubSpot journey, what tends to happen is the big companies with loads of money fart around, kind of mentally defining these apps and figuring out what they should be, because they can afford to. Then, where the features lock in, the SMB guys go, “Oh, we'd like that,” and you don't have to do as much of the figuring out what it is.

In the case of this generation, you probably come with a very much pre-trained app where, if it's call answering or something like that, most of it's already done and you just have to configure it at the SMB level. It may be over the next year or 2, because I do believe—and this is, I think, one of the reasons we love HubSpot—that anything the big companies have, the small and midsize companies want too. They're not different; they just need it packaged tightly and priced tightly so they can consume it in bite-sized chunks.

7. Meta’s Monster Quarter: Growth, Cash Burn, and the Real AI Strategy

Over the next couple of years, it may well be that for things like phone answering and simple order dispatch, there'll be a whole bunch of pre-trained, pre-baked systems: “This is how it works, Mr. SMB. Just turn it on, and you, too, can sound like a big call center.” So I'm optimistic. It won't be trained on a company-by-company level, but I think it will deliver big-ass value.

You said there about big companies with lots of cash. I do want to progress this, because there's something I'm fascinated to hear your guys' thoughts on. Meta—what a freaking ripping quarter. It was a 38% year-over-year increase in adjusted EPS, 22% revenue growth, but a 22% drop in free cash flow. How did you guys read it? How long does this go on for? Is this the start? Is this near the end? How much patience do people have?

Brian Halligan

It's funny you led with the “How long does it go on for?” because 2 weeks ago, when I said, “How long does it go on for?” you looked at me like I had 2 heads, right?

You see the shit I put up with, Brian. I just take it.

Brian Halligan

No, I'm polite. I don't give it back. I just take it.

I will say, on all these things, the takeaway—and I like the way you framed it here, and not the way you framed it in the note you sent me prior—is this isn't an AI-enabled success. This is: I have an awesome existing business, and it kicks off so much money that I'm allowed to spend that money on building this great AI vision. I can probably do that for as long as my existing business kicks off cash.

My big aha from this week's earnings—and I would say I got broadly AWS right and I broadly got Microsoft wrong—is that not all the AI stuff is working for the hyperscalers. My big takeaway is that all their existing businesses are working so well and kicking off so much cash that they can keep doing this for the next year. They said they're going to keep doing this for the next year because they want to play in the new game. That's the takeaway.

Real men with $700 billion in market cap and $70 billion of free cash flow get to spend $40 billion of that on servers. It's a great country.

Is there any nervousness at Sequoia, Brian, at all, that the good times might end soon? Is there any draft Sequoia memo, version 3, ready to go out? Just search and replace with Gamma—just have Gamma, please. Dust off the “RIP good times” and update it for AI.

Any discussions at the partner meetings you’ve been in about that? I mean, David Cahn wrote the piece about the gap between capex expenditure and revenue.

Brian Halligan

Yeah, I think we’re in a kind of—one of the questions is, are we in a bubble or not? The argument against the bubble is: look at Anthropic and ChatGPT. The growth rates are ridiculous. Or look at Harvey, or so many of these companies that are at the app level.

As an investor looking at this stuff, I get nervous about tech companies selling to tech companies and Silicon Valley companies buying from Silicon Valley. There’s a lot of trading going on, and there’s a lot of growth in there. I like what ChatGPT is because it’s mere mortals using that thing. I love what Harvey’s doing; they’re selling to lawyers. I like what Rogo is doing, selling to investment bankers.

I get nervous because I think 2001—and definitely 1999 and 2000—it was just Silicon Valley companies buying and selling from each other, and it created kind of—

2022, though. 2021 too. A lot of it—

Brian Halligan

For sure. Yeah, for sure. So I kind of like these ones selling to mere mortals—

But it has to be a bubble at some level. The capex—this is the capex bubble. It can’t last forever, right? We can’t—

Jason Lemkin

That’s Harry’s point. Look, AI is bigger than the internet, most likely, right? Bigger. So the investment makes sense, but when you see Meta’s cash flow decreasing, there’s some kind of bubble here, right? Hopefully we all get out, but it’s some kind of bubble.

