Transcript: (disclaimer: may contain unintentionally confusing, inaccurate and/or amusing transcription errors)

Ben: Hello Acquired LPs and readers of The Generalist. For anyone from The Generalist audience, I am Ben Gilbert and I'm the co-founder of Pioneer Square Labs, a startup studio and early-stage venture firm in Seattle.

David: I’m David Rosenthal and I'm an angel investor based in San Francisco.

Ben: We are the hosts of Acquired, the podcast about great technology companies and the stories and playbooks behind them. This is a first for Acquired on a number of fronts. It is the first time that we've covered a company just before its IPO. It is the first time that we've done a show like this where we cover a full company. Just for folks who are members of the Acquired Limited Partner program. It is the first time that we are partnering up with Mario Gabriele, the person behind The Generalist and a frequent source of Acquired research, the S-1 Club.

Before we introduce Mario, here's a little bit about the company we are discussing today, Roblox. Roblox might be the most important company of the next generation, those 16 and under. A full 50% of kids under 16 have used the service in the last 30 days, which is a crazy, insane stat that just blew my mind in the research here. It was supposed to go public just over a month ago at a $3 billion market cap, but instead, they decided to not do that, take some private investment at just shy of $30 billion valuation—that's 10X.

There are now rumors that it’s going to go public in a direct listing very soon here at double the share price of that December round, meaning it could be a 20X compared to their valuation from December. The company's fascinating that we'll talk about today but just like a nutty moment in our financial system where this is possible for this to happen.

David: Totally. I think it was $4 billion and I think that was the last private round, which was somewhat less than a year ago.

Ben: I apologize.

David: But we dramatize for effect here on Acquired. By any measure, this is nuts.

Ben: Totally. The question becomes about this company, is it a platform for games? Is it the next generation of what social networks look like? Is it a remnant of an esoteric game designer thing from 15 years ago based on a physics engine? The answer is probably all three of these things, so we're going to dive in here today on this odd and different LP episode and crossover with Mario. David, tell everyone who Mario is.

David: Yeah, we are so excited to have Mario here with us. We were introduced by a friend of the show and former LP show guest himself, Jake Saper over at Emergence Capital. We’ve just been super impressed with Mario's rating over at The Generalist, which is less than a year you've been doing this now, right?

Mario: Yeah, that's right. It depends, I guess, on how you calculate it. I was probably writing it a little bit on the side for longer than a year, but full time on it since August of last year. It still feels very early innings.

David: Yeah. You've built a great audience. We’ve done some fun collaborations. We did a clubhouse together with Packy McCormick the other day. We were jamming with Mario a couple of weeks ago about the stuff we could do together, and Roblox just seemed like such a natural topic because of all the buzz around this IPO. Of course, you spearheaded the S-1 Club and published the S-1 Club analysis of their S-1. It just came out last week, and I think it's basically the bible on the topic. So we thought—certainly I am thinking about—is this a stock that I want to own, make part of my portfolio?

Who better to talk about this than you? We thought let's just hit record and share it with everybody. We’ll link to the S-1 Club piece on Roblox in the show notes. For people who want to follow along, but go over to readthegeneralist.com, no longer on Substack. You got your own platform, owning your own audience, I love it.

Ben: Going even indier than going indie.

Mario: So indie these days.

David: So indie, I love it. It's great. We’re members, we’re premium members, so worth it. Everybody go check it out.

Ben: David, before we dive in, I do want to say this is not investment advice. We are not professional stock pickers. We may have interest in the companies we discuss on the show in this episode. I just wanted to tell everybody that. With that out of the way, and this being a hugely important IPO or actually direct listing, we did want to discuss all the different lenses to look at this company including the bear and bull case. David, with all that out of the way, takes us in.

David: Well, unlike a regular Acquired episode, we're not going to do the full history and facts, but we have to start with a few. Mario can jump in and help us here and then we'll go through all the great analysis that he did. 

This company—it's crazy as Ben alluded to in the intro, November 19, 2020, just a few short months ago—filed their S-1. And then less than a month later after the Airbnb and DoorDash IPOs went crazy (as we discussed on this show), they pulled the listing, said that the market conditions—it's not that the window had closed. The window was too open. It was too open to go out and instead did this private financing at just a hair under $30 billion valuation raised over $500 million. Led by Altimeter and Dragoneer, announced that they would go public in a direct listing in early 2021. 

Rumors are that it could be as soon as a couple of weeks from now, sitting here at the beginning of February. It's quite a story. We're going to step through each section of the S-1 Club piece and start with the history, as always. Mario, if you wanted to jump in and take us from here. Who is Dave Baszucki, the CEO? What is his background? How did we end up with this crazy Lego/Second Life thing?

Mario: Someone who I think knows David fairly well described him as the first true James Halliday character from Ready Player One, sort of this massive maker and dream of a fantasy world. You can see the early signs of that in Interactive Physics, which is the company that he and his brother founded in 1989. Interactive Physics basically, for those who are in high school around that time might remember it as this little wonkish game where you could set up a ramp, a pulley system, or just these different objects that simulate movement, momentum, and so forth.

The idea was that you could get a sense of the actual forces of physics. It was supposed to be educational but ended up getting used quite a lot as essentially a game. Baszucki even said, "Users of Interactive Physics software use it for fun rather than school! Kids would build all kinds of funny contraptions with the product." That really ended up setting the tone for this builder game world that became Roblox later.

Ben: Yeah, it's funny how with some of these things, you need to put software out in the world and then just watch what people do with it. They had no intention to build a game system, but you got to follow the use case when people are using it for that.

Mario: Yeah, exactly.

David: It's like the old TI calculators, which I'm sure kids don't use anymore.

Ben: I'm sure they do. There was some kind of weird regulatory capture monopoly type thing there.

David: Well, I know they have an app now that emulates the 82, whatever it is, the TI whatever.

Ben: Point being, the TI-84 or the TI-83, which were exactly the same thing and slightly different plastic packaging was $100 in 1991, was $100 in 2010, and was mandated by all the textbooks. There was some weird going on there.

David: Yeah, definitely something weird. I don't know about you guys, but my main use case for that was like playing Mafia Wars and Drugwars.

Mario: I don’t think I had Mafia Wars, but I remember some more innocent games on there.

Ben: This is a foreshadowing of David would be playing very different Roblox games than you, Mario.

Mario: Yeah, exactly. I guess if you go to the real long tail there's probably some dark stuff. But the top end of the distribution is pretty light. Anyway, at Interactive Physics, the brothers Baszucki hired this extremely talented engineer called Erik Cassel. You guys can correct me if I'm pronouncing it incorrectly. 

He was working at Cornell's physics department doing some really interesting work and read about Interactive Physics in the MacUser magazine. He flew out to take an interview and the three of them ended up building Interactive Physics and the parent company Knowledge Revolution into a significant business that eventually sold for $20 million to MSC Software. That's the end of that chapter one of the Baszucki-Cassel collaboration, but the two of them kept chatting.

David: This all happened in Silicon Valley. Baszucki had gone to Stanford—I don't know where he's from originally—but kept working on the company there. So this was post-Silicon but the software era, the PC era of Silicon Valley.

Mario: Exactly, and as is the case with so many of these stories, they built Roblox out of a small office together in 2004 with pretty vague [....] of what they want to do. But knowing just that they wanted to take that kernel that was the unexpected outcome of Interactive Physics—this desire to play and build something around that.

They built this first iteration called DynaBlocks, which early on they have to explicitly on the website call out that there is no affiliation with Lego because the worry is that DynaBlocks later Roblox sounds too similar. The iconography, the avatars, the imagery they use is too similar. They have to be explicit that yes, these are our digital Legos but by a different company altogether.

