The Exchanges

Q That's complicated, right? Because you can have a user that's not the same person as the buyer, and if you make your product fully available to everyone, you may have use cases that you don't really want. How do you think of that?

A So it's, it's, as far as we can tell, it's way more complicated than building a consumer company, um, in that, in a consumer company, you know, to first order, if you have users, you can monetize. Right? And so, like, basically, all you wanted to do was get users, then the assumption was is that you could use advertising or something to monetize it. And even that's changing now, by the way, um, as GAFA takes control. But that's kind of the way you thought about it. In the enterprise, that's not the case at all. I mean, you can have an organic motion that you simply can't build a sales motion on. So I'll give you a few examples. One of them is you can have an organic motion that sell, that is adopted by one person in an organization, but the buyer is someone who is entirely different, and the person that organically adopted doesn't have budget. Another common failure mode is, let's say you have an organic engine, so you're, like, a bunch of people are using this product, ah, but you made it too good, and so now you've cannibalized any ability to monetize later. Um, another problem that I'm seeing in a lot of companies is, let's say you build a good organic motion, but not a great one. So you're doing an open source project, and it's pretty good, but not great. The problem is, is now if you build sales, as soon as your sales surpasses your organic motion, you're SOL. It's really …

AI assessment note: “adopted by one person in an organization, but the buyer is someone who is entirely different”

Q Is that true both at the beginning of a venture and at scale?

A Yeah, so no, so I think that, like, it's, it's very dangerous, like, um, to do services, because you can really get addicted to that revenue, and you can build your company to assume it's gonna come in, right, like, from the board on down. So I think that keeping service separate as a separate P&L has to be a first-class problem, if that makes sense, and everybody understands this. And eventually, you do want the partner ecosystem To take on those services, right? You want the boutique bars. You want the integrators. You want the consultants to actually do that relatively low margin business. So I would recommend early on, keep it separate. Use it for account control. In my experience, when you hit about 30 million-ish dollars a year, you actually have enough of a market to actually incent the partners. By the way, it's very difficult to get partners to do your services before you have a market anyways. I mean, you're some pipsqueak company there. They have some pipsqueak product that nobody knows. To actually get someone to spend the time to learn the product, help you sell it, and integrate it is hard. So I'd say you do that work, you get the account control that you're doing it, you get the credibility while you're doing it, and then bring the partner ecosystem later once you have an actual business to offload to them. But make sure your board knows this, because otherwise y…

AI assessment note: “In my experience, when you hit about 30 million-ish dollars a year”

Q That's, that's, um, that's great. I guess there's also the, um, issue of, um, the big cloud providers, um, taking the open source code and creating their own instance of it. Is that, is that something you worry about when possibly investing into an open source company?

A So I personally don't worry about it a lot. I mean, I know that there's a lot of heartburn and angst around this, and the heartburn and angst is the following. It's, you know, you've got these huge cloud incumbents. As soon as you make a popular project, they can kind of host it, and as a result, you know, they'll erode your value. And I think for very mature products that have been around a long time, there's some truth to that. Um, but I think, like, listen, if you're in the first hundred million dollars Because of revenue, and you're the innovator, and you own the code, I think it's incredibly difficult for these These larger incumbents, you know, Google, Amazon, and Microsoft to compete against with you. They can't build a sales force that will, that will articulate the value to the customer. They can't keep up with the versions. They can't operationalize it. I mean, I think it's very difficult. And so as a VC, I'm definitely in the camp of, you know, I'm not worried about the incumbents for new markets, but I do think that if you have a popular open source project and it's hit maturity, Like, it's been around for a long time. It's something you really have to think about. Um, but I do think from my, for, for my investing life cycle, it's not something I, I, I, I deal with. That said, every re-invent, I swear, I play therapist for, you know, I'm on 12 boards, and I, I reall…

AI assessment note: “So I personally don't worry about it a lot.”

Q What do you make of current, you know, the current state of the VC market where, like, every year, um, you know, rent sizes get bigger, valuations get higher, and it just keeps going and going and going. Do you care, or is that irrelevant?

A You know, I, you know, I don't even look at the VC market. I look at the market. Man, the market's enormous. Man, these companies are bigger than we ever thought they would be. Right? I mean, a taxi company, man, it's forty billion dollars later, right? Like, Facebook is larger than anybody thought it would be, Salesforce is larger than anybody thought it would be. I mean, here's the way that I view it. I mean, I view, like, the first, you know, couple of decades, the internet was about information processing, and now it's, like, grown, like, eyes, ears, arms, and legs, and it's going after every mature business on the planet, whether it's forestry, or it's mining, or it's taxis, or it's lodging. I mean, it's business, which is trillions of dollars, and so, For me to worry about the difference between 10,000,020 million valuations seems like such a myopic view of the fact that technology is working, and it's disruptive, and the market is like basically the globe. But I tend to be an optimist on these things.

AI assessment note: “worry about the difference between 10,000,020 million valuations seems like such a myopic view”

Q And that's presumably, presumably going to be a sales-driven effort, right? That's not gonna be, you're not gonna start, like, doing marketing, or is that, is that the correct way of thinking about it? You need to have people that go to buyers, potential buyers, and tell them about the widget?

A Yeah, so I think this is changing over time, and we can talk about that, but traditionally to do market category creation, um, you had to show up. It's kind of one of these things where they say Jesus didn't evangelize over email. It's very difficult to get somebody to actually understand something fundamental if you're not sitting there convincing them and then walking them through the, the kind of intellectual path or maze that you had to go through to understand why, why it's a value. And as a result, most category creating technologies went to market through a direct sales force, where you actually had somebody representing the product normally, by the way, the technical founder to begin with, to describe what it is. Um, so that's, that's typically how market category creating technologies have gone to market. However, I would say that this is actually changing a little bit.

AI assessment note: “most category creating technologies went to market through a direct sales force”

Q So the typical thing for VCs to say and think is that in enterprise software you should largely avoid, um, services because they don't scale, they don't have the right margin profile. What do you think?

A Yeah, so I, I actually think, ok, so traditionally in the enterprise, the world worked as follows. I mean, in, in enterprise software, which is, you have a team that understands the space, they create software, you give it to a sales force, and then they sell it, and they sell it just as software license, and as a result, ah, you have the traditional software margins, right? So it's direct sale, traditional 80% margins. I think that model is basically going away. It's It's being eaten on the bottom by bottom up, which we can talk about later, and I believe on the top it's been eaten by services. So if you go to most VCs and you're like, hey, listen, I'm gonna build this company with a heavy services component, they'll, you know, give you, like, these common lines, and it's hard to scale, it's like low margin, blah, blah, blah, blah, blah. But here's the reality. The reality is for every dollar that most customers spend on a vendor product, they're spending two or three dollars on services. Services. And by services, I mean things like contract engineering, or integration, or even just, you know, traditional pro-serve, right? And those dollars normally go to Capgemini, DiData, Infosys. There's a whole ecosystem that gets these services dollars, right? So, there's more of the services dollars, but more importantly, the, the services, whoever provides the services becomes kind of …

AI assessment note: “for every dollar that most customers spend on a vendor product, they're spending two”

Q Actually, you know, the question that, um, everybody always wants to know, and I know for a fact there's no Hard and fast rule, but what is this? Like, can you quantify, like, what is traction in that world? Like Series A or Series B?

A Yeah, yeah, yeah, yeah. So, um, ok, so let's assume that we're past a team bet. Like, any VC will invest in just a team because it's a team. And that's a team bet, it tends to be low dollar amounts, and that will never change, right? But let's say that we're actually investing in, like, the idea of a product or something like that. So, here's, here's, you know, I don't have any hard and fast rules, but a couple things. Like, Um, we're, we're absolutely concerned with the second derivative. It has to be positive, right? Linear growth, by the way, doesn't seem to work. So, like, the curve needs to look like this. Like, the, the, the curve needs to be pointing up. Um, the second thing is different technologies, um, have different types of, um, engagement metrics. If you're writing something like core that's gonna be in the Linux kernel that's some nerdy thing, it's gonna look very, very different than if you're working on, like, a front-end stack, right? If you look at something like, Vue, or Gatsby, or Hugo. I mean, those are hugely popular, but, but you have to notice that these tend to be frothy spaces. So what I would recommend is, look at other projects that are in a similar space, understand the metrics, whether it's GitHub stars, or it's downloads, or it's community, or it's mailing lists, or whatever you want to track. Understand, kind of, what your peer set is, and make s…

AI assessment note: “we're absolutely concerned with the second derivative. It has to be positive”

Q And, um, maybe just to, uh, close for my, my, my questions, so I want to open up two people. Um, does this strategy, does that have any impact on fundraising strategies? Is that more or less the same thing, but you need more capital, less capital? Uh, should you raise more early or not?

A Yeah, so, so I do think that this trend, this, this, this kind of macro trend of bottom up is an advantage to startups for a number of reasons. Number one, it plays to your strengths of being the product experts, not necessarily customer experts, right? You know, you can iterate. It needs to be simple. You can focus on the product, which tends to require less money to do. So I think, yes, I think that required less money to get traction to understand. I think you can A-B test in early markets, and I think it plays your strengths at really understanding products. The other side of that is, I do think that we're moving away from the days where you can show up with a slide deck and, and the core technology and, and get real funding. And the reason is, is, you know, just like I can't predict what consumers are gonna like, what dog walking app, like we, I can't predict what enterprise are gonna like, what open source app, or what online service. Nobody can right now. I mean, it's unbelievable some of the companies we've seen that have enormous traction, um, that seem like very simple things to me. Actually, I would have said that about Slack. I mean, I've been an IRC user since, like, undergrad, right? But it's a total phenomenon with a tremendous amount of value. Um, so I think the other side to that is I do think that you should really focus on bottoms of traction, not revenue. Ju…

AI assessment note: “focus on the product, which tends to require less money to do.”

Q You mentioned open source. Is open source one of the ways of building that wave?

A Yeah, I mean, it is, it is, but I, I almost feel like open source is just being superseded by, by any software as a service, and, and what I mean by that is the following. It's like, as soon as you, you host something as a service, people don't seem to ask so much about open source, right? Like, I mean, you know, GitHub is hugely proprietary, right? But it's based on kind of open interfaces, and, and certainly open source software, You know, AWS is hugely proprietary, but it's based on a lot of open source software, and even if everything is open source, often, like, all of the operational code required to stand up these large environments is proprietary, or there's a bunch of tribal knowledge to run these things. I mean, it's hard to run a large, like, operations, and et cetera, et cetera. So, you know, as our, as the consumption model shifts to as a service, that's how we want to think of services, whether or not it's open source, I think, is, is becoming less, It's becoming less relevant over time, if that makes sense. That said, there are some domains where people still want stuff on premise, and they still want to see open source, and so there, it's good for distribution, but if I were you, I would assume the consumption model is as a service. That's what I would assume, and then I think the decision of whether it's open source or not, it's somewhat irrelevant. Go ahead if…

AI assessment note: “Yeah, I mean, it is, it is, but I, I almost feel like open source is just being superseded”

Q super interesting content by the way people can go online and like on YouTube and on the Anderson Horowitz website that you've written and on Twitter. Uh, written a lot of super interesting things, so maybe let's start with that, uh, this concept of category definition. How do you sell a product when the category does not exist, and the customer may not even know that they need the product?

A Yeah, yeah, so, ok, so this all speaks to my total naivety, uh, coming out of school. Which is like, I always thought, ok, here, I, actually, I think this is probably the number one error technical founders make, and it's the following, right? So, alright, so I'm in my lab, and I make my cool thing, my cool widget, whatever it is, and then the big logical fallacy we all make is that whatever, that widget, that thing we've made, actually has intrinsic value. So it's like, I could go, and I could, like, put it on the sidewalk, and I would leave, And someone would pick it up, and they would understand, oh, this is amazing, it'll change my life, and I'll pay lots of money for it. And I think most technical innovators believe that because they're, they're part of the creation process, they understand why it's cool, they've kind of done all of that background work. But the reality, this is just simply not how markets work. And in fact, I kind of learned this the hard way. So I've done two companies now. I did, I did one while I was, um, uh, uh, in my PhD program, and then this year after my PhD program. Both of them were successful exits, and in both cases, I tried to give away the technology, and nobody would take it. So I decided to start a company. And I think this is illustrative of the, of the following. It's, um, A market category is something where people know what you're sell…

AI assessment note: “A market category is something where people know what you're selling”

Q Yeah. So we'll let, let's turn on category creation just for, for a second. Does this have an impact on how you price your product?

A Yeah, yeah, so, so, so I think probably, and actually, so Ben Horowitz told me, so Ben Horowitz was on my board, so he invested, he was the first investor in, in, uh, in this era, and, um, one of the lessons he taught me that has stuck with me the most is the following. Um, especially in enterprise, there is no single decision that you make that's more important or directly tied to your value than pricing. And the reason that is, is if you look at enterprise companies over time, almost all of the variable cost is go to market. Actually, R&D becomes a sub-linear cost, right? As you have more customers, yes, you need more engineers, but it's sub-linear. But sales, but go to market is linear, and if you look there at your variable cost, you basically have your top line, which is how much money you can get out of customers, and then your CAC. And the CAC, you know, the customer acquisition cost, the cost of a salesperson, that's set by the market. You don't get to set that. So what you really get to set is the pricing. So your margin basically comes down to how much you can price your product, assuming that, um, you're doing direct sales. Right? So pricing is incredibly important. So there's two ways to think about pricing. Most people, technologists like myself, they think bottoms up, right? So you're like, oh, I've got this new widget, and the widget is going to save you 10 horse…

AI assessment note: “I think probably, and actually, so Ben Horowitz told me”

Q So, uh, we alluded, uh, to this a couple of times, and I, I know it's a theme that you're very excited about. Do you want to talk about this combination of, uh, bottoms up sales in the enterprise on one end, and sales on the other end?

A Yeah, yeah. So maybe I'll, I just wanna, I mean, you guys probably know this, so I, I'm not trying to pander, but it actually just helps my, my thought process as I go through this. So, so typically in investing, you look at consumer and enterprise very differently. So let's talk about consumer. So here's the thing about consumer. It's very rarely about core technology. It's more about, like, do you create a product that people like? So if I'm listening to consumer pitches, and I have 10 companies that come in that all walk dogs, I have no idea which one's gonna win, right? It's like answering the question, what are, like, nineteen-year-olds in Toronto like? I have no idea, right? So consumer investing has always been, listen, we can't predict the whim of the crowd, But we can understand trends, so whatever my theory is, the graph of adoption is smarter than my theory, so I'll wait till you have adoption. And the reason is, is when you're in a consumer, you don't have, like, it, like, it takes too much money to reach everybody to send a salesperson, so you have to do marketing, and so you end up doing this kind of marketing-led, um, distribution, which leads to these kind of organic or bottoms-up adoption models. Which is why you have to wait to see whether things are, are, are being adopted by the crowd. Now, sales is very different. Like I mentioned previously, for sales, you…

AI assessment note: “In enterprise, you're like, listen, the enterprise is”