Rory O’Driscoll

“Bubble” is a laden word, right? It implies—I just think you invest. It’s the usual two things. There’s an enormously enabling technology, and it’s getting massive traction at the apps level, as Bryan mentioned. But you still add up all the apps’ revenue, and it probably comes to $25–30 billion maximum, while the capex to build that is running between $400 and $600 billion.

So you’re investing $400–600 billion a year to enable a $25 billion ecosystem to go and keep doubling. Maybe next year it’s $50 billion. My takeaway is that the long-term trend is almost certainly real. If you fast-forward 10 years, that $25–30 billion of apps revenue could easily be $300–400 billion. So, in the long term, it’s not a bubble.

There’s probably going to be a period where things get ahead of themselves, right? The marginal player will get caught, just like the marginal player got caught in 2001. The overlevered player who’s taken on too much debt—financiers will get caught and get burned. The big guys will retrench for a year or two and then just grow into it.

That’s the most likely version of the movie: some pain at some point in time. It’s boring because it’s not amazing AI maximalism, and it’s not bubble doomerism. It’s just that we’re doing what we always do with a new technology. We’re spending like crazy because it’s the only way to discover the frontier, and until you discover the frontier, you’re not investing enough.

We’re doing it as an organization, or as an organism almost. Capitalism is working. We’re spending money trying shit that works. Some of it won’t work, but unless you try, you just end up like Europe. Sorry, Harry, I just shit on you. You’re so Irish.

We’ve had 800—

Jason Lemkin

He talks more European than you do, Harry.

Harry, we’ve had 800 years of you guys shitting on us. Every chance I get to shit on your back, I’m going to take it. So just get used to it, man.

Aside from the navel-gazing on this, because I don’t know what the answer is, I would just say I’m incredibly impressed at massive scale with how fast Microsoft is growing and how fast Meta is growing. The companies people aren’t talking about are the very old-school Microsoft, SAP—and how about Oracle?

Jason Lemkin

Oracle.

Absolutely. How about Oracle? My date is August 26, 2022. That’s when everything kind of hit the bottom in share price. If you go from that date and you look at Oracle and SAP, they’re both growing about 230%.

Jason Lemkin

Are you talking stock price or revenue growth rate?

Stock price.

Jason Lemkin

Yeah, you’re becoming a stock-price baby.

Yeah.

8. CEO of the Year? Why Jensen Huang Leaves Zuck & Satya in the Dust

Jason Lemkin

The lower you start, I mean, I think part of that is a function of the fact that they started at a much lower pace. But I think you are right, though. The stunning thing in Oracle’s particular case is how they’ve, rightly or wrongly, taken that free cash flow and invested it in GPUs and have now made themselves relevant in cloud. Provided that market keeps growing, that’s clearly worked for them so far.

Rory, I know you love unfair questions. I’m going to give you the chance to grant a CEO of the year award, and you can grant it to Satya or Zuck. Which one do you give CEO of the year to?

Rory O’Driscoll

You know, I’m deciding I’m now going to be the new humble me. It’s inappropriate for me, a little grasshopper, to comment on which of those two amazing CEOs is the best. They’re both GOATs.

I think the CEO of the year is Jensen.

Yes, that’s—I’ll go with that.

[Speaker?]

Yeah. And I like him. But the other thing I like about Jensen is that he’s rethinking the role of the CEO. He’s rethinking the CEO playbook. I think he’s a pretty good inspiration for CEOs out there today. I’m a huge—

I’ve certainly gone as long as you can go, right? He’s the number one on my CEO list—

Jason Lemkin

Because I think, of the two, the reason the two people you cited are not—well, in one, it’s very much an existing business carrying and doing something in the new world, whereas with someone like Jensen, or Sam at OpenAI, or Dario, we’ve created the new world.

I think the two people who have the two categories in the stack that didn’t meaningfully exist at scale in SaaS and cloud land, and that exist now, are the GPUs, which is all Jensen, and the models. Neither of those categories even existed. Obviously, not only do they exist, but they appear to be dominant relative to the other parts of the category, either apps in AI land.

So I think Rory’s right. Those are the contenders for CEO. The other guys are contenders for managing the cash machine brilliantly at scale and keeping it up and to the right, which turns out to be a pretty lucrative way to spend your adult life.

[Speaker?]

I’ll tell you why you absolutely have to go for Satya over Zuck, by far. It’s a simple—it’s more of a structural reason, and this is certainly what I learned in my tenure as a VP at Adobe and as a founder. This stuff’s easy. You just call the troops together. Zuck can do what he wants.