Ben: Lego is famously very protective of their intellectual property despite the fact that you could go get some plastic and injection mold it into something in the shape of a Lego. You'll notice that there are really not any of those on the market and that is because they are hardcore about making sure that you don't build a system that rips them off or integrates with them in any way.

Mario: That's super interesting. That makes sense. They certainly did not want to fall foul of them early on.

David: But definitely, at the same time, the fact that these were sort of black, cartoony characters in a digital world, they knew that this is going to hold some appeal for kids.

Mario: Absolutely. There's definitely a familiarity there.

David: In 2004, they started DynaBlocks that becomes Roblox, and then I think it's at the end of 2006 that they launched into the world. They worked on it for two years and then launched this product. It's actually a suite of products. It’s crazy to think this happened in 2006. People think of Roblox think Lil Nas X all this, but it's not just that. What is this system?

Mario: I mean it started out with a more simplified version of what today as the Roblox client. You basically had a hangout spot, your friends had hangout spots, and you could jump over to each other's little virtual homes and mess around, have fun, and socialize. Over time, that evolved into the more complex gameplay you have now that compromises the Roblox client.

When you look at the product suite today, it's really these three different lines. You have the Roblox client, you have the Roblox studio, and then you have the Roblox cloud. Each of those fit into this ecosystem.

David: I don't know the full history of Second Life. We're going to have to pull all this together in a future episode. Basically, it's like Second Life for kids is what this quickly evolves into, right?

Mario: Yeah or just totally open world, infinite iterations Sims. The client is the place that people are probably most familiar with. It's where you log in as a player of games and navigate to one of the 18 million titles they have on there.

Ben: That's so insane.

Mario: It's nuts. The funny part is that six million have never been played.

Ben: It's like podcasts.

Mario: Exactly. You guys are like the Shinobi Life of tech podcast, right up there.

Ben: There are 18 million games—we’re cheating hardcore and flashing all the way forward to today. Mario, do people go into Roblox and do stuff just in Roblox, or is it the Epic Games Store where there's really nothing to do in the outer layer. You have to go into a game to do something.

Mario: I'm very likely to utter a shibboleth in saying this because I have not played as much Roblox as I'm sure many, but I believe you really don't get into the meat of it beyond decorating your avatar. You can make some cosmetic changes to your very local little house space before you jump into a game, but again, this is based on my minimal experiences playing Adopt Me! and preparing for this.

Ben: I guess where I'm going with that is Roblox, despite being an open world because there are so many games to go play—there are 18 million of them—most of the activity actually happens inside those games. When you say the open world, it's because there's a ton of games you could play, and it's not terribly hard to build a game yourself. When you're comparing that with Second Life or The Sims where it truly is an open world. There's no guidance, no game mechanics. It’s just like, okay, now you get to walk around digitally.

Mario: Yeah. I think each of the games themselves, in some respect, take on that quality. If you go into Welcome to Bloxburg, a lot of the activities are you working at a pizza place, hanging out in your house and going to high school, or whatever those things you want to do are. That's a big part of the power is that people now use it just as the watering hole for 9-13 year-olds. It’s also where a lot of the challenges come in because since socialization is at the heart of it, moderation is not far behind.

David: There's an important point up there that is very different than Second Life and the like, which is that Roblox isn't making the content or the games. It's people on the platform who are making games, and we're going to dive a lot more into this as we go. But there is the separate Roblox studio product, which really these days is basically pretty close to Unity in terms of what you're both able to do with it and the complexity and knowledge required to build high-quality games and experiences. It's a full-on programming environment, game development, editing platform, and it's separate from the client.

Ben: It's like Unity but it only works on Roblox. You can't release your game anywhere. You can release it everywhere where Roblox has a client app.

Mario: Exactly. The innovation (I suppose) there is the use of this programming language—Lua, but that simplifies the process of creating a game greatly. That has allowed younger folks, children to create games of their own. Over time, as they grow up, the games mature as well. It's this on-ramp to game development. 

It's pretty clever from a business strategy perspective in the sense that if you can give someone the tools to earn money early on as a young game developer and attach them to that asset, then there's obviously an ongoing incentive for them to keep building it even if they themselves might age out of using the platform.

David: It's just like Interactive Physics and the TI calculator. You can consume—most kids, mostly myself included—a little bit of tinkering, but mostly I was just playing games that I downloaded from my friends that they had downloaded from other people on my calculator. But if you want, you can open up basic right there in the calculator and make your own games.

Mario: Totally.

Ben: Mario, take us forward from 2006. How did it go when they launched it? Did they launch with all three of these components—the cloud, the studio, and the client—all at once?

Mario: I actually don't know if all three of them launched at once. I’m pretty confident in saying that both studio and client launched at once, but I would imagine that cloud must've been at least greatly changed over time as they've gone so cross-platform. You can access Roblox on every mobile device, VR headsets, PC, Mac, et cetera. That's been pretty key. If you're a kid, you can just jump into it wherever you want.

To return to your question, it's a pretty slow go for Roblox after 2006. They didn't raise a ton of money. Altos Ventures was the first outside investor. Their Series C was $2.9 million.

Ben: That's the letter C, not seed.

Mario: Yes, C.

Ben: Third round of funding.

David: I assume it must have just been friends and family before that and they just called it A and B. That was the first institutional. That was in 2008 I believe.

Mario: That sounds right.

David: I think it was. I think they're operating for a couple of years. This is such a case study of exponential growth, economics, and then the economy itself. For years, there was nothing going on here by venture community standards. They were operating for two years with no institutional funding, then the first round came in, then the next round, and then several years after that. 

The company was making money operating profitably or breakeven-ish, but this didn't attract a lot of attention until people started to realize this is a whole economy going on here. Let's get into the business model of how this works.

Ben: Just to color what David is saying there with a stat that Mario pointed out in the S-1 Club, over the first 12 years of operation, they only raised $11 million. Saying it was not a darling of the venture community for its first dozen years is a wild understatement.

David: Compare that to companies growing today and raising venture dollars in today's environment.

Mario: It also speaks to how much venture is the view of gaming writ large has changed over that period with companies like Twitch, Epic, and everything else that have also risen in prominence similarly. But it's really only in 2017 when Baszucki mentions that Roblox is starting to get some network effects—and maybe we'll talk about this when we talk about the seven powers later—that the money starts really pouring.

Ben: David, back to your earlier question, let's talk about the economy. How does the business model work, Mario?

Mario: This is one of the parts of Roblox that I think is most interesting is that there is this thriving economic system, and it's quite an idiosyncratic one in many respects. Basically, Roblox makes money by selling users Robux. That's their in-game currency. You as a user are buying Robux to buy in-game experiences, to buy cosmetic items for your avatar, and all of those items or the vast majority are created by developers on the platform. 

Essentially, Roblox is engaging in a revenue split with the creator of those objects and experiences and a pretty steep one in fact.

David: There's an exchange rate that Roblox sets. They're like the central planners of the economy between USD—or I assume lots of other currencies too—and Robux. Users buy-in, they buy Robux at a certain rate, but then developers or anybody can cash out back to USD, I assume at a much lower rate.

Mario: Exactly. It's a clever default, but if you're a developer and you earn money from one of your games, for example, you are first paid in Robux. Not only is the rate of exchange lower, but also, there's the sort of implicit temptation that you might just use your Robux back on the platform, spend it on someone else's game, and so forth.

Ben: If I'm a 15-year-old game developer who's like, look, I made my first game. This is super fun. I get some Robux. I'm not going to cash in with that.

Mario: You're going to binge that on garlands for your avatar's hair.

