How to Build a Multi Product Business with Carta’s CEO and CMO

Thank you. Thanks for, thanks for having us. We got 15 minutes to talk about building a multi-product business, so we'll try to get as much done as we can. We'll talk about products and markets, and the most important thing is the connection between the two. So let me start with how we think about products, and we talk about this concept a lot at CardUp called atomic units, and I talk to product managers and founders about this. And it's easier to describe than it is to define. So let me just give you some examples. I would argue the atomic unit of the Insta— Instagram platform is the photo, right? Everything's built around the experience of sharing a photo, commenting on a photo, liking a photo, uploading photos. I would argue the same is true for Airbnb, where the atomic unit of the Airbnb platform is the rental unit, right? It's one night in somebody else's unit, and the entire platform is built around How do I optimize that experience of creating inventory for units to stay in, and how do I get somebody in that room? Similarly, Slack, simple, easy one. The atomic unit is the message. It's how do I get a message from point A to point B, whether it's threads, DMs, channels, connections, all of it is around how I get messages from point A to point B. And then of course we're Carta, so we have to talk about the atomic unit at Carta, which is the security, or the asset, right? Everything we do at Carta is built around how do we get an asset from one person to another person? Whether we're tracking on their cap table, their fund. And when we talk to product managers at Carta, everything we build has to be built around this atomic unit. And so we only build 3 types of products at Carta: products to capture assets, products to value assets, and products to transact on assets. And if you look at any product at Carta or any product manager, if they're not building one of those things, we're doing something wrong, because that is the atomic unit of our platform. All right, so that— I'll go to markets.

So if that's how we think about products, how do we think about markets? Any time we enter a new market at Carta, we sit down and we do a little bit of an exercise, like a pre-mortem, basically. And we say, 5 years from now, if you imagine that somebody has completely overtaken a market, has just like dominated that market in a way that nobody could have anticipated, what would they have done to make that true? And usually when we do this exercise, it boils down to about 3 things. The first thing is that they have done something to the microstructure of that market to make an N-of-1 market out of it, to make it winner takes all. The second thing is they will have done something. They will have found a unique business model change, a business model innovation that allows them to really capture that market. They'll found some sort of edge that changes the business model of that market. And the third thing is that they will go ahead and they will find a way to acquire customers faster than anybody else has. And if you think about if the market has become an N of 1 market, a winner-take-all market, it's super important that you're able to capture customers as quickly as possible. So they'll have found some sort of way to acquire customers at a velocity that nobody else can keep up with. Now, notice what's not on that list. There's absolutely no mention of TAM. So long as the microstructure of the market is N of 1, No market is too small for us. It doesn't matter how big it is or how small it is, no matter what the business school books say. The second, and this is kind of, kind of interesting, there's, there is no explicit requirement that the market be immediately tangential. So a lot of times when you're thinking of a new market, you'll think, how is this accretive to the business? How does this fit in with the existing products and services that we have? We don't think about that. All we think about is, do we have edge in that market? Do we have a way to approach that market that is unique to us and not to anybody else. And if we have edge, then it will become obvious to us how it is tangential in the future. As an example, when we entered foreign 9As years ago, cap tables and foreign 9As were totally different. There was no tangentialness to those two markets. Cap tables were software, foreign 9As was accounting. Like it was totally different. And now years later, it's obvious there is no chance you would buy a cap table service with— that's not also doing your foreign 9A. The markets have become the same. And third, there's no requirement that there's edge in technology. You may say, OK, we need technology edge to be able to enter the market, but technology edge isn't enough to sustain your advantage in that market. You need some sort of business model advantage there because eventually all technology will become commodity. Word processing was innovative at one point, and now it is commodity. Cloud storage, innovative at one point, now is commodity. Only technological edge at one point will not suffice to let you keep that market. You need something else.

And so when you think about these products, I build a platform with an atomic unit I've built around it. How do I make this market or this platform an end-of-one market so that— so the microstructure facilitates one and only one winner? There's a lot of ways to do this, economies of scale and so on. But the most common way in software is network effects. And we talk about— there's many forms of network effects. I'll talk about 3 specifically that we think a lot about at Carta. The first is relationship network effect. And so when you think of this, you think about the obvious, you know, Instagram, LinkedIn, Facebook, and so on. The second is data network effects, and you have lots of these companies like Google, Bloomberg. You could argue Facebook also has a relationship network effect and a data network effect. And then of course there's liquidity network effect, and these are your marketplaces, NASDAQ, NYSE, Airbnb. EBay. What's really interesting is the best companies, the best software companies in the world, they'll have to get one of these right. You can't be a software company without getting a real— a big software company without having some form of network effect. The really generational ones will get like two. So, you know, Facebook will get relationship network effect and data network effect. They're trying to get liquidity network effect with Libra and other forms of payments. They're not— I wouldn't— I would argue they're not quite there yet, but they're definitely working on it. There may be only one company in the world that I think gets all three, and I'm I would probably say it's WeChat and WePay. They've figured that— figured this out, and it's an incredibly valuable and defensible proposition. Obviously, we think a lot about this where we're mapping products to, hey, our cap table is a relationship network effect product. Our 49A is a data network effect product. And then our liquidity or Carta X offering is a liquidity network effect. And we're trying to layer all of these things on together. And then when you think about network effects, they're really simple. Like, they're hard to get going, but they're, they're really simple to understand because once you get going, you just define what the North Star is, right? So if it's a network effect for Airbnb, it's the number of rooms rented, right? If it's a data network effect, it's the number of times that people are accessing Google Search and getting information out of that. If it's a relationship network effect, it's the viral coefficient of a new person brings how many new relationships to the network. And these things get very easy to measure. The one caution that I always give product managers and founders when they're building network effect businesses is they'll always say, well, once I have the network, it'll be really valuable. And they're always right, almost by definition. It's a tautology. The hard part is how do you, how do you create the network? How do you bootstrap in? And there's this, this phrase that I got from Andreessen Horowitz, which is come for the tool, stay for the network. And that's what everybody misses when they build network businesses, is you have to have a tool that somebody, that people will come for first. To bootstrap into the network. There has to be intrinsic value, uh, for the first person because you can't get n+1 unless you get n. And you need 0 to 1 first, uh, and you need someone to— you need a tool that people will come for that. And then once that start— flywheel starts to spin, then they'll stay for the network effect, and that's what creates the defensibility.