Satya having to get—I just watched Adobe trying to go to the cloud. It took 3 years of convincing everybody. What Satya has done—inviting Sam Altman back in, doing the deal, managing the deal, investing the money, doing this kooky deal to buy 49% of OpenAI, investing all-in on Azure for AI—he doesn’t have the power to do this on his own, right?

I mean, Brian and Dharmesh can get together and honestly, you guys can just decide what you want to do. I know I’m being simplistic, but I bet you’d agree. For Satya, the amount of meetings and orchestration and stuff you have to do as a non-founder—it’s much harder.

He’s like a refounder. He basically acts like a founder. He gets stuff done like a founder. It’s super impressive.

Brian Halligan

If you accept the constraint that you can’t build the core technology internally, which is what Microsoft had to accept, and he accepted, right, execution since then has been perfect. You found the only other people that had the technology. You gave them a convoluted deal, you sucked a lot of value out of them, and you have the ability to resell it. All those things are awesome.

At some level, you must kind of wish, as the CEO of Microsoft, that there’s a little part of you that says, “If my guys were only smart enough to build the shit that OpenAI was building, I wouldn’t have to do all this crazy stuff.” But maybe that’s the nature of being smart enough to accept that this large bureaucratic company can’t get it done and live with it.

Maybe that’s the answer to why it’s awesome. He’s lived with the reality of, “I wish my people could do this, but they can’t, and I’m not going to keep banging my head against the wall. I’m going to do this very hard thing for a non-founder to do. I’m just going to cut this weird deal with these other dudes, give them $10 billion, own 49%, and basically ride into the AI business without actually having the core model that you should have had to be able to do it.”

It’s kind of a one-man show. Literally, he must feel—maybe the real sound in his head is, “Thanks a fucking lot. The rest of you guys—I had to figure this out with one BD guy while all you guys were sitting on your ass not shipping AI.” Maybe that’s what he deserves the medal for.

Jason Lemkin

Not easy.

Brian Halligan

Not easy, and he has to suffer and take 49% of the losses.

Jason Lemkin

Massive losses flowing through their financial statements, right? You could argue asterisks and daggers, but it’s not cost-free. There is a cost, right?

Yes, but 3% of OpenAI—they’ll be fine. I mean, 3% of OpenAI in return for, you know, probably a trillion dollars.

You get criticized for this as a non-founder CEO. You get criticized for every line versus Zuck. They’re like, “What are we going to do?”

Jason Lemkin

Just being ballsy enough. You’re right.

What are we—what are we going to do with Zuck? What are we—what are we going to do?

Jason Lemkin

You’re exactly right. Just being ballsy enough to write a $10 billion check for something weird is, in and of itself, heroic.

Bryan, do let us know when you have to go. I know you have the time.

Jason Lemkin

I have to go. No, no, no, no. That’s not how it works.

You’re the best, man. Thank you so much.

9. Cognition's $15B Deal & Mass Layoffs: The Most Savage M&A Move of 2025

Jason Lemkin

You’re good. Thanks for inviting me.

Guys, I do just want to go to a couple of private rounds, just because they’ve really stood out to me. We’ve said, about companies that we’ve talked about before, Cognition is now rumored to have done a round at $15 billion. The new combination is obviously Cognition and Windsurf. Both had $85 million in revenue, so combined, you’re at $170 million, being priced at the new $15 billion. How did we think about this?

Jason Lemkin

It didn’t move a ton, in the sense that what you’re saying is the rumor was it was being done at $10 billion, and now the rumor is it’s been done at $15 billion.

I mean, you know what? The cynical take is pretty much the same as the Anthropic rumor. It was done at $100 billion, and now it’s being done at $170 billion.

Jason Lemkin

I think what it says is demand is high for premium assets. It’s a little like the private IPO: You float a price of $100 billion and you end up at $150 billion. You close at a price of $10 billion, and you end up at $15 billion. I think there’s just a lot of demand for premium, perceived-premium AI assets, and price is how scarce assets get allocated. So, there you go.

What’s the question? I just thought this layoff-and-buyout thing was just crazy.

Jason Lemkin

Separate question, but yes.