Ben: What is that exchange rate? There's some complexity and they are clever with subscription models, but if I just put $1 in or if I just put whatever it is in, how many Robux do I get just for a one-time purchase?

Mario: On the developer side, the exchange rate is set by Roblox at $0.0035 per Robux. On the buyer side, they're really trying to push you towards packages directly from the site, obviously because of app store cuts and we can talk more about that.

Ben: It's cheaper for me to get Robux directly from the Roblox website than it would be to get them on the iPhone.

Mario: Exactly. You can buy the packages that we highlight here 80 Robux for $0.99 or 450 Robux for $4.99 a month. That's another way they play with the value is pushing toward subscription rather than one-off bulk purchase.

Ben: It sounds like on that one-off purchase, they're taking effectively a 20% cut just on the purchase because you're getting 80 Robux. I'm trying to think through an entire transaction where someone puts money into the system, then it flows through, and then the developer gets money out of the system. How much does Roblox get to keep along the way from the time the consumer puts it in until the developer can pull it out?

David: I think it very much depends on how the Robux was purchased and time in the system. There's a time element to this too, and how many times it re-circulates in the system. There are two super brilliant, really novel things about this that exist in other virtual economies too. 

One is that there's no transaction cost that Roblox has to bear. If they were operating in a marketplace for games, like Steam or whatnot, users would buy stuff in USD. They would have to process all that payment. Those dollars would have to get shipped off to the developers. Stripe somebody that’s making money here, and it's hard to do all that. With Roblox, once the dollars come into the system it's all native.

Ben: There's a payment process required upon buying Robux, but after that, you can have tons of transactions at effectively zero cost instead of incurring credit card fees every time.

David: Which then they do. It re-circulates in the system. If you're making Robux as a developer, yeah, you could binge on it. Well, I don't know, Mario might know, but I'm assuming you can build into your gaming experience distributing Robux back out to the users of your game as part of the economy of the game, and then they can just recycle on and on and on and on and on within the system. However long they remain in the system, it's like an interest-free loan to Roblox.

Mario: I actually don't know about that specific case, but I think the principal regardless is definitely true.

Ben: That working capital thing is so interesting. Part of the reason that they're not public yet is because of this question around how they're doing accounting for Robux in the system. But if you just pop up a level from caring about the minutiae of how exactly they're doing the gap there, they are totally getting an interest free loan, and they totally have a negative cash conversion cycle where they get all this money before anybody has to do anything with it. It could be a lot of time before they need to pay any of that out to developers.

David: If ever.

Ben: Wild.

Mario: Definitely. It's crazy stuff. The tough part of the problem is really that the developers often end up with a very small proportion of the total spent on their experience or in their game. Roblox themselves estimates that it's about 24.5% of the transaction that goes to the developer. A few deeper Reddit community forums in the Roblox world suggest it's below 20%, like 19.7%. 

That is not just one of those figures that makes you raise an eyebrow, but it also gestures towards one of the fundamental worries that people might have about Roblox, which is whether they'll be able to keep developer talent happy for the long run, especially as this war for good developers keeps up.

Ben: Did I just hear you right? That you only get to keep 20%-25% as the developer. Whereas if I'm releasing an app on the app store I keep 70%?

Mario: Exactly.

Ben: Why am I developing for Roblox?

David: I think I tweeted about this yesterday. A take away that, Ben, you pointed out was actually really insightful from our episode with Alfred Lin on Sequoia's Playbook. There are two ways you could have high gross margins. 

You could be a software business that sells software and have high gross margins for that, but that's deceptive because there are transactions that people are accomplishing with your software. You're not taking a percentage of that. You're low gross margin on the net end transactions. Or you can be a full-stack company or marketplace where you're running the whole system, and those tend to be lower-margin businesses. Some of those businesses are high margin like Facebook or Google. They run the whole system, and they are a high gross margin business.

Ben: Facebook doesn’t pay a dime out to creators.

David: Exactly, or Instagram. In cases where it's possible to sustain that, the only way that's possible is if you have an immense amount of power in the Hamilton Helmer sense because otherwise, all your users would just leave you. We’ll get to this as we go along, but clearly, Roblox has an immense amount of power. It’s all within the ecosystem. The users that are using the client are playing the games, the games are made for the client. You could not like that as a developer, but if you want to access these 150 million users, well, you got to do Roblox.

Ben: You're saying that it's not actually apples to apples. There are very few developers who are out there saying, should I make an app store game or should I make a Roblox game? They're going to Roblox users and saying, make Roblox better. We’ll even pay you for doing some of it. In the same way that Snap has experimented with you Snap more and we'll pay you for being on the social network and enriching it with great content.

Mario: Exactly. Also, just the sheer fact that it is a less mature audience in many respects that come in probably first as a player of this game, the attention is captured, and then there's this opportunity to build a world on top of that. It's maybe less of a financial calculation at that stage of the user's life.

David: Right. It's like you could build stuff in Minecraft. Great, lots of people and kids do. You could build stuff in Roblox and then make some money from it. It's like, great. I learned how to develop and I learned how to run a business.

Mario: Exactly.

David: Then all that IP gets locked into the Roblox universe.

Ben: Mario, did you do any research on who these top game developers are? Are some of these stories of 15-year-old who got crazy rich who has a hit game on Roblox?

Mario: They're definitely on the young side. I can't remember which game it is at this point. I want to say it's MeepCity, but it could be a different one. I think it was essentially modified from a different game, sort of heavily inspired by previous games. It started without a huge amount of fanfare and then metamorphosized into a hit game. 

More broadly, looking at the S-1, Roblox highlights a few of their developers, and they're probably slightly trying to spin the narrative in a certain direction. But they're all on the very, very young side. There's no photos of a 48-year-old dude running a Roblox game that's crushing it. It’s like a 21-year-old who was playing it for 10 years.

David: I'm sure they exist out there. I know there are definitely older folks and studios that are making Roblox games. But Ben, to your point, that's less attractive. You have to really believe you're going to get a lot of users in Roblox to do that versus I’ll build in Unity or I’ll build in Unreal and I’ll ship it to every platform and keep 70% of what I make. 

Let's talk about the market because I think you did a really good job, Mario, of breaking down—all this sounds great, lots of power, but we’re talking about kids’ games here. How big is this? What's the TAM? What's the opportunity?

Mario: One thing to say about the S-1 Club that I think we haven't necessarily hit on yet is that it is a very collaborative effort. I want to say that Alice Lloyd George, who's a great VC, did a lot of amazing research on the market section in particular. That's true throughout, it really is a collaborative effort. 

Digging into the market, we've seen—over the past few years and we touched on this—just how much ventures’ interest in gaming has grown. That's really because it's become a massive market. I mean, it's a $159.3 billion market as of 2020, growing almost 10% a year, expected to reach $200 billion by 2023, and there are some specific characteristics of gaming that make it particularly attractive.

Ben: It's so funny you say growing at 10% year over year because I do remember a few years ago when we interviewed—I can't remember if it was Nolan Bushnell or Trip Hawkins, but whichever one it was commented on, it's crazy. I invented video games for the consumer market, and now it's over a $100 billion market. It's literally compounded 2-3 years since then at 10% per year.

Mario: Yeah, and it's exceeded most expectations. I remember looking back at some of the previous estimations. I think it's outperforming certainly. There are statistics that we actually didn't use in here, but I think it's something like 3X the growth rate of gaming versus other forms of media. Whether that's film, TV, et cetera. It's capturing more and more of that attention, particularly amongst younger audiences.

David: The whole Fortnite phenomenon over the last few years, which is probably a slightly older age demographic than Roblox I think just shows for younger generations, these aren’t just games. These are what they're doing instead of watching Netflix.

