The Hacking Open Source Business Podcast

HOSB - Open Source VC Funding - The Insider View from Joseph Jacks ( OSS Capital )

October 19, 2022 Matt Yonkovit & Avi Press
The Hacking Open Source Business Podcast
HOSB - Open Source VC Funding - The Insider View from Joseph Jacks ( OSS Capital )
Show Notes Transcript

In the 3rd Episode of the Hacking Open Source Business Podcast Joseph Jacks (JJ), the founder of OSS Capital, a VC firm specializing in investments in open source businesses, shares his insider view on building, funding, and growing open source businesses.  Avi and Matt ask JJ to cover his investment strategy, what he looks for in founders, how he evaluates business growth, and more!  If you are a maintainer thinking about trying to get seed funding, an open source business executive evaluating growth, or an aspiring investor in open source you won't want to miss this episode.  

Checkout our other interviews, clips, and videos:

Don’t forget to visit the open-source business community at:

Visit our primary sponsor, Scarf, for tools to help analyze your #opensource growth and adoption:

Subscribe to the podcast on your favorite app:

 Hacking Open Source Business Ep 3

[00:00:00] Matt Yonkovit: Hello everyone, and welcome to another episode of the Hacking Open Source Business podcast. I am one of the co-host Matt Yonkovit, and I'm joined by a Avi Press and our guest today, JJ from O S S Capital. Jj, how you doing today?

[00:00:16] JJ (OSS Capital): I'm doing good, Matt. How are you doing?

[00:00:17] Matt Yonkovit: Good. Good. , maybe if we could start, could you tell us a little bit about yourself, and about OSS Capital? 

[00:00:25] JJ (OSS Capital): Sure. So myself, I was born in, Long Beach, California in 1989, so I'm, 33 years old. And, I was, homeschooled from the age of, two to, 15 or so when, I graduated high school at home. So that means my parents went to the party store and bought a diploma and handed it to me. And , then I started working.

[00:00:46] I learned about open source. I was like one of the early sales people at this, uh, French software company called Talent, which, , was, ,well known. And there's still around, , for like ETL software, like data integration, data movement software. So like kind of a technical enterprise software [00:01:00] product.

[00:01:00] But they had this open source community and so kind of observed that their, um, way of building products and hiring people and building community was just like really different as as an employee there. And just like working in sales, you kind of touch all these different functions and you really get, get a feel for like what the company's like culturally and structurally and stuff.

[00:01:18] And then I worked in a couple of other, uh, enterprise software companies after that and really reinforced, like working in a totally proprietary software company, yet you really see how different, open source company is versus like a, a proprietary company. So that kind of stuck, stuck with me for.

[00:01:33] A while and, I was very lucky, uh, to learn about the Kubernetes project in 2014, and start the very first, like Kubernetes startup, like training services, tools, um, consulting stuff. Built a couple of, um, you know, web UI things around Kubernetes. And then I also, uh, designed and started co Kubecon, which is the Kubernetes Community Conference.

[00:01:52] And I was very, very fortunate and lucky to, to do that. And it was kind of an obvious idea at the time. Kubernetes ended up becoming this huge project. You know, [00:02:00] I do zero to me, like next to no contribution to that, but, uh, ton of amazing people from Google and Red Hat. A lot of the companies and the, the Linux Foundation, um, created this sort of mini foundation called the cncf, which is like, Google donated Kubernetes too.

[00:02:16] And, um, I was involved in that and, uh, donated Kubecon to that organization. And then, um, and then I worked on another startup that kind of died, and it was basically like a open source programming language for distributed systems. Um, like a compiler that would like let you, build, software without, uh, needing to kind of manage all these disparate tools like tracing library and locking server and, you know, cluster management and, web proxy and all these different things.

[00:02:39] Like let the compiler kind of manage that. Um, but a little bit too futuristic. Um, for, for, for, for, for the time. then I started blogging basically about a lot of these observations that I mentioned earlier around like why open source companies are different than, you know, fully proprietary software companies.

[00:02:54] And that blogging kind of led into really like a kind of an investment thesis because, um, [00:03:00] over the prior years, I kind of got to know a bunch of VCs and the sort of the conventional, um, kind of VC approach to, to these types of companies was, was, was really clear , in that, you know, the, the, the vc uh, community at the time was sort of very much like, you know, open sources are really great kind of marketing and lead gen approach and sort of like, that's it basically.

[00:03:18] So, so, so sort of like the beginning and the end and my, um, observations and kind of points of view and which I'm still kind of, you know, learning more about myself and kind of thinking more about, uh, as I've been doing for the last few years is really that it's, it's much more holistic and it's kind of.

[00:03:32] You know, fundamentally, uh, different on, on kind of all dimensions. It changes everything. It changes, you know, culture and hiring and go to market and business models and you know, of course also customer acquisition and, and lead gen and that sort of stuff. But, uh, changes really everything. Um, and so, The blogging evolved into, the observation that, you know, may maybe there should be a fund or VC focused exclusively on, investing in, in, uh, in, in open source startups and doing nothing else but [00:04:00] that.

[00:04:00] And so I was very lucky timing wise because the last four and a half years we've been in, well not, maybe not withstanding the last year, uh, the biggest bull market in the history of money in markets. Um, so I was able to raise our, our first, uh, fund, uh, for a 50 million fund and, um, make about, uh, 35 investments and, um, also raise our second fund, which is about the same size, and, um, been investing out of that, uh, for the last year.

[00:04:26] And, um, yeah, OSS Capital was started in September 18 and we're, um, we're still the only fund, uh, we're very small, very early stage focused fund, but we're still the only fund focused exclusively on, um, what we call open source startups or commercial open source startups. Uh, a lot of blogging also started a conference that Avi spoke at, um, called Open Core Summit, um, which, uh, we got really busy.