Well, unpack that, Jason. I guess today Cognition laid off 30% of the folks they bought, and they offered to buy out all the other 200 employees. They gave them a 9-month package. They told them they either had to work 80 hours a week, 6 days a week, in the office, or they should take a 9-month package.

Jason Lemkin

And listen, no criticism—great people here, right? I’ve had some of my best portfolio companies use Devin, which is pretty interesting when I talk to them, right? But doing this hero acquisition, then laying off 30%, and then telling everybody to either work 80 hours a week in the office or take a 9-month package, the story keeps getting more complicated. Then finding out that the founders and the investors put in the $100 million—Google didn’t—the story is more complicated than it looked at first, right? There aren’t quite as many white hats, and everyone’s a gray hat, it turns out.

Okay, but let’s just unpack that. Why make the acquisition, then, if they’re going to get rid of 30% and then say, “Hey, all of these terms are posted if you’re buying the team”?

Jason Lemkin

I don’t think they’re buying the team. I think it’s clear they weren’t buying the team. Listen, my limited experience with Devin is that it’s an AI engineer, okay? I talked to all the folks in my portfolio. The 2 actually toughest-problem CEOs are using it, okay? Stuff where they’re reluctant to use AI. They like Devin, but it only does a little bit. They pay for it, it’s fine, but they’re not deploying it across their whole team like Claude Code or something.

So, they have a niche product that’s done well, and they want access to a top platform to get into everybody—a broader platform. They bought a brand. They bought $80 million of revenue to maintain, right? And they saved themselves, who knows, 3 months to 9 months. They basically did a deal that looks non-dilutive at the end of the day at this $15 billion.

Maybe there’s just a lot of spin on a deal that was just for brand and an accelerated market entry. I don’t know. The fact that they’re offering to let every single employee go certainly means they don’t see a lot of value in the folks that are left. There’s an implicit cultural statement, and again, they’re basically saying, “We’re all working 9 to 9, 6 days a week, and you guys aren’t. If you want to sign up for this, do; if not, leave.” I mean, you’re right, there’s an element of clarity to it.

They’ve got until August 10 for all employees to decide whether they’re staying or going. It’s a lot of change from the last pod.

Jason Lemkin

But all predictable, I mean.

Bye, guys.

I’m not laughing because the human implications here are so stunning, right? I mean, Google’s such a jerk: “We don’t want anybody. We want the company to die.” Then the investors and the founders having to take the $100 million out of their own pocket and leave it in Windsurf. Google didn’t do it. They had to do it after the deal was handshaken, apparently.

Then it all happens. Then they get bought at the 11th hour, and now everyone gets a buyout package. It’s just—I mean, even in a 9-to-5 world, it’s too much.

Jason Lemkin

Yeah. And what you don’t know is, because they all had the common part of that, they got their equity cashed out. What you don’t know is how that compares to taking that $100 million, closing the company down, and splitting the money. So, yeah, you don’t—I genuinely don’t know how they ended up.

But what it says is this: If you’re not one of the key players, the minute you move away from the cap table to making it up as you go along, you’re very vulnerable. You’re basically depending on the kindness of strangers, as Blanche DuBois would say, which is always a mistake, right? You’re depending on arbitrary people deciding, quote-unquote, what fair is, and people’s decisions on what fair is change over time, right?

Versus in a normal M&A, where you know where you stand. It’s a Delaware corporation or a Nevada corporation, and you get what you get. I think the lesson here is, once you get away from that, everyone’s just winging it, and it’s hard to know from the outside.

Did you quote A Streetcar Named Desire?

Jason Lemkin

Yes, I did.

In a VC pod? My respect and love for you has gone through the roof.

Jason Lemkin

There you go. There you go.

That is fantastic. You know what?

10. Ramp’s $22B Raise: Genius Move or Suicide Round?

Jason Lemkin

Just a little bit of the humanities, right? Well, now that computer science is on, I’m going to go back to doing my English lit exam.

Yeah. Well, there are 2 more things that I want to discuss. Ramp raised a $500 million Series E at a $22 billion price. Iconiq led it. This is like the 5th or the 6th round they’ve raised in an 18-month period. So, is this in line with company progression, or is this late-stage capital really trying to find a home?

Jason Lemkin

Probably a bit of both, because remember, Ramp, unlike most software companies, part of what they do is lend money. They basically give people corporate credit cards, on which they earn the interchange. The way you earn the interchange as the issuer is that someone has to fund them for that period—the 30-day float period.