Ben: Or instead of social networks.

David: Yeah, or instead of Instagram.

Ben: It's entertainment and social networks. Both of those markets were growing on their own anyway, but gaming is just for a certain audience combining them together.

David: If you think about it, Roblox has the youngest users if—there's a big if that we will get into as we get farther along in the episode—they can keep these young users. This is like Netflix plus Instagram for them or Netflix plus WhatsApp. Think about the value that’s going to generate a —50% rate, is that what you said, Ben, in the intro? 50% of…

Ben: Of sub-15 year-olds I think in the US used this product in the last 30 days.

David: Unreal. There's also a really interesting piece of the market here in China, right Mario?

Mario: Exactly. Roblox has done an amazingly good job capturing the US. The Asia Pacific is a colossal market. I think the number on that is 72.2 billion market size as of 2019. You can really see that it's essentially 50% of the market. Roblox has been I think very pragmatic about entering China. They have entered into an agreement with Tencent through sort of a JV where they own 51% of this China entity and Tencent owns 49%. 

Tencent's role is really to walk Roblox past the governmental licensing roadblocks and allow them to operate there. They've successfully done that it looks like from the two licenses that they got in December.

Ben: Tencent is such a freaking juggernaut. I feel like they show up in every episode somewhere, David.

David: They're amazing. But yes, in China, you need a license to operate a game. This has been true for a long time.

Ben: Didn’t they pause licenses for games a couple of years ago?

David: They did. Tencent, as we covered way back on our episode with them. Tencent really originated—well not in the very, very beginning, but it was scalping messages in the beginning. It was QQ—an ICQ knocked off, but then got into gaming. Tencent was an operator of many games in China and got the licenses, built those relationships with the government there, and has an established track record of bringing western games into China—League of Legends, etcetera. 

What you're referring to, Ben, is Honor of Kings, which is the mobile version of League of Legends that Tencent then developed in-house after they had acquired League of Legends. People started playing it so much in China that the CCP revoked their license temporarily. There's no way to operate in China without the strength of somebody like Tencent that has those relationships that can get Honor of Kings back reinstated. No western companies is going to be able to do that.

Ben: Right. Looking at power dynamics, if what Tencent is doing is saying, yeah, we'll make sure that you can get to market here. We will take half of the company for that, and people are willing to say, yeah, we'll totally do that. That is like an unbelievably strong power dynamic to be able to provide that value.

Mario: Yeah. It's an interesting defensive maneuver also. We'll get to competition I'm sure a little bit later, but Reworld is sort of the Chinese version of Roblox—much less mature, much earlier in his life. If Roblox paid homage to Lego in its initial incarnation, Reworld is very much worshipping at the shrine of Roblox. It's a good move for plenty of reasons, but you can't discount the fact that it's certainly going to make the path significantly more difficult for Reworld.

David: All right, let's talk about management and investors in Roblox. This is a fascinating one. We talked about Dave and Dave's been there. He was the founder and CEO. Basically, the whole rest of the management team is new.

Mario: It's an interesting one. Looking into it, you don't see a lot of gaming experience as it happens. You see a lot more retail experience and even a little bit of social network experience, but even that is not particularly pronounced. It's only really when you get to Roblox’s China czar who had some TV experience at Sega that you see that expressed in any meaningful way. But yeah, the rest of the team is fairly new and from some spaces that you might not necessarily associate with Roblox as a core business. 

You have the CTO, Daniel Sturman, coming from Cloudera. That makes total sense in the sense that this is a huge infrastructural project to keep up and running, and has been historically an area of weakness, but not necessarily gaming related.

Chief Business Officer, Craig Donato, coming from Nextdoor. Before that he was at QVC. That’s an interesting twist because Roblox is notoriously quite aggressive in up-selling in-game, and so perhaps that interactive sales experience is valuable. 

CMO experience from Walmart. You have a true car experience, more ecommerce experience. It really comes from the gamut.

Ben: They're not trying to be EA. They're trying to be something that covers a variety of sectors. They don't think of themselves necessarily as a game company.

Mario: I think that's right. They describe it in the filing as trying to win over human co-experience. That is fittingly vague, but also capacious enough to account for social networking, purchasing, ecommerce, et cetera.

Ben: That’s so funny. Just like a meta point on this, these incredibly vague market definitions are so—on the one hand, highfalutin ludicrous. On the other hand, they're trying to paint that literally anytime where two people are connecting over something, we could foreseeably address that in the future. When you're trying to estimate our total addressable market, you should be in the high trillions. It's sort of BS, but on the other hand, you’re like, they're definitely not a gaming company. How else do you want to describe it?

Mario: For almost anything else, I think you sort of are tempted to laugh it out of the room. But with something like Roblox you're like, I mean human co-experience sounds a little bombastic, but it doesn't feel far off. It's like a matter of degree rather than a matter of order.

David: You did this cool thing in the S-1 Club where you have word counts of things that appear in the S-1 how many times different words appear. In Metaverse, it’s one of the ones that appear on there. It is a very easy Tim Sweeney Epic-like the ambitions of this company. To me it makes a lot of sense, regardless of whatever marketing jargon they want to use to describe it. It makes a lot of sense that executives from EA, from Activision Blizzard—maybe Activision Blizzard and Blizzard particularly with World of Warcraft are pretty similar here.

But if you're used to making package software, that's not what's happening here. They're doing a couple of things. They're operating cloud technology on the scale of AWS, Google, the same scale as Epic with Unreal, and then they're operating an economy just like Google, Facebook, or the like. They're not making content, they're operating an economy.

Mario: 100%.

Ben: Before we move on from this point of talking about other Metaverse-type companies. If you were to ask me five, six, seven years ago who's going to own this space, and obviously that was well before Fortnite came out or even really was it in development. I wouldn't have told you it's going to be Epic. 100% chance I would've told you it was Minecraft. Is it just getting bought by Microsoft, or how is Minecraft not the company that we're talking about right now?

David: That’s a great question.

Ben: Did they have a strategic misstep? Because I assumed their user base is also extremely large and has continued to grow since being acquired by Microsoft.

Mario: I think that's right. There's a statistic here that 50% of 9-12 year-olds in the US, Canada, New Zealand, and Australia play either Roblox or Minecraft. If Roblox has overtaken them a little bit, it's no mean feat. Minecraft is still incredibly popular.

Interestingly, Roblox purchased this company called Code Kingdoms—it used to be called Ceebr—that basically teaches young game developers how to build games, and it offers classes both for Roblox and Minecraft.

I still suspect there's a huge amount of overlap there, but to your point, it does feel like selling to Microsoft fundamentally changed the trajectory and ambitions of that business.

Ben: I'm on businessofapps.com. We're doing this live, so caveat's about how reputable the source is. It's from Sensor Tower, PCMag. In 2020, there were 131 million Minecraft users. Mario, how many users were there on Roblox?

Mario: I think we have 150 or something like that.

David: 150 MAU?

Ben: Right in the same league. On revenue, Minecraft did in 2020 about $415 million. What was topline revenue for Roblox?

David: Bookings was about 1.3.

Mario: There we go.

David: We'll get into this in a minute, but bookings is the number to look at.

Mario: I wonder if it's the same for Minecraft. I wonder if we should be looking at bookings there too.

David: I'm going way out on a limb here, but there's the Microsoft acquisition that may have derailed Minecraft's economy building a little bit. My understanding—this is where I'm way out on a limb—is that experiences in Minecraft that are not centralized operate on servers. You have to run your own server if you're going to host a Minecraft experience. This is where we were talking about Roblox being a cloud company. That's not the case in Roblox. It's all centralized in the Roblox cloud. 