So when you've built that tool, how do you start to map the network? Well, it's really, again, it's pretty simple. It's hard to execute, but it's simple to understand. You start with that central node and you simply pull the thread of that atomic unit and see how it interacts with other nodes along the network. And what you'll do is you do that is you'll start to create a graph of not only all the different nodes in the network, but the edges between them. And there's opportunity in both. So taking an example, what did we do at Carta? We started with companies, the central node of our network, and we used the atomic unit, our asset, and we just mapped How does that interact across the rest of the network? So companies issue options to their employees. Another node, another edge that connects them. And they also issue securities to their venture firms, their investors. Another node, another edge that connects them. And what's really compelling about these network effects is that the further and further you pull that thread, the more you discover. So actually, the venture firms are also issuing membership units to their LPs. And what's so special about the ownership graph that we're mapping is that actually if you pull from the company to the investor to the LP through the pension fund, the shared ownership vehicles, ultimately you end up at an individual, at a person, at someone like you and their 401(k) or their retirement plan. And so what's really special about this network is all the leaf nodes are just people.

And I'll add one thing about the topology of these networks is really important. And so, for example, in our world at Carta, We have a heterogeneous network. Each node is different, right? A company node is different from a venture fund node, which is different from an employee node, different from an LP node. But there's also homogeneous networks like Airbnb or Uber, right, where you have a driver is a driver. It's commoditized. Each node is exactly the same. What you're, what you're differentiating on is the usage against those nodes. And when you think about your own network businesses, I'll encourage you to think a lot about the topology of these businesses. I'll give you one more example. The difference between Airbnb Airbnb and Uber. Both are liquidity network effect businesses, but I would argue that topology is very different. One is what I'll call a global topology, and the other is a local topology. So Airbnb, an apartment in Paris helps me as a user, as an apartment hunter in San Francisco. There's a global connectivity of each node that brings— that Airbnb brings into the platform helps every other node in that platform. But Uber, a new taxi driver in Paris doesn't really help me. I mean, you might argue on the margin if I travel to Paris, but the instant value of that new, that new Uber driver in Paris doesn't do anything for me. And so it's a local topology. The topology of network effect is localized to the cities. And I would argue that Airbnb has a better business than Uber because it's much easier to defend a global network topology than a local topology where Uber has to defend each city. On its own because they can't combine the cities together in one global topology. If you think about this graph, so whether it's a homogeneous or heterogeneous network, the playbook is really the same. Once you've connected these nodes together, you just sell services and software into the nodes, right? So you monetize the driver, you monetize the passenger, you monetize the room, you monetize the renter. You just sell these services and software to the nodes. And then for most of these companies, especially the fintech ones, the edges are really simple. It's the pipelines of cash movement. It's where the money moves. I've moved money from a renter to a rentee, and all of the money moves through these platforms between nodes. And for us, it's really simple, right? We have companies on the platform issuing stock certificates. We move money for them. We have venture funds calling LP committed capital, and they're doing distributions, and we move money for them. And that's how we expand out the network. Oops, I'm sorry. I went the other direction. So anyway, the key takeaway is we'd walk away— I'm running out of time, sorry— is the 3 things that I think need to be true when you think about these network effect businesses for yourself, whether you're building a startup on your own or you're building a new division within your own company, is that the, the most important thing in these network businesses is that you have to figure out how to make it an number one market. By definition, if it isn't a winner-take-all market, it can't be a real network effect business because it will fragment and you don't get the value of the network effect. The second is the most important thing around this business model innovation is it's the only durable advantage that you have, right? If you read like the book of Loonshots, right, what was the durable advantage of American Airlines versus Pan Am? It wasn't inventing new airplanes. It was that they had a better business model that would endure, because technology, by definition, you will always lose that over time, right? It's sort of the old Sun Microsystems adage: there's way more smarter people outside your company than inside, and somebody is going to beat you at technology at some point. And so you need that durability. And then the high-velocity customer acquisition model. The only thing that makes venture companies venture companies is that they— you know, the old question, what's the difference between a startup and a small business? The startup was designed to grow fast from day one. And so you— to grow fast, you need a high-velocity customer acquisition model. Otherwise, the venture economics don't work. That's it. Thank you. Come visit us at our booth. We'd love to meet you. Thank you.

Thank you so much.