Now, I don’t know in their case: Are they funding it themselves, or do they have some kind of flow-through? But in any event, issuing a corporate credit card, by definition, is way more capital-intensive than simply building a software company, and they’re building both. So, it probably consumes more capital than the typical software company. The faster you grow, the more capital you consume. At some level, there’s going to be just a capital need for that.

On top of that, you obviously have the phenomenon of it being a wildly hot, successful company, perceived as dominant in this category. By virtue of that, it’s just going to attract a lot of venture capital interest. If people keep offering you money at increasingly higher prices, you’re probably going to take some. So, probably a bit of both.

The other thing with these rounds, when I look at Ramp—$22.5 billion, very quickly, but $500 million—it’s only 2%, right? Clay just did a round at $3 billion, which is stunning growth too, right? But they sold $100 million, right?

So, these little, tiny rounds—the absolute dollars may sound large, but they’re not even rounds when we’re talking about 1% dilution, 2% dilution. The VCs get a markup out of it. There’s also a weird set of questions about whether they really count.

I remember back in the day, I had a markup at a $3 billion investment, right? It was a very small number. My anchor said, “Don’t recognize it. It’s not big enough. It doesn’t count, right? It doesn’t count. Don’t. Just don’t.” And he was right. I’m not saying that’s the same here, but if I’m Ramp and I could sell 1% in a series of rounds, there’s no cost, is there?

But do you actually think these are good rounds for companies? I think they’ve been called suicide rounds before, where you raise $100 million at a $3.1 billion valuation. It just sets a really high price for the company to grow into, with actually not that much money going into the company.

Well, that’s the downside, right? Maybe it is a suicide round.

Jason Lemkin

You know, I think if you need the money, you have to get the money. And if you need the money at a high price, it’s better than a low price. I don’t think raising money at too high a price is only “suicide” if you have to raise again.

If you don’t have to raise again and all you do is have some investors who’ve overpaid and they take 4 or 5 years to grow into that valuation, well, that’s tough shit for the investors.

[Speaker?]

But from the company's perspective, it's fine, and you're glad you got the money, right? So I think the number one thing you don't want to do is raise $1 or $3 billion when you needed $400 million, and then go back out 6 months later. You haven't had the growth, and in theory, you're only worth $1.5 billion, but then people get the cognitive dissonance of, “It's a down round and you're screwed,” right?

That's a suicide round.

That's a suicide round. But if you raise $500 million at $2 billion, and even 3 or 4 years later you go public at $1.5 billion, well, tough shit on the guys who paid $2 billion, but life goes on.

Ramp's raised $1.9 billion, so it's just going to keep consuming this capital for one reason or another, right?

[Speaker?]

And as I say, you can actually work it out if someone had the time. I've heard they're doing roughly $700–$800 million. Interchange is a good slug of that, and you get 2.5% on interchange, so you can work out their total transaction volume. You have an average, probably, of a 15-day rotating balance.

So you probably have 4 to 5 times revenue in terms of floating cash amount. In other words, to do $700 million in revenue, you might have a $3–$4 billion capital requirement because you're floating all these—I mean, you're replacing Amex. You're floating all these guys on their credit cards. So you do need the money.

I mean, I get like 20 emails a week from Brex telling me to deposit more in my account.

Yes.

Jason Lemkin

Maybe it's not a coincidence. I'm constantly getting, “Jason, your borrowing may go down below $2 million. We need money instantly today.” Leave me alone, guys. Just leave me be. I'm fine.

That's exactly right. No, I mean, in the end, fintech companies are fintech, and at some point, you need—

One of the non-negotiables is to have low-cost capital, and at the moment, bizarrely enough, venture equity is lower-cost capital than pretty much anything out there. As Jason said, “2% dilution or get a banking license.” Hmm. I'll do the 2% dilution.

Going back to venture land, CRV raised $750 million. They shrunk the team, and they're not raising their late-stage Select fund. Is this a sign of a more rational venture landscape? Is this a sign of LP appetite being less willing to do opportunity funds? How did you guys think about and analyze this move?

[Speaker?]

I think they do early stage well, and they probably decided the best way to make money is to do the thing you do well, then do it well and keep the message clear. I actually thought it was very smart of them, right?