I would imagine that significantly hampers their developer/creator TAM in Minecraft. In Roblox, it's like I want to start building a game and creating. Great, Roblox takes care of everything for me there. Whereas in Minecraft, I got to set up a server, I got to run a server.

Mario: Presumably, it's also considerably less flexible. Minecraft games all share more aesthetic DNA. It also feels like Minecraft themselves push their own games or game themes (so to speak) more expressly than Roblox does. In that respect, Roblox probably benefits from the UGC element a bit more directly.

David: Roblox studio is a very powerful application. It really is basically Unity. You can build anything. You're not just using templates and building blocks in there.

Ben: At risk of over summarizing, very similar user counts between Minecraft and Roblox, but Roblox has managed to monetize it probably over 3X, at least in the last year where Minecraft was.

David: Before we get into talking about accounting, which is going to be the most dramatic part of this episode for real.

Mario: Some accountants out there are like, finally, our moment.

David: We pulled the E brake, heading off the cliff, and then change in revenue to bookings.

Mario: Investors, got to give shoutouts to huge friends of all of us in Altos who we've mentioned as the first institutional investors in Roblox. They're wonderful investors, not well-known and they like it that way, but very, very different DNA and mentality than the rest of Silicon Valley.

My former firm where I worked as an intern when I was at GSP, Meritech, came in as the first growth investor in the company. I can't even remember which round Meritech came in, but I think in the same round that Meritech and Index co-led once growth really started to pick up. I think that was where the company had only raised $11 million before then. Then they've invested in every round since. Now, there’s Index, Andreessen Horowitz, and of course Altimeter and Dragoneer.

Ben: It's quite the list.

Mario: It is quite the list, but that's all in the last few years. The mindset of this company is not the traditional Silicon Valley for a long time.

Ben: This is the playbook you should run. You want to raise as little capital as possible and be profitable until you know for damn sure you have a runaway success on your hands, and then go raise a ton of capital all at once on your terms.

This [...] way that they ran the company is very similar to how Epic Games was run. We talked with Lad about this. He couldn't get funding for Webflow for all those years, and then boom it sells $80-something million round. I'm sure they had plenty of dilution happen in those early years. In fact, by looking at the percentages that these early investors still own, we know they took a lot of dilution in their early years with Roblox. But if you can raise little capital until you're a runaway success, and then raise a lot—that just seems like the way to do it.

Mario: 100%.

David: All right. Let's talk about accounting.

Mario: Here we go.

David: We were talking about the business model and this recycling of capital in the economy. As you can imagine, this makes for a nightmare for GAAP accounting. The company reports bookings, which is dollars flowing into the system in any given period. But then some portion of those dollars flowing into the system will ultimately get paid back out to developers. There's a very complicated revenue recognition formula here that essentially makes no sense. But you got to have something, I guess.

Ben: What did they try to do, and then what did they have to change to?

Mario: I actually don't know what the change is going to be, and that will be very interesting. The trickiness here is that there is this accounting standard called the Accounting Standards Codification 606, ASC 606.

David: It should be 666.

Mario: Yeah, no kidding. That was imposed upon public companies starting in 2019. For the most part, that's fine. But gaming companies, in particular, are not very well-served by it because of the strange economies and strange purchases that you might make on a gaming platform.

In Roblox's case, they have to recognize their revenue not when someone purchases the Roblox on the platform or when that person spends that Roblox, but over this indeterminate or amorphously defined LTV period, this lifetime.

Ben: That was the craziest thing reading the S-1Cclub, Mario. When the second shoe dropped, when you're like it's not when people buy the Roblox, I'm like, that's tricky because they're going to have to figure out how to account for the amount of time between when someone buys the Roblox and when they spend the Roblox. But no, it's what is your LTV or what is your customer lifetime? Is it amortized across the customer lifetime?

Mario: Yes, exactly. Growth—who is an anonymous Twitter account—is very brilliant about this space, really spearheaded this, and also did some hilarious writing here. I think he managed to make accounting seem very funny. But yes, exactly.

Roblox's obligation to a USC user doesn't cease once you spend your Roblox because presumably, in some instances at least, you spent it on a digital item—whether that's a crown, a piece of armor, or whatever that is. That in and of itself becomes a manifestation of that value that Roblox needs to—

David: Keep servicing, essentially.

Mario: Exactly. It's a bit convoluted and makes it not the most legible.

David: We'll talk about narratives in a second around this company. Mostly, the narrative is to the moon in this company. To an extent, there is a bear narrative. One of them is, this company is losing massive amounts of money and is unprofitable. That's just wholesale false and is a consequence of this accounting.

If you look at the revenue and their net income, yes, they are net income negative. But what you should really do is essentially take their topline bookings number, which is dollars flowing into the system which for the first three quarters of 2020—we don't know yet the fourth quarter of 2020—was $1.25 billion.

If you really want to be conservative, you could haircut 20% out of that. That's eventually going to go back to developers. Fine, do that. You're still sitting on a billion in the topline. The company is bringing in lots of money. It's very free cash flow positive.

Ben: Their EBITDA margins are 25%-28%. That's not quite Facebook territory, but that's on its way.

David: And for a high-growth company too. Being hyper-conservative would be stripping out the 20%. Being less conservative would say this Robux is going to circulate in the system a few times before they pop out. Yeah, 20% will get popped out eventually over three or four years, or maybe longer. You could time discount that back to maybe you haircut this by 5%.

Ben: This is the Warren Buffet playbook. This is Float. This is like the brilliance of owning an insurance company. Roblox doesn't seem to be doing this, but they have so effectively a giant crowd-sourced interest-free loan that's just sitting there that they could do other things with. Anybody who's describing this company as an unprofitable company right now doesn't realize that in some ways, that huge liability is actually an underutilized asset.

Mario: That's a good framing. I totally agree.

David: On the growth side, the bookings for the first nine months of 2020 tripled versus bookings for the first nine months of 2019.

Ben: Which is pandemic-related.

David: Of course, it's driven by the pandemic.

Mario: It's still nuts.

David: It's still a little nuts. From ‘18 to ‘19, they grew about 50% topline bookings.

Mario: That sounds right.

David: A little over 50%, maybe 60%. They went from 500 million to just under 700 million. No, that would be more like 40%. Anyway, lots of high-growth tail ends at their back here too.

Ben: What we're painting here in both that stat that I threw out on the EBITDA margin napkin math and on what you're talking about and how they're able to scale—this is just a tech company. This is a pure play tech company. I should give credit to Scott Galloway on Pivot. He pointed out 79% of their full-time staff are engineers.

If what you want to do is invest in technology businesses on the internet because they have zero marginal cost and zero distribution cost, not only is that true, they also don't have sales people like you would have at a Snowflake or something. They have a bunch of engineers who are building a system that makes games that have digital goods that uses a digital currency. This is as technology internet-native a business as you could possibly have.

Mario: 100% agreed. That and moderators. Those are the key functions.

David: Let's talk about that because that is a piece of this. But it’s a piece of so many—Airbnb, Rover, Amazon, customer support moderation, Facebook, Google. This is an expense that scales.

Mario: Yes, although compared to some of those other companies you mentioned, this is obviously more acute in some respects because of the demographics of the user base. Over 50% are under 13 years old. It's a place where a lot of socialization happens. There are pretty robust moderation systems in place.

I think Roblox serviced something like 9 million tickets or issues that propped up last year. Their average response time to it was 10 minutes. They've clearly invested a lot in it, and we touch on this in the piece, but they've managed to improve their efficiency there. That's one of the places where it seems like they've gained some margin. But it still is a massive vulnerability.