[00:04:48] We were gonna do the third one and, you know, the fund started to get a little bigger. So we still plan on continuing to do open Core Summit, but started Open Core Summit in, um, late 19, about a year after the, the fund was started. And we did the first [00:05:00] one in person before Covid and the second one happened online, uh, using hopin.

[00:05:03] And that was in late 2020. I was a, I think the one that obviously spoke at and, um, the third one, uh, was supposed to be happened, uh, last late last year. But, um, you know, things, things got a little bit busy. But yeah, that's, that's basically my background. We're a pretty small team. It's, it's me and, uh, three other people.

[00:05:18] My partner Heather mek and I, um, run the firm and she's a very well known licensing attorney with a focus on open source, uh, licensing specifically for like 20 years. And, she's really awesome. And, uh, yeah, that's about, that's about it.

[00:05:29] Matt Yonkovit: Okay. You know, you mentioned you've invested in a lot of different companies. Could you maybe just give us a couple of the companies that, are in your portfolio right now?

[00:05:36] JJ (OSS Capital): Sure. Um, One that's probably pretty well known is Uh, cal, like uc, Berkeley cal, or calendar, uh, ca l And it's, it's basically a, a, an application that lets you build an interface for booking your, uh, you know, booking your calendar. Um, so you could have like or And, you know, you give people a way to basically find time, [00:06:00] uh, and book you, uh, in like two clicks.

[00:06:02] Um, so the closest equivalent to, to something like. Uh, is Calendly, which a lot of people may have, may have heard of, or Savvy Cal. There's a lot of tools. The difference with is it's an open source project, and so the whole code base is open, the whole code base is extensible. Anyone can play with it and use it and extend it and build tools around it.

[00:06:17] And there's a really active developer community, over a hundred contributors. Um, it's like a, a top, you know, top 0.1% whatever opensource project. And, and in the, in the getup space, uh, and has, uh, a lot of users. And so their business is, um, really interesting. Another company that we've invested in, um, maybe that you, you might know of is, uh, in the developer community like space.

[00:06:38] It's called Hopscotch. Um, it's like a developer, tool for building, uh, APIs and testing them and kind of versioning and, um, iterating on the API design, uh, sort of stage, um, before you sort of roll out an API and put it in production, kind of maybe closest equivalent to, uh, in the sort of proprietary software world to something like that would be something like Postman.

[00:06:55] Avi Press: Yep. Yep, yep. Yep.

[00:06:56] JJ (OSS Capital): um, uh, uh, other tools. So yeah, we've invested [00:07:00] in a lot of developer oriented tools, infrastructure tools. We've also invested in a lot of application level things. Um, our latest, uh, investment we did, um, last year, but we just announced it a couple days ago, is no good db, which is like an open source air table, um, which gives you kind of like a smart spreadsheet kind of interface on top of any data store or, or any

[00:07:18] Avi Press: cool product.

[00:07:19] JJ (OSS Capital): Yeah. They've been growing super fast and, you know, zero to millions of downloads and that sort of thing in

[00:07:24] Avi Press: So with these companies, um, the, the theme to those three, at least it's not the case for all of your companies, is that it's, you know, the open source version of X. And so I'm very curious, um, how do you think about when that is, when that's a good approach for a venture back business, and when is it not?

[00:07:41] Like how do you evaluate when someone comes to you with, Oh, I'm building open source x.

[00:07:46] JJ (OSS Capital): Yeah. I mean, it's, it's kind of almost become a meme at this point. Like there's like the open source version of. As like a formulaic thing. Like developers are like, Oh my gosh, I'm just gonna, you know, the moment this big proprietary, you know, space starts to [00:08:00] become like, really, uh, clearly a thing. I'm just gonna build an open source version of it.

[00:08:04] And, um, so pretty sensitive to that. We don't actually really go out and try and find the open source versions of big successful proprietary software products. Um, systematically. It's much more like looking out in the world and trying to identify people who are very authentically, you know, driven to create something, um, that is either solving their own very particular problem or they're just like super passionately connected to it.

[00:08:30] So like with, for example, peer, peer, the, the main founder was previously running a recruiting startup. And one of the big workflows and use cases in running recruiting service are constantly meeting tons of people and interacting with candidates. And, um, he, he wanted to build some customizations and extensions around Calendly and realized he couldn't because it was proprietary.

[00:08:48] And so he was like, Wow, I'm just gonna, you know, there's gotta be an open source thing that's kind of like this. He went to Google and he was like, open source version or, you know, alternative to calendarly. Nothing came up and he was like, Gosh, I'm just gonna build this. Um, same story with no good [00:09:00] db uh, uh, Navi, the founder was building, um, uh, a tool that let you build rest and graph graph Q APIs on top of my sequel.

[00:09:08] And cuz he just wanted to have a nice interface to be able to interact at the application layer with those different databases without needing to build all this custom infrastructure and uh, uh, you know, glue between the database and the application. Um, and ultimately you realized like what would be more interesting is if you could actually connect that, uh, user experience to the end user.

[00:09:29] Um, In a spreadsheet like interface in, in a way that a human can understand as opposed to, you know, making it easier for programmers to like write data to a database. Uh, and that's obviously a much more interesting, much larger market in terms of application software. Um, but it was built authentically to solve his initial kind of pain point around, you know, that, that, that, uh, that use case of, you know, building rest or graph APIs on top of my sequel.

[00:09:52] And so he kind of like morphed the project in that direction. Um, you know, with Li and Postman, you know, he really just wanted a more performant, lighter weight, [00:10:00] um, browser native version of testing client for building APIs. And it was less about like, here's an, here's an open source alternative to Postman.

[00:10:07] He was like really solving his own problem. Um, Hopscotch was actually originally called Post Woman, and Post man actually like, kind of went after him and he had to like, rebrand the project to post to to, to hopscotch. Um, you know, so we don't really kind of systematically go out and do that. However, it is our thesis really, like, you know, we think that like, so this is kind of like our slogan, we say like, Software Z in the world.