There are some firms that are pulling off these multiplatform strategies, but it's just a step-function increase in complexity, right? If you can do it—and obviously we all know the names who have—great, you have a multiproduct firm. But if you're going to be marginal at it, the non-negotiable thing is to at least do 1 thing well, right?

CRV clearly decided that rather than muddying the waters trying to do this multistage strategy thing, they would just execute really well on great early-stage investing. So I thought it was probably smart in a world where you just want to have a clean message.

And to your comment on whether it's a sign of wider LP appetite, no, I think what you're seeing is that LP appetite is varied. You can say, “Hey, I'm really glad, CRV, that you focused your message down,” and the next day you can say to Founders Fund, “You've got the most amazing growth fund on the planet. Let me give you another $1 billion.” The day after that, you can say to Elad Gil, “You're just amazing. Let me give you $1.5 billion on your own.”

They're trying to figure it out too, and maybe the takeaway is that the whole industry has changed so much that there's a lot of different ways to play the game. More than anything, I think they want to see that people know what game they're playing and play it well, right? Rather than trying to have envy of someone else's game.

By CRV saying, “Hey, this is what we do, and we're doing it well,” you can say, “Okay, I know what I'm getting from that,” right? Does that make sense? I wouldn't want to—

Yeah, versus keeping it clean. I mean, look, the other thing is that now we're in an age of everyone raising as much capital as they can and deploying infinite capital as startups stay private forever.

Jason Lemkin

But deep down, if you're in it for carry over fees, if you want to get into carry mode faster, I'd rather have 2 $750 million funds split in half, like Founders Fund did, than 1 $1.5 billion fund, right? It's just better for GPs, isn't it? I'd rather get into carry mode faster.

I don't know CRV's results, but they've had some good investments. If they're looking at it, especially some partners who maybe generated more carry than others, they're like, “I'm not in it for $1 million or $2 million a year in salary. I'm in it for big carry checks, and I want to get this thing deployed in 24 months, 30 months,” right?

I know we've lost this over the last 18 to 24 months, but if we look back on all of history, in normal times, you want to optimize your fund size to achieve the maximum carry you can in a given time, right? Then just go raise another. In an ideal world, you might even raise a fund a year, so you could get into carry mode as quickly as possible.

You lose a lot of things. You lose time and other things, but you want to get to carry mode fast. You don't want to leave it all to your grandkids, do you?

Jason, come on, man. Do you regret doing an opportunity fund?

Jason Lemkin

I don't regret it because I'll make money, but it wasn't worth it for me to do the opportunity funds. It's not enough. I'll make maybe 15% more money, but it's not enough money. If I had a $500 million opportunity fund and could deploy it, that'd be different.

So maybe CRV looked at it, and maybe they did some of their deals that didn't make a lot of money. It was a lot of hassle, and they're like, “Hey, I only made 10% or 15% more carry. My LPs don't love it because I burned a lot of capital from them. Let me concentrate where I make a lot of carry.” That's my guess.

And for me, Harry, it's the same thing. Ninety percent of my carry will come from the main fund, so I'm like, “I don't want the drama in my life.” I'll still make maybe 3x, but that's it. It's not a lot of money.

I think that's what the math, interestingly enough, runs out to, because you look at it and you go, “How many deals do you have in your main fund? How many of them are amazing? How much can you get in?” Maybe only 20% of them are amazing. How many of them can you deploy late-stage dollars in? Maybe only half of that, because the late-stage rounds get pricey very quickly.

It turns out that unless you end up with one of the very few companies that are not just amazing but super amazing, where they can be a $20 billion outcome, your ability to deploy lots of capital relative to your early-stage fund is actually much smaller than you think. So the size of the opportunity fund that you can deploy just within your entities is smaller than you think.

And you're right, Jason. Then you end up saying, “Is it worth it?” Now, you can decide, as some people have, “No, I'll build a whole late-stage growth strategy,” and knock yourself out. You can do anything and put billions to work, but then, at that point, you're becoming a different thing, right?

Plus, maybe less discussed, maybe CRV is too big for this, but almost every seed manager that we know that's been successful can spin up an annex fund. It's not a permanent decision to not raise another vehicle. If all of a sudden you got into Anthropic early and it's turning out pretty good, you could raise a couple hundred million in an annex fund or an additional fund.