David: Both the vulnerability and risk, a cost that scales with usage of the system. Also, it's something that we got to commend them for too. Of course, there have been issues. You guys talked about it in the piece. Roblox dating is like this shadow thing that happens that they're constantly trying to crack down on. Operating essentially a social environment for underaged children around the world and keeping it safe is an immense task.

Mario: It is a gargantuan effort, and I think they've done some really interesting stuff around tech screening to make it more difficult to say anything explicit. They seem to have built a really good moderation system that is human-powered behind the scenes as well.

The tricky thing is, I wonder how much more difficult that may or may not get as the user base changes over time. Maybe a larger portion of users are older, they can take a more hands-off approach there. The corollary would be that as more older users are on the platform, the younger users are potentially more vulnerable to more mature commentary.

David: This is good. I want to get deep into this competition in a second.

Ben: David, I feel like you're going to ask the question, are all these people going to stick around after they turn 16 and after they start looking at other platforms? Just because you're capturing the next generation, it's only a good business in the long-term if you can keep them. Are they keeping them?

David: I was going to say billion-dollar question, this company is already a $29 billion-dollar company. That's the couple-hundred billion-dollar question, right? This deserves a pretty deep double-click here. Honestly, for me, the biggest reason to give me pause about the stock and the DPO.

Structurally, everything that we just said makes Roblox—beating Minecraft—a fantastic economy, a fantastic place for kids 13 and under. It was a long time ago when I turned 14. But I remember when I turned 13 or 14, I didn't want to have anything to do with the media I was consuming, things I was doing, memes that I was then part of when I was 10, 11, or 12 years old.

Ben: Right, it's kids’ stuff.

David: That's a huge question for Roblox, I think.

Mario: 100% agree. Once you turn 13 or 14, you start to become a little edgelord and want to take in the most mature media that you're allowed to.

David: Oh, yeah. I remember I got my hands on our family VHS copy of Pulp Fiction when I was 13 or 14. I probably burned the tape out of that thing.

Mario: You're exactly right in identifying that as a key question as to whether a retail investor could expect to make a 3+ X on something like this.

Ben: Let's talk about what a 3+ X on something like this would be. What we’re talking about right now—even at this $30 billion valuation, if we're being generous and talking about bookings at one point too, we're in the category of 25X bookings on the price. It will likely be closer to 40X, 50X by the time the public actually has access to this thing.

You're buying a stock here that's trading at call it 40X bookings and probably 100X-120X EBITDA. Already, tech margins right now are in an insane place. What do you have to believe about the underlying business’ growth in the near future in order to get a 3X on those already very, very, very high multiples?

Mario: You're right to point out that the retail investors are probably not going to get anywhere near $30 billion anyway. A 3X from wherever it sits at the end of IPO day suddenly looks quite tough. I wouldn't be surprised if we get reasonably close to a 3X from the last round by the end of the first day of trading.

Ben: Can I just make a meta point here? If these prices just feel so disconnected from the underlying businesses—it feels like what we've decided for a certain crop of public companies, they're more like currencies. We just did the Bitcoin episode. They aren't coupled to an intrinsic value. They're much more based on the greater fool theory because of a collective buy-in. They feel a lot more like currencies than they do something that's strictly tied to the utility value of the underlying asset, which is I am entitled to future cash flows from this thing.

Every number on every dimension has to be at the very maximum of what you could possibly believe—growth rate, revenue, number of years you want to factor into your DCF to make any of these things fit. Let's just throw it out there right now that if you're going to get the 3X on this thing, it's probably more likely because of hype than it is because the business kicks ass.

Mario: On any sort of short-term timescale.

David: Yeah, it very much depends on the time horizon. We style ourselves, investors. All of us—Mario included, although feel free to speak for yourself—are fans of the compounding mindset on this show.

Mario: Please be careful with that word. I'm immediately triggered. For those that haven't heard that tale, check out The Six-Billion Dollar Stare piece for more context. But yes, that sends a chill down my spine.

Ben: We're going to go down this rabbit hole. What are you talking about?

Mario: You don't know this?

Ben: Nuh-uh.

Mario: Oh man. Last (I want to say) November, I, with a collaborator, wrote a piece on public compounders—companies that regularly grow 20% or so a year. It caused a massive kerfuffle because a hedge fund called Durable Capital read the piece and decided that phrases like good to great and compounder were their IP.

David: Oh my gosh, I'm sorry I forgot about this. What a kerfuffle. I didn't mean to trigger you.

Mario: You must have been like, what is he talking about? Now the word compounder brings sirens to mind and all sorts of other terrible things. They threatened to sue me for writing this piece if I didn't take it down. When I asked them why, what words, or what claims they were making, they shuffled their feet and threatened to call GCN.

I asked again and still and still, and it basically just became a question of whether I was going to take this piece down or not and whether people are allowed to use the word compounders or good to great. Thankfully, I had some pro bono representation from a media law firm that was able to back me on this.

I shared this story on Twitter and it went a bit nuts because FinTwit was like this is absurd. People have been talking about these concepts for decades, what nonsense?

David: Can Warren Buffet not use these words?

Mario: Yes, it was a very strange one. Anyway, that's the rabbit hole.

David: Wow.

Ben: Thank you for sharing that with us. David, let's talk about the time horizon a little bit because I think I made a snotty little joke there about near-term pop being because of hype.

David: I don't know what near-term and long-term is these days anymore. If you're thinking long-term and you believe in the markets, the power of this business, the attractiveness, the growth, and everything we've just laid out here, the biggest question to think about is how defensible it's going to be in the long run and competition, and is there anybody else credibly that could come and take this momentum from Roblox?

A friend of the show, Nick Fajt, co-founder and CEO of RecRoom in Seattle, who we had on way back in the early days to talk about raising his seed round from Sequoia. Rec Room—we'll talk about this, I talked to Nick last night—is in many ways Roblox built on modern technology, and Roblox for 14-18+ year-olds. Designed for older kids, you're able to do much more, much less restrictions on your communication.

They have also seen incredibly explosive growth through the pandemic and over the past couple of years. They're at about 5 million MAU—much less than the 150 that Roblox is at—but Rec Room is much younger and growing much faster. They grew over 500% in the past year. They're monetizing as well or better than Roblox rate.

Ben: They were just VR when we had Nick on. They're on every platform now, right?

David: Yeah, so there are a couple of really important points here. One is this user base difference between Roblox, which is really 13 and under, and Rec Room which picks people up as they age out of Roblox.

That's an issue for Roblox because of course, they could compete, but they can't structurally change a lot of Roblox because it's designed to be a safe environment for younger kids. They would probably have to do—and they've talked about this—maybe having a separate instantiation like a Roblox 13+ or something.

Mario: Mature.

David: Yeah, Roblox Mature. The other aspect is that RecRoom started in VR and is now on mobile, console, and every platform out there—now on screens. Nick makes a really, really great point that if you believe that all of this Metaverse-type stuff and economies in gaming experiences are going to be going into virtual and augmented reality in the future, it's really hard if not impossible to take a platform designed for flat screen experiences and turn it into VR.

You can go down. You can go the other way. You can go from VR to flat screens, but the way he explains it is so simple. He uses the analogy of Quidditch and basketball. If you have a Quidditch team, they could probably play basketball because Quidditch is sort of like basketball but you're flying around in the air. Basketball teams would be really hard to go play Quidditch because they're just used to two dimensions.

Ben: I love that.

David: What that means is when you build a platform for VR, you can do anything with your hands and your body. You can move and have any experience. When playing a game on a screen, it's usually hit this button on this controller to jump or hit this button on this controller to use the tool in your hand, and your avatar performs that action. In VR, you actually do that with your hands.

Ben: Has anyone ever successfully moved from 2D to 3D? I'm trying to think of any of the successful VR things and figure out if they come from a 2D predecessor.

David: There are. Roblox, you can play in VR, but it's the same thing. You can look around in VR and you hit a button on the controller to use the item in your hand. Whereas in Rec Room, you hold the paintball gun, you wave it around, and you shoot it.

Mario: Interesting. If I can offer a brief defense of Roblox’s ability to capture that segment, 50% is under 13 but obviously 50% is over 13. It's something like 54:46, and 17-24 year-olds have been growing at a faster pace than those under 13. If you look over time at that percentage under 13, it has been declining albeit by not a huge amount.

We've seen at least some ability to capture an older market in the way that they're offering different aesthetics and the way that they're bringing in people like Lil Nas X. The fact that there is VR available at least suggests they're aware of that. My only final defense would be that if you gave Viktor Krum and Lebron James a chance to go head-to-head in each sport, I'm taking Lebron every time.

David: every day. I love it.

Ben: Should we just stop the episode there? It's not going to go anywhere but down.

David: The reality is this is a rising tide that all boats are going to the lift. I do think this is really interesting. The other difference between Rec Room and Roblox is in Rec Room, it's just one platform. There's no separate studio. All the creation happens in Rec Room. That means that they have a much higher percentage of users who become creators. You can do stuff like remix rooms and experiences. 

You're in a room? Great. You want to change something about it, you want to remix, you want to publish it yourself? Super easy to do. You can collaborate with other people on creating inside the app. You don't exit to a separate environment. It will be fascinating to watch over the coming years how this plays out.

Ben: David, you just gave two bear cases. One was if you believe VR is the future, then you have to believe that—at least in this category—these things should be native VR. You painted the other one, which is that they're not necessarily going to be successful in retaining all these users that they get in order to continue compounding them.

Mario, great counterpoint that so far it seems they actually are doing a pretty good job retaining them and even acquiring older teens. Before we move on to bull, are there any bear cases that either of you has?

Mario: I think the big one for me is really retaining developers, and that seems particularly worrisome just because of the distribution on Roblox already. It's an incredibly top-heavy distribution. Adopt Me! is the top game on the platform. That has 3X the number of concurrent players regularly than the #2 game, Shinobi Life.

It's really that the vast majority of the platform is concentrated in the top 3 or 4 games, and then even beyond that on the top 20 games. It makes it incredibly hard for any new developer to come in, make a living from it, and try to gain share. That ends up being a fairly unhealthy dynamic when you look towards the future of game development and game creation.

Ben: Do they need more developers? One argument is it's working pretty well. What if they, at any given time, do just have 10 games and figure out how to make sure that they're getting 10, and they're getting 10 quality ones?

Mario: I think if you want to extend the LTV and especially appeal to a broader audience, you're going to need a lot of different experiences. I would expect definitely some degree of concentration, but it feels particularly pronounced here. It feels like the current version of the platform is not necessarily where they want it to be in terms of demographic appeal.

David: It's like the lesson from our TikTok episode two years ago—that's crazy—about you wanting to have dynamism among the creator economy of people rising and falling and not get locked in. TikTok has done this amazingly well. Why are so many people flocking to TikTok other than it's cool—that's a big part of it—but why have they done so well where Snap has stagnated?

It's because you can break through. You can be anybody and post something, and if it's cool, you'll get distribution. If that's not the case, you're not going to keep attracting dynamism.

Ben: Not to mention it's always interesting as a consumer because the content is always completely different, always very timely, and a clever remix of everything else that's ever come before on the platform and what's going on in life right now.

David: Those are spot on, the bear cases.

Ben: We painted lots of bull cases. What's the biggest one? Did someone have to walk away and be like, this is why this company is insanely valuable? What's the this?

Mario: I feel like it comes back to what you said at the beginning, Ben, which establishes that entertainment is a massive industry. Establish that social networks are a massive industry. Right in between them and across both of them, is Roblox. It's this insane mix of the way you want to spend your time instead of watching television, and also the way you want to spend your time instead of going to socialize with someone somewhere. 

That feels like an incredibly powerful combination that they have effectively found a way to really monetize well. That's incredibly powerful, in my view.

David: I would add too the economy aspect of this, which really is a new thing that we haven't seen before. They operate their own economy where they are the FED. They set the exchange rates and get to not only pay no transaction fees from moving things around but recycle money in the system. Imagine if Facebook or Google-operated that way, there would be a whole other order of magnitude of value.

Ben: Well, Facebook tried with Libra.

David: Right.

Mario: It shows how hard it is to bolt on, right?

David: Yup. Should we discuss powers?

Ben: Let's do it. Mario, for folks from The Generalist listening, are they familiar with the power framework at all?

Mario: I would imagine so, but a brief synopsis is never amiss.

Ben: All right. We have this section because previously on the LP program, we had a friend of the show, Hamilton Helmer, who wrote this incredible book called 7 Powers. I decided here not to put in front of me how he actually describes what power is and just try to articulate it in as plain of English as I can, which is, here's why you deserve to have profits, and here's why you deserve to have lots of profits.

If you really want to dig in, it's differential profits above your nearest competitor. The 7 Powers are the different types of reasons why you are entitled not only today, but in the future to have enduring, sustainable differential profits over your competitors.

Those are counter-positioning, scale economies, switching costs, network economies, process power, branding, and cornered resources. Rather than explaining each one, let's just pick one, two, or three that we think Roblox has, and then we can define them as we go in describing how it applies to Roblox.

David: They have a bunch of them, this will be fun. Mario, do you want to go first? Pick your favorite poison here.

Mario: It's tempting to just run the gamut, but one of the low-key ones is counter-positioning. Games have traditionally been seen as a waste of time by parents. By coupling it with education and particularly development like coding-based education, they're a few more marketable skills that a parent would be happy for their kid to learn. By framing it in that sense, it has a little bit of counter-positioning power.

David: Let's just go down the list. I think it has scale economies on the Roblox cloud, part of the business. As we talked about it without knowing enough about Minecraft, it's a major advantage they have over Minecraft that they're able to centralize all the compute storage networking, et cetera resources needed. Switching costs, 100%. 

Ben: Switching costs to me feels like the biggest on the developer side. If you're a teenager who wrote something in Lua and deployed it on Roblox, not only is that user base not easy to acquire on other platforms, but also the dev tools system is completely different. You’d need to pick up a whole new set of skills and hire a whole bunch of new people in order to move to somewhere that's not Roblox. It feels like there's a tremendous amount of lock-in on the developer side.

David: I think on the user side too. Mario, go ahead.

Mario: No, I was going to echo that, which is you accumulate all of this wealth and social capital that has no transfer system. It's like owning old money from some ancient empire that now you can't use in the present day unless you are in that milieu.

David: It's not just the Robux, it's also the stuff that you buy with the Robux—your avatar, your skins, your outfits, all your [...] you have in the game. You’re not taking it to Fortnite.

Mario: That allows you to flex.

David: Network economies, duh. Roblox with no friends is not very interesting.

Ben: It's like me being on Roblox.

David: Process power is so elusive. The canonical definition Hamilton uses for process power is the Toyota production system. I don't think something like that exists here.

Ben: It's a rare one to have, as is cornered resource.

Mario: The only argument I would make is by using Lua and having that system built around it, have they made it uniquely easier to create games? Would that fit the bill or is that too far?

David: It's interesting. It's definitely easier than Unity and Unreal. It's not as easy as Rec Room though.

Mario: So it's probably not. It feels like it has to be a little more unique.

Ben: I'm trying to remember the formal definition of process power. The thing that I do remember uniquely about it is that it can't really be fully duplicated because it's so complex that you can't pick it up and bring it to another organization because there are so many components. You can't write it all down in a book and be like, go execute this book.

David: That's why Hamilton used the Toyota production system. They open-sourced it. They were like all other car companies, other industries, come to Japan, learn, study from us. Nobody can do it like Toyota. We're not trying to, it's just so embedded in the organization.

Mario: That's pretty cool.

David: Branding is interesting.

Ben: I'm always tempted to use branding, but often, it's something else masquerading underneath. The way I always think about this is if you have two commodities that you're holding in each hand, and one is branded with Tiffany's and one is branded with Ben's ABC item that you can't charge or generate nearly as much profit on Ben's ABC item. I don't really know that that exists here.

I think every reason why Roblox can make more money than anybody else is attributed to all the other powers that have built up the success of the business, not because they have this brand that people are willing to pay more for.

Mario: That's probably right. There may be some argument that they've taken some of the brand magic from Lego and reapplied it in a slightly different way and can piggy-back off that to some extent. You're right, I don't think it feels like that's one of their core strengths.

David: They have—shoot I'm blanking on the name—but the Chinese competitor.

Mario: Reworld.

David: Yeah, Reworld is doing the same thing to them. That doesn't feel that powerful to me. I think they do have cornered resources though. I think the cornered resources are the existing games and experiences on the system. You can't port those out. They're not going to work outside of Roblox.

Ben: But the definition of cornered resource is something that you can port out, that can be identically valuable for a competitor.

David: I think it's that competitors can't access.

Ben: I was just thinking about this from the way that Packy defined this in the Supersapiens idea. That contract would be equally valuable to another competitor, but Supersapiens was the one who locked it down.

Mario: But then didn't we get the Lithium mining guy who was like, hell no that's not a cornered resource.

Ben: That's right.

David: I think Hamilton uses patents. Preferential access to valuable either resources such as a blockbuster drug.

Ben: But those games aren't equally valuable to another competitor, that's what I'm saying. Another competitor doesn't have the [...] that makes it valuable—the user base, the Roblox.

Mario: It feels like all of these definitions are almost a little bit on a spectrum. Maybe that Helmer would be like, absolutely not. I know what you mean, David. It feels like a very valuable thing that they have sole proprietary ownership over.

David: We may be butchering the definition of cornered resources. If we are, we apologize. But I don't think that these games—even if the developers were to leave, the games that exist in Roblox cannot exist elsewhere.

Ben: Why not? I totally think they can.

David: They were written in Lua for Roblox. You could go recreate them, but you can't move the games and the user bases over.

Ben: Their user bases, that's the thing that you're throwing out. That's why I don't think it's a cornered resource because the user base isn't a cornered resource. You literally could go write a thing that takes Lua and spits out something that looks like that Lua-based game in the Roblox runtime. Maybe the Roblox runtime is a cornered resource?

Mario: Isn't the IP potentially a cornered resource from the game?

Ben: Probably. That seems fair that you couldn't ship an iOS game with the same IP as the Roblox game without getting sued. That feels reasonable, but is that the reason why Roblox is able to be super profitable? I don't think so.

Mario: Agreed.

David: Okay, maybe no cornered resource.

Ben: I only try to be strict on this because I feel like I could apply all of them to every business. Then I go and reread the book and I'm like, Hamilton was actually way more narrow in how he defined this than I've always thought about it. Semantics—it's a fun section to do. We end up debating what did Hamilton really mean half the time.

David: The answer is we just need Hamilton to join us for part two. We should just dial him up.

Ben: Here’s the power section. We’re calling Hamilton just for this moment.

Mario: To play judge. We can each plead our cases.

Ben: As we wind down here, we should say again, not investment advice. How are you guys thinking about if this thing goes public? How interested are you in adding it to your portfolio?

Mario: I'm interested, let me say that. It feels very hard to say these days because of the pricing that you might get. But long-term, I will definitely be looking for opportunities to buy. I may do some amount on IPO day, but I'm going to keep a close tab on it if I ever feel like it gets to an exciting place so that I can double up and triple up because it's an insanely powerful and impressive business.

David: I'm on a similar boat. It's interesting. I don't know what I'm going to do yet. Some of it will certainly depend on the price. I say some because of what I'm about to say—it's not often that you see a business with this kind of power and this kind of potential. It's very, very rare. I very much believe that the management team that Dave, that the current shareholder base behind the company is thinking extremely long-term here. It makes it very attractive.

In ordinary circumstances, all else being equal, this is one of the very few things that I think about as a candidate for what I call my compounding portfolio where I think I have only five equities in there right now—Amazon, Spotify, Tencent, Zoom, and Berkshire—that I think of as long-term. I can hold this for a decade plus, two decades, and be happy. I think this is a candidate for that.

I am thinking a lot about the Rec Room age dynamic. If Rec Room were going public and we were able to invest in that, which would I rather? I probably would want both and just play the theme as a whole. The price is interesting. This is probably going to go out at a nutso price.

I was thinking similar things with DoorDash, Snowflake, and even Zoom at the original IPO. This price is too expensive. Just like you Mario, I'm going to wait for an opportunity. But there was no opportunity. I don't know what to do in this environment about this, but that's how I'm thinking about it.

Ben: I don't know what to do in this environment about this is the bow that we can put on many episodes, and is a great place to leave this one. Mario, really awesome to do this collaboration with you.

David: Ben, you can’t get out of it.

Ben: I feel like I'm the least educated on this one. I just don't do as much of this as I think you do, David. I would be very interested in this company, but I'm not going to go out and individually buy the stock. I'm going to watch it, I'm going to be very curious, I'll follow it. I hope to do a future Acquired episode on it.

Mario: If it was a SPAC, you'd be all over this.

Ben: The SPAC strategy is a little different. I can articulate that another time.

David: I love it. Which as it should be. if you're not willing—for whatever reason—to follow this thing for a decade, at least my view is you shouldn't invest in something if it’s not something you're willing to follow for a decade.

Ben: That's one strategy. That's the time-tested strategy where you can time the market, but imagine you could.

David: Oh boy, the world we live in.

Ben: Indeed. Mario, one more time, where can folks from the Acquired LP community find you, find The Generalist, and read more?

Mario: You can check out readthegeneralist.com. I would love you to check it out and see if you're interested in learning about more S-1s and companies in the private markets. You can follow me on Twitter. I’m @mariodgabriele. Thank you guys so much for having me on. This was a real treat. I really appreciate it.

Ben: Super fun. You should do more audio. For any new folks who are listening to this from The Generalist, you should email Mario and tell him he should do more podcast stuff because this is great.

Mario: Please tell me that because I'm very nervous. I texted these guys right before this being like, I am absolutely shaking to do this.

Ben: For anyone who's an LP or for anyone who's from The Generalist, we do have a slack—acquired.fm/slack. That is a great community to talk about all things—current tech news going on, current investing going on—and it's a really fun place to hop in there, hang out, and discuss all this stuff with everyone. Go check it out.

If you're new, check out the main show. Most of our content is free and are these three-hour epic deep-dives into most recently Bitcoin, but usually companies. This is a much shorter, more abbreviated, more conversational casual version. Mario, much like this is more casual and conversational than your writing. Which every sentence has so much gravitas when I read it. I love it.

Mario: What a charmer.

David: Only the three of us could be like, an hour and forty-minute episode, this is brief.

Ben: Nice and tight, yeah. Awesome. Listeners, see you next time.

David: We'll see you next time.

Mario: Thanks so much.