[00:10:28] We think open source is eating software faster than software's eating in the world. So like, there's this, there's two curves that you could visualize. There's like a big curve of like software in the world and that's like a huge curve at this point. There's a sort of a slope on that curve that's kind of starting to slow though, because like so much of the world has been eaten by software.

[00:10:47] So it's a giant, you know, $10 trillion of value captured by, you know, all the consumer internet companies, all the SAS products, all the big, uh, uh, proprietary software companies. But the area under the curve of open source eating [00:11:00] software. It's a tinier, it's much smaller area under that curve, but the curve is growing faster.

[00:11:05] So that's basically what we see as, uh, a much more interesting and exciting and compelling opportunity. It's not that building an open source company is the only way, and, and that's the only path to like success or, you know, that that's, that's the right way or the wrong way. It's, it's just the, the thing that we happen to focus on, and it allows us to have a lot of, um, clarity.

[00:11:24] Going back to your question about like, yeah, you know, there's a lot of markets that you could look at and say if they're large enough and if they're sort of, um, exhibiting certain characteristics that show there's a slowing rate of innovation, or there's perhaps some monopolization or some kind of oligopoly dynamic forming there.

[00:11:42] Open source is just this incredible force for accelerating innovation, um, accelerating commoditization and as a consequence, fundamentally changing market dynamics and, and characteristics of markets, uh, which obviously unlock, you know, new business opportunities.

[00:11:58] Avi Press: Right, and [00:12:00] so I figure. Listeners that, might be working on their own projects exactly in this space and think that they have, , struck on one of those areas potentially, um, when you're talking about how, how you're evaluating, um, potential investment opportunities, and I think this is very similar to a lot of seed rounds where, you know, it's really about like the founder and the story and very, very story oriented.

[00:12:23] I'm curious if there's anything, whether it's specific to open source or otherwise, what are you looking for in these, in these commercial open source businesses? And like, what metrics are you looking at to evaluate, uh, those investments? Like, is, is it really all the kind of, um, the more qualitative storytelling or are there other more, um, quantitative things that you're looking at when it comes to an open source project?

[00:12:45] JJ (OSS Capital): I mean, it's always, it's always some combination of qualitative and quantitative. On the qualitative side, it's, you know, heavily weighted on humans and people because you can't boil down humans to, you know, sets of zeros and ones, uh, unless maybe you can't, maybe we're in a simulation in the future possible.

[00:12:59] Cause [00:13:00] that's a separate

[00:13:00] Avi Press: An exercise for the reader.

[00:13:02] JJ (OSS Capital): Yeah. Uh, but humans, humans tend to be these like squishy. Very difficult to express in, you know, mathematical terms, you know, kind of, uh, beings that you really have to get to know, and you have to kind of understand like motivations, drives, uh, aspirations, um, inclinations, uh, you know, sort of personality traits.

[00:13:22] Are you, are you predisposed in a certain way or not? So like, that's just really just human relationship development, understanding. If, if we have, you know, in investing, um, really, and, and I, I've been doing it full time for four years, it's not really that long. Venture capital, private, private equity investing in very early stage companies, you typically have like seven to 10 year feedback select plus to just know if you were kind of right, not exactly right.

[00:13:47] And you're never perfectly right either. You're never perfectly right. You're only partially, right. Right. So like for, for, for example, you can invest in a company and, you know, they could become, you know, you invested a hundred million dollar valuation and then they could become a [00:14:00] $20 billion company.

[00:14:01] You're, you're not perfectly right. You're only partially right, Because that company fast forward five years further, uh, could have been, um, a zero billion dollar company and they were just a $20 billion company because some strategic bought them for that, uh, out of fear that they were gonna disrupt their core business.

[00:14:18] But had they not continued to innovate, maybe they would've died. And so, like you were only right on a temporal dimension at a certain point in time, given certain execution up to that point in time. This is like why there's like so many false positives in private startups and this tech ecosystem in general is people just think that like things are kind of consistently, exponentially linear and they continue going forever, uh, if things continue to grow.

[00:14:41] And the reality is you could have these like false starts that kind of like wa off and, you know, sort of flash in the pan dynamic where something kind of appears to be working but can't, you know, be sustained over a long period of time for, for, for a variety of different reasons that are very case specific.

[00:14:55] So for us it's, it's, it's a, it's an exercise in humility to kind of just go and say, [00:15:00] What we, what we invest in is a function of aligning with our thesis, which is a very, very specific and narrow thing. It's extremely narrow. Um, and, and, and a combination of that with do we think a huge business could be built over a certain period of time that maps to the constraints that we're operating within, which are sort of tenish years.

[00:15:23] Like each fund, uh, that we invest out of has a 10 year life. And so we have to operate within the confines of that in order to sort of say, will that company potentially be potentially a, a, a big, uh, multi-billion dollar, you know, IPO scalable kind of company within a decade from when we make an investment.

[00:15:40] Um, and then, uh, you know, qualitative and quantitative, most of it is highly indexing on people. Are they humble? Are they mission driven? Are they, um, skilled and technical enough to actually execute the technology roadmap necessary to get to a viable product in market? Um, do they have self-awareness?

[00:15:58] Sufficient to allow them to grow and [00:16:00] evolve into many different types of individual as the company grows, uh, is extremely unlikely and rare that you have a founder that stays the same characteristically, mindset wise, plasticity wise, learning wise over the arc of the company. Um, the reason for that is the company goes from five to 10 people to 50 to a hundred to 500, and the founder has to evolve accordingly.

[00:16:18] And if they don't, the company doesn't really succeed unless you bring out, um, uh, you know, outside leadership joins the company or people from the, from the inside get promoted. And so the best companies , are led by people who are very curious and, um, you know, intellectually curious and, um, you know, motivated to constantly learn and are, are humble about that.

[00:16:35] And that's, that's hard to identify sometimes. It's, easy to sort of want to believe that people have those characteristics. Um, but it's extremely difficult to determine. Uh, you know, those things are durable without actually working for, with someone for a long time. The benefit that we have of looking at the open source project is we have some sense of data around the consistency of their efforts and, and a consistency of, um, commitment and interest in [00:17:00] a given area.

[00:17:01] And so that, that gives us some sense of being able to go and say the person's actually very motivated and interested in that domain cuz they've been actually doing it without it being, being paid or without having a compensation mechanism behind it. They just care. Uh, and, and it just so happens that caring and really being, um, you know, personally drawn to a problem space.

[00:17:21] Intrinsically not extrinsically. Meaning, you know, extrinsically is like you do it for over payment and intrinsically is you're doing it for, you know, some sort of altruistic reason or sort of for the sake of art or for the sake of personal, personal gratification. Um, the intrinsic motivation, is really critical to, driving, like mission driven behavior, I believe.

[00:17:43] Um, I, I personally think that very little mission driven behavior is driven extrinsically, uh, little to none. Um, in fact, you know, you hear Elon Musk doing these recent interviews saying that like all of his companies are philanthropy, right? Tesla as a form of philanthropy because it cares about, you know, uh, uh, accelerating, uh, the [00:18:00] sustainable energy.

[00:18:00] Future Boring company cares about Phil Philanthropy cuz it cares about eradicating traffic. Neurolink is philanthropy because it cares about solving, you know, uh, cognitive, diseases. And, people who have, uh, Basically paralysis and so on. And so like, there's, there's that dimension that we really pay close attention to.

[00:18:18] On the metrics side, we don't really look too much at metrics beyond the rate of growth. Like, the rate of growth is probably the number one thing quantitatively. Um, if the rate of growth is very slow over, over, even a short period of time, whether it's weeks or months, that tends to make some impact on, like, that things could be less interesting.

[00:18:35] Um, the, the rate of growth being, you know, somewhere in the neighborhood of like five to 10% per week or 20, 30% per month, consistently over

[00:18:43] Matt Yonkovit: JJ can I ask a question on that specifically? Like, when you talk about rate of growth, how are, how are you looking for people to define that when they're bringing you the pitch deck? Are you looking for, I know some people look at GitHub stars, some people are looking at, you know, Slack users, some people looking at, commit frequency, [00:19:00] other people, you know, downloads.

[00:19:01] Are there certain ones that you say are the growth indicators you're looking for?

[00:19:08] JJ (OSS Capital): Yeah, it's a really great, great, great question, Matt. So, um, I, I have to say, first of all, of our core investments, there was no pitch deck and we engaged with the creators and the founders as humans, um, before they conceptualized the deck and before they conceptualized the business. And so we in fact had to actually have a reasonable sense of conviction well before the creators themselves even had the conviction, which makes it even harder.

[00:19:33] Um, and so by rate of growth and more specifically referred to rate of growth of external contributors is something that really matters and is very interesting. Rate of growth of Get Up Stars is probably. One of the, the least interesting things because it, it's, it's a very, very thin sense of sentiment and sort of popularity.

[00:19:51] It's, it's not completely, lacking, of some insight in terms of the, the, the proxy for, you know, developer excitement or, [00:20:00] um, uh, you know, diffusive interest. But, um, I think that, uh, on its own, it doesn't give you a strong proxy for, um, actual usage. Actual adoption. I think Mitchell was actually just tweeting about this.

[00:20:11] The, the co-founder of Hashi Corp pretty recently. He used the physics analogy of, um, potential energy and kinetic energy, and I believe he, he put the GitHub star metric in the potential energy bucket as opposed to the kinetic energy bucket. Um, which is a, a very strong way of sort of ca categorizing and classifying different types of metrics.

[00:20:31] Uh, I'm not sure if you did this, but I would certainly put the contributor and the rate of growth of contributors in the, in the kinetic energy bucket, meaning it's something more immediately. View a visible in, in the sort of like immediate spectrum of stuff that affects the momentum of something like from a physics perspective.

[00:20:47] Um, and, uh, I don't know, like other metrics that we look at are sort of the rate of the, the, the rate of development, um, execution, which, which, you know, frankly pine boils down to, like the number of [00:21:00] commits you're making into the repository on a month, monthly or weekly or daily basis. Like, that's also very gameable.

[00:21:06] Like, you could have a bot that's just like making, you know, uh, small single line, uh, you know, uh, uh, a commit changes. Um, you know.

[00:21:15] Avi Press: though, how often is that being done?

[00:21:16] JJ (OSS Capital): Yeah. And, and it's,

[00:21:17] Avi Press: it's a probably

[00:21:19] JJ (OSS Capital): easy to spot, you know, if you just look at repo and you're, you don't even need to be a really sophisticated program or like, see if there's like, okay, it's not a human actually making those contributions.

[00:21:27] So I think rate of, rate of iteration, uh, rate of kind of. Not necessarily improvement, but like rate of change, rate of, uh, adaptation is a very, very valuable measuring proxy, even if it's literally, even if it's a single individual, um, in, in the project. Um, just really dedicated to, uh, improving things at a, at a fast rate.

[00:21:48] Um, you know, that's, that's another really big signal we look at. But going back to Abby's question, like I think the number one thing is people, like at the very earliest of stages, it, it really does boil down to is this person highly mission driven? Do they have, we think [00:22:00] the technical skills and the, um, personality traits necessary to manifest their vision from a zero to a one state?

[00:22:09] And then from there, do they have the sort of raw ingredient. Needed that we think constitute enough that would allow them to grow into a person that would be extremely inspiring and motivating to lead an organization That is often the leap of faith. Most people do not have that, especially first time founders.

[00:22:25] They do not have that. They have to learn those things. They have to learn how to be an, an inspiring recruiter and inspiring magnet for talent, um, uh, beyond themselves. They also have to learn how to grow into, like, doing literally everything and wearing all hats to hiring people who are specialists in different domains and, and, and, and enabling those people to be successful.

[00:22:44] Like those, those things you have to learn as a, as a startup founder.

[00:22:47] Matt Yonkovit: Let me dig into a couple of those things that you said. You said a couple things that really intrigued me and I actually got my pen and paper out so I could write 'em down. Uh, a little old school there, but that's okay. So when you're talking about, the founder throughput and their ability [00:23:00] to adjust, are you seeking out projects where you see active founders then?

[00:23:04] So are you looking at, you know, the founders commits the how much they're contributing to the code and the project and then approaching them. Directly and saying like, Oh, we, you know, have you ever thought about this as a business? Um, you also mentioned you're not really doing pitch decks.

[00:23:17] You're starting before they even thought about this as a

[00:23:20] JJ (OSS Capital): Yeah, a a absolutely. Um, most of the time there's been a couple of exceptions, but most of the time, um, we identify something, uh, just by constantly looking at the open source community and new projects that are coming out and things that. Like, look interesting objectively. And then we basically cold outreach, introduce ourselves and uh, I'll personally do that most of the time, like on Twitter or over, or email,

[00:23:46] Avi Press: That's very cool.

[00:23:46] JJ (OSS Capital): or LinkedIn, you know, um, uh, uh, and um, the, the response rate and the, the sort of engagement rate is kind of a hundred percent most of the time because we're super unique and, you know, we're not [00:24:00] really viewed as venture capital and we're not viewed as this sort of like trying to extract information and, you know, for the sake of some, you know, particular agenda or something like, uh, we're actually very interested in like broad mission is like to encourage and, and, and you mentioned this word founder.

[00:24:14] They're not really founders at that point. They're just open source maintainers. They're open source engineers. They're like, there are software developers out there on the internet, like building their stuff in the open with open source. And so like going in, introducing ourselves to them to say, Hey, you could actually be a founder of a company, own the vast majority of it.

[00:24:31] We'll fund you with, you know, end numbers of millions of dollars, and you'll be in control of that company, will help you build a business that can do two things. One, um, allow you to sustain and grow your, your community and your project more systematically than you would otherwise through donations or through through some other approach.

[00:24:49] And two, uh, if you're successful in actually building a business in a commercial enterprise, potentially allow you to create a huge amount of personal wealth as, as well as contribute back to, you know, um, uh, [00:25:00] the world in, in, in a, in a, in a, in a bigger way. And, you know, creating a lot of wealth for other people as well.

[00:25:04] Also us, but also the employees of the company that you, that you build. And so it's a combination of, you know, helping open source creators realize that they could become capitalists and founders of companies while also continuing to kind of be, you know, this is not the right characterization, but sort of socialist communists in building out the open source, you know, communities and continuing to do that.

[00:25:24] We also have a very, very high, uh, uh, level of, um, you know, I guess opinionation is, is the word on open source and what open source means. And so every investment that we've done that we actually strongly encourage, we don't have control over this. Um, but we, we encourage and we recommend that the projects remain open source licensed forever.

[00:25:45] Meaning from the time we invest to the time that the company, you know, over time, you know, becomes huge or not, whatever happens, like continue to maintain an open source license, which means something very specific. Um, there is no other VC on earth. I've actually kind of [00:26:00] more recently realized this and. I, I, I thought it was sort of more safe to say, you know, there may, there may not be another one.

[00:26:08] Concretely, categorically, there is no other BC that, that says, We want you to be very committed to open source with permissive licensing at the core of your technology forever. Like we would like you to do that. Um, uh, in, in fact, the reason that, that there, there is no other, is, this is an extremely counterintuitive thing to do as a capitalist, um, because it introduces massive uncertainties and complexities in terms of how you'll be able to monetize your technology.

[00:26:34] Um, and, uh, so, but we, we, we like to think that there are enough provable cases where this is not only possible with better for the world, and it's actually more sustainable way of building.

[00:26:45] Matt Yonkovit: So JJ, let me ask specifically about that, because this is interesting and, we were gonna talk about this in a, in a bit. It was on my list of things that I really wanted to dive into. When you talk about that open source forever, that that's, that's great. But the definition of open source seems to be a [00:27:00] little changing.

[00:27:01] You know, when we talk about licenses, and I don't know where, you and, you know, the, the OSS capital folks fall on this, but we've seen new licenses kind of come into the space and there's a lot of questions on whether they're truly open source or not. Some call them eventual open source, source available.

[00:27:15] We've seen, you know, BSL, SSPL, Commons Clause. There's other things that have started to kind of float out there. I'm not a negative on any license personally. I, I don't like when people start as open source and they give something away as like Apache and then all of a sudden, Nope, we're gonna restrict everybody five years later.

[00:27:34] Right. Um, I think that leaves a bad taste in a lot of contributors mouths. I'm curious your take on that, where you see this again, I'm happy with any license somebody wants to go with, but where do you draw the line?

[00:27:47] JJ (OSS Capital): Yeah. So our, our answer to this is unfortunately very nuanced. Um, but at the same time, it's actually.

[00:27:52] Avi Press: Good.

[00:27:53] JJ (OSS Capital): So, so the, the top level thing is going back to our thesis, which is kind of like the, the billboard slogan version of [00:28:00] like , slogan, version of , what, uh, constitutes like our whole, mission and purpose, Um, which is, we think, we think software's eating the world.

[00:28:07] We think open source eating software much faster than softwares in the world, as those two curves I mentioned earlier. Uh, in order, in order for that to actually make sense and mean anything, open source has to have, uh, a very definitional concrete, uh, meaning behind it itself. And you, you, you mentioned something which is, you know, maybe there's the open source definition starting to evolve and change a little bit.

[00:28:30] I actually strongly disagree with that, and I don't believe that that's actually happening, um, for a lot of kind of very concrete reasons, but primary. you know, open source itself, uh, is built on these four, uh, four kind of rights frameworks, primitives, if you will, otherwise known as freedoms, which that, that word freedom is, is, is a bit, you know, subjective in a lot of ways.

[00:28:54] And, and I think, you know, uh, uh, we had to sort of change the original construct of [00:29:00] what open source was from free software, uh, built on this word freedom, which, which has very loose meaning in different kind of linguistic contexts and different language contexts to this thing called open source. Um, uh, which encapsulated all of those four freedoms.

[00:29:16] But, you know, sort of rebranded and distanced itself, itself from the creator's. Uh, highly ideological motivations, which is by the creator, I mean, Richard Stallman and Richard Stallman in the eighties, who invented the, the free software framework and the G canoe, um, you know, tool chain as well as the, uh, the, the, the free software, um, license licenses themselves.

[00:29:36] So, um, Uh, we believe open source means something very specific, meaning it, it's a fundamentally, uh, rights, uh, driven, uh, discrimination free framework applied to source code, computer source code. It, it basically says without discrimination, uh, to any entity or human, um, you, you have the ability to see the source code, modify the source code, distribute the source code, and,[00:30:00] basically commercialize the source code with, with or without the changes to, to, to the, to the original, um, uh, source code.

[00:30:06] And there's, there's open source licenses that fall within those core tenants with some degrees of kind of permutation in terms of how trademarks are treated or in terms of how attribution, attribution, uh, is treated and in terms of how reference rights are treated in, in patents and, and so on. But fundamentally, those four rights are encapsulated.

[00:30:24] Um, there is a new branch of sort of. Uh, an in between fully proprietary software licensing and open source licensing that is called Source available. And a lot of people call this sort of pseudo open source or maybe open source definition is stretching. It's actually not. This is a new class of licenses, which are sort of an in between.

[00:30:46] And my partner, Heather, the mes actually been the sort of the, the, the primary attorney and licensing expert in the world of the, the, the legal domain to have contributed to designing a lot of these licenses. You mentioned one of 'em called the bsl, which was written for Maria db, uh, a decade or so ago, [00:31:00] uh, which is a temporal timebound license that says in the beginning it's source available, meaning only one company can monetize, only one company can control and, and, and contribute to that code base.

[00:31:09] Um, but over a certain period of time, after maybe a certain release cadence, what have you, then the code base flips over to being, um, at that point in time. Uh, uh, open source, meaning it could be Apache or mit, that's actually a very nuanced license. Uh, it's not open source. It's a time bound, initially source available.

[00:31:25] Eventually open source, uh, license. Um, what, what we see, which kind of goes against the narrative based, based on the data, which is like, you know, observational, like looking at data that's not really our opinion, um, is we have a database of about 500 or so of, uh, the most widely recognized and sort of understood to be, um, up and coming open source companies, commercial open source companies.

[00:31:48] And something like on the order of less than 8% of those companies use a source available, uh, license at the core of the technology, which would actually go and say they're actually not technically open source. They're [00:32:00] sort of, you know, proprietary software companies that have a pseudo open source license at the core source available.

[00:32:05] What we still classify them as open source because they operate very much in that, in that kind of framework. Um, so, and that increase has definitely been, you know, notable in the last kind of few years. There's been a lot of new, new startups that have just chosen, like from the beginning instead of Apache or mit.

[00:32:19] I'm gonna, I'm gonna use one of these sorts available licenses. So that's definitely a decision point, but it is so. Fractionally small as to really not be justifying this sort of narrative of like, there's this huge movement towards like all the open source companies moving towards using source available.

[00:32:36] It's actually a tiny, tiny, tiny fraction. The vast majority still choose a very permissive license from the beginning. Now there's a flip, flip side, uh, discussion to this, which is like, after a certain point in time and years in the future, how many of them will want to do what Elastic or Mongo did, which is change the license in the future after they've reached a certain amount of adoption, critical mass, you know, community, community, uh, you know, uh, [00:33:00] value.

[00:33:00] And, um, you know, that that's really kind of an unanswerable question so far because, uh, frankly, the category is so young. We haven't, we haven't seen the category grow enough. By category, I mean the type of companies that this represents. Commercial open source companies, which there's like literally less, There's around a thousand of them maybe in the world.

[00:33:20] We have like extensive data on like 500 of them. Um, There, there, there's not enough data to be able to say, Are we going to see a full, you know, full, full scale transition from very permissive open source licensing to source available? My senses were not. In fact, I would propose, and I think this is a, this is probably the, the, the clearest I've gotten on this pretty recently, so think, thank you for giving me the opportunity to kind of express this.

[00:33:43] I think that the, the, the mechanism that is a source available license will actually, in fact serve to be a transition point into more open source, not the other way around,

[00:33:53] Matt Yonkovit: it'll 

[00:33:53] JJ (OSS Capital): lot of 

[00:33:54] Matt Yonkovit: as SSPL or source available and then move to open source later?

[00:33:58] JJ (OSS Capital): I actually think that that's more [00:34:00] likely what will happen because the, the sort of, the size of, of the space of proprietary software that exists as part of all of the, you know, trillions of dollars of value captured is much larger than the size of the space of open. That, uh, constitutes commercial open source companies.

[00:34:18] And so if in fact we're relatively correct about the open source eating software faster than software Z in the world, you know, two areas of the curve, uh, it follows that source available could actually be a way to go and say, for existing proprietary software company incumbents, look, we're actually gonna maybe gradually take some of our product and source available license it kind of like how Unity does with Unity game engine.

[00:34:41] And then over time they're gonna be able to say like, Wow, there's, there's so much stuff here that's kind of non-business differentiating. Like it's not gonna really matter too much. Like if we actually just like fully open source this, like we're still gonna be able to generate the same revenue, we're still gonna be able to build the same products and services.

[00:34:55] People are gonna pay us for those products. So they're not paying for the source code, they're paying for the product experience. They're paying [00:35:00] for the differentiation they get from, they get from that experience. And I think over time what we're gonna see as source available is used as a mechanism to expand the amount of open source as a further transition from that and not.

[00:35:14] Shrink the universe of open source more towards proprietary, which is kind of the broader narrative, which I think is kind of the exact perfectly inversely wrong way of looking at it. I think over time what will happen is a subsequent step from that is the industry will go and say, Wow, maybe source code itself actually doesn't have any value.

[00:35:30] Maybe we can just kind of open source really everything that we're doing, and in fact just build valuable products and services and charge money for those products and services, and people will continue paying money for those products and services. I would argue for the vast majority of consumer software, for example, you could open source almost all of it, and the consumers would continue paying money for the products and services that are generated by those, uh, pieces of source code, right?

[00:35:53] Like the Zoom client, the the Riverside client, like all the application software we use, consumers do not have the level of sophistication to [00:36:00] care. Or to, you know, cause disruptive, you know, phenomena for those businesses. Um, they're just gonna continue paying for the products and services. They don't have the ability to go and like marshal the source code and like prepare a binary and build it for their different client experiences.

[00:36:14] They're just gonna continue paying for the software and people pay for the software because of convenience and a d differentiated experience. They do not pay for software for access to the source code. So my, my belief actually long, long term, which is like a, you know, much more kind of controversial thing maybe to say is that I, I think the biggest lie actually ever told, which continues to be actually believed by the vast majority of the industry in software and digital technology markets, is that source code has any value at all to begin.

[00:36:41] In fact, I think it does not have any value at all. And so as a consequence, if you open sourced pretty much almost anything, um, I don't think that markets would like collapse and like fail cease to exist. What I do think instead that would happen, what would be, we would have a far greater pace of innovation, a, a much bigger magnifying [00:37:00] glass put on the, you know, actual sources of innovation where things are occurring and a far, far higher, uh, sort of accountability standard in terms of, you know, how, uh, how those innovations kind of.

[00:37:12] Matt Yonkovit: Isn't some of this though, driven because of those fear? I mean, it's the fear of the cloud, right? It's that article that came out a few years ago on the strip mining. Everybody was like, Oh my gosh, you know, the cloud providers, they're taking all of the open source, eh. And so that really generated this movement to how do we exclude those cloud providers from taking our software and then running it on their own.

[00:37:32] But it also led to a lot of open source companies not trying to necessarily commercialize their open source as much as their, as a service products. Um, and sometimes using the open source just as a, as a gateway. Um, what would you say to a company that's looking at the big cloud providers and going like, I don't want them to steal my stuff.

[00:37:51] How do I prevent them from stealing my stuff..

[00:37:54] JJ (OSS Capital): Yeah, so a few comments there. So, Salil de Bondi is the guy who wrote that article, and he's someone I know and he's categorically wrong, both in [00:38:00] terms of his framework for thinking and his conclusion chain that led to writing the article. And he's also wrong in terms of the actual arguments that he's constructed to write that article.

[00:38:09] Uh, thirdly, um, actually believe that the subsequent, you know, sort of as you characterized it as a quote unquote movement that occurred, uh, is actually not a movement. It's a false successive conclusions from it previously. False logic chain that, motivated that, that piece of, uh, you know, characterizing cloud providers at strip mining, the business opportunities behind these open source.

[00:38:31] Um, you know, I, I, I think I'll kind of unpack that very briefly before going to answer your broader question. the reason it is fundamentally flawed and cognitively, you know, sort of laughable to say that the cloud providers are capturing all of the value or too much of the value in open source, uh, is as follows.

[00:38:49] Um, you know, you have open source projects that have served as the basis for creating basically every major internet company and SaaS [00:39:00] product that exists under the sun since the rise of the internet. You know, in the, in the eighties and nineties, uh, in particularly the early two thousands to today, which was about, you know, over the last 20 years, um, the value captured in aggregate by all of those companies measures in the tens of trillions of dollars in the tens of trillions of dollars.

[00:39:16] Um, cloud providers overall are maybe a trillion dollar market today. If you, if you look at the sum total of aws. Azure and Google, uh, and maybe a trillion or so. The value captured by technology industry, as reflective of market cap, um, has already far surpassed, um, any of the value that you could go and say the cloud providers are capturing.

[00:39:37] And by the way, that's been done without any pushback, without any vitriol, without any similar strip lining type posts. And, and there's one simple reason for why that is. Um, because it's been indirect, it hasn't been direct. And what I mean by that is when you take an open source project and you include it in the dependency graph of your product when you're building a [00:40:00] product, and in that dependency graph is a critical dependency and fundamental to being a system of record or a core library that's doing rendering or core library that's doing.

[00:40:10] Um, you know, fundamental that you didn't have to write from scratch. By the way, that could be the compiler, that could be the database, that could be the application framework, that could be the tracing library. It doesn't matter. Um, an application software vendor exposes the product to the consumer, and the customer pays for a set of abstractions that have been compiled and stitched together by that application vendor.

[00:40:30] Could be internet company like Twitter, could be a software company like Salesforce, and they do not care, uh, about those core libraries that are made up, uh, of that product. What they care about is the product. They're gonna pay money for the product, right? And so at the same time, um, what's happened is the open source creators behind those libraries have had enormous benefit.

[00:40:51] and rising of a tide and validation and contribution and, and battle testing of their code. Because it's [00:41:00] powering the entire internet, it's powering the entire SaaS infrastructure marketplace. That, that, that we see today is a, you know, today, a trillion dollar market, you know, last October is it $2 trillion markets?

[00:41:11] So presses have come down by about half, but it's still a trillion dollar category. Internet application, software and consumer services is another three or 4 trillion. And then you have to kind of count up all the major internet companies, is another four, four or 5 trillion. So in total, it's. North of 10 trillion in, in total value capture across the entire digital economy of products and services.

[00:41:30] 90% of that is made up of open source, 90% of it, probably more. Um, and so we haven't seen any pushback to that phenomenon. And so that's where I would say the argument is just very deeply flawed, which is, you know, the cloud providers are capturing too much value. We must discriminate against them and change the license such that these poor helpless, you know, ignorant amus, open source maintainers can somehow capture some, some, some value.

[00:41:58] I think that is equal parts [00:42:00] condescending, phenomenally poorly constructive and flaw as an argument. And it actually strongly encourage you to watch a nearly two hour keynote conversation that I actually did with Si de Bondi, which I, I feel very few people have actually taken the time to watch cause this is a very extensive discussion, um, in a respectful way.

[00:42:17] And I, cuz I do believe SI is very smart and. Intelligent person, but his cognitive logic chain is so deeply flawed. It's, you know, quite, quite laughable in my mind. Um, I would encourage people to watch that. Going on from there I think that, you know, what we have seen as, as a, a massive, um, sort of validation point for this way of building companies is there is no need to discriminate, uh, against commercialization in order for your business to be hugely successful.

[00:42:46] What you, what you have to do though is, is embrace the pace of innovation that is forced upon you by a much larger surface area of collaboration and competition. And when that happens, all it, all it does is it, it [00:43:00] benefits everybody. It benefits the consumer, it benefits your competitive ecosystem. It benefits society, it benefits the technology cuz the technology is pushed in a much more, uh, productive direction than it would otherwise.

[00:43:11] Um, it, it really, it really benefits everybody. It does make, um, capitalism and, and, and sort of value capture, you know, frankly, a lot harder. But this is just going back to the fundamentals and basics of, of capitalism. Like, you know, pace of innovation is the only thing that actually matters. The moment you start to slow your pace of innovation, you become vulnerable.

[00:43:32] Maybe you've built this monopoly or this oligopoly and, and big profit profit stream, but at that point you're gonna get disrupted by someone else, or you're gonna have to disrupt yourself. The only thing that open source does is it accelerates that by a couple of orders of magnitude, and you have to kind of be comfortable with that, with that phenomena.

[00:43:49] Matt Yonkovit: Yeah, that's a great point. I agree completely that it is kind of a weird spot and I don't believe that it is the strip mining effect, I've, Avi and I have been talking about that [00:44:00] for months and months now on, various channels. So yeah, we completely agree.

[00:44:04] Avi Press: But I think like overall, the, the idea that value capture is more challenging in this space, I think just, it just is, um, you know, d due to all of these, um, the way that we are shipping code to each other and then all of these values. And so I'm curious with your portfolio companies, are you seeing, um, are you seeing particular challenges that are unique to open source companies that seem to be common across these?

[00:44:31] So I'm curious, what are these challenges and are, how are you helping companies through them? How are you seeing companies work through them? Uh, yeah. Talk about that a little bit.

[00:44:41] JJ (OSS Capital): Yeah, I posted this maybe a couple months ago, and I'm not sure if you, you caught it Avi, but it's, it's really good, uh, you know, reinforcing pitch frankly for scarf and what you're building. Um, uh, I don't even, I'm not sure if I actually called out scarf specifically in that post, but what I find to be the most common thread of, of, um, sort of [00:45:00] structural difficulty for all emerging commercial open source company founders, um, is that, uh, from inception, these projects are built, distributed, maintained, and iterated on, um, via a, a centralized, closed social network owned by Microsoft called And, um, you know, on, we, we have certain data and certain artifacts of information that give us a, a reasonable sense of sort of the, the, the, the metadata of activity and sort of, um, energy that exists around these projects.

[00:45:38] But we don't have any, anywhere near the amount of information and insights that you would get, uh, say if you were frankly building a fully proprietary widget. And for anyone to use that widget, they have to sign up on your domain with an email and sign up for a freemium account and every click interaction, um, Context switch, uh, uh, you [00:46:00] know, a data mutation, uh, that they do within that piece of software that you wrote, you have forensic visibility into, right?

[00:46:06] And so there's a big spectrum. I, I think of everything in terms of continuums. Nothing really is binary in the world except, you know, the open source definition and maybe death and taxes, um, uh, and maybe a few other things. Um, but, you know, you have fully proprietary sa products like Riverside thing we're using today, or Zoom that, uh, you don't have forensic telemetry and detail on every single click stream bit by packet of information going through that software.

[00:46:30] And, you know, frankly, you're, you're able to have, you know, far more knowledge and information about, um, how to improve that product experience. Then you then you would otherwise, with open source, it's the complete other end of the spectrum where you build this piece of software, you know, you distribute the software somehow, whether it's on your own GI repo or on mostly, or on Docker hub, and the software gets downloaded and then you're not able to actually understand how it's used.

[00:46:55] Not able to understand, you know, where the user is running it, uh, in terms of the system configuration or the [00:47:00] operating system. You're not able to see how people are operating and interacting within the software, which functions they're invoking, you know, how often they're invoking them and so on. And with that, with that causes is this fundamental disconnect and gap between the product creator and the product user feedback.

[00:47:17] So as a result, um, the, the, the most common structural set of challenges is this kind of like, um, degree to which you can shorten that gap, a as, as close as possible in order to get closer to your user, closer to the source of feedback and insight that would allow you to improve your product in different ways, in different contexts.

[00:47:38] And, um, and, and frankly, just get a lot more insight and information from the software, uh, user, um, and in order to build something more tailored to them in, in whatever domain it could be a customer, it could