Do you not think, though, that Benchmark's capping or discipline on fund size is one of the core reasons why maybe they've lost some talent this year, when you look at the likes of Miles or Victor leaving?

I don't know if I agree with the characterization. I think they've had—I mean, Jason did the list, actually. Either you or Jason did the list last, Harry. They've done amazing deals. If being left behind is that set of deals that they've done, I don't think that's challenging.

Jason Lemkin

Right.

But you're both right. You're both right. The names that they've picked—extremely well, right? Incredible investments. At the same time, if you look on social media, Benchmark isn't listed the way Andreessen and Sequoia were a generation ago. It's just not. Does it matter? Harry's built a big brand. He's concerned it matters. You could argue both sides, but when the industry was smaller and Brian Halligan had to drive up and down Sand Hill to get a deal done and it took months, brands were just different, right?

It's still an A-tier brand, but it's not in every conversation on X, right? It's YC, Sequoia, Andreessen—that's it, really.

Do you not think that if you were scaling fund size and scaling strategy, that would have increased the likelihood of being able to keep great talent like Miles and Victor from leaving?

[Speaker?]

I think it's presumptuous.

[Speaker?]

I'm not going to tell Benchmark how to run their business. They've done it. I've been in this business. I remember them starting in ’95 and talking to them. They've done a pretty damn good job of running that business. They don't need my help.

I think it's what Jason said. Stepping back from the individuals, the meta question you're asking is—let's stipulate, because I think it's true—what we're asking is: Is the very best specialist fund able to compete in this market with the very big, full-stack firms? That's really the meta question you're asking.

I'm willing to stipulate, and there's a lot of data that says it, that Benchmark has been among, if not the best, specialist funds. So I think this takes it away from commenting on individuals, which gets personal very quickly, and more toward the meta question: Is the right strategy a specialist fund, or do you need to be a full-stack player to matter?

There's no doubt that if you're full-stack, you have more coverage, you have more news, and you have more news flow. There's also no doubt that—I cited the data from the Rotman guy at DST a while back—your picking goes down a little, but your volume goes up, right?

I think the biggest risk you face as a specialist is that you get crowded out by the noise, and people don't know you're amazing enough. Therefore, you lose some of the at-bats to the people who have more brand. The risk you face as a big brand is that, in your wild urge to put all the money out, you end up overextending yourself, and you get subpar returns.

You fast-forward 5 years, and you look back and go, "Oh, we had lots of noise, lots of good individual deals, but as Jason said, it didn't add up to compelling returns because we had so many other deals." I think the truth is both strategies will work if executed well, and both strategies have their risks. I know that's kind of a stupid answer, and you're right, it is, but I think there's lots of ways to make money.

The one thing you don't want to do is be inconsistent. You have to have a strategy that plays to your strengths and that can work for you, and you have to understand the risks that your strategy entails, including the risk of a specialist strategy. We see it every day. When you're competing against the guys who have infinite deals, infinite deal flow, and infinite money, it's hard, and sometimes you lose, right?

Equally, for those guys, sometimes you put $100 million into something and it just doesn't work. Unless you have an outlier to cover all those mistakes, that's going to be their problem.

Guys, I want to wrap up with one final question. We had Halligan today. He was fantastic. Who would you most like to have as our VIP guest on the show next time?

Rory O’Driscoll

You guys are so much better at this than me because you interview for a living. I don't have an opinion. I thought Brian was an excellent guest because he brought a different perspective to bear as a CEO. He was strong on the things he knew. Obviously, we knew him from the deal.

Both of you have done way more interviewing than me because I do none. So, by definition, you do more. Whoever you two think, I'm totally willing to try. How about that?

I think it would be great to have Marc Benioff. I think he would do it. He'll be different from Brian, right? But my idea that I didn't have until today was Jeff Lawson.

I thought the same. I thought Jeff would be fantastic, dude.

Didn't occur to me until today. I would love to hear all of his reflections, all of his thinking. I mean, he's a founder's founder, and that hadn't occurred to me. It's a great idea, right?

Rory O’Driscoll

And how would we, in fairness to him, not make it just be everything you learned getting fucked up? You want the activist story, but you don't want it to be a celebration.

He's rethinking everything from his stack in the age of AI, right? Let's go for it. I'm going to do Jeff and I'm going to do Benioff, and let's get them on in the next couple of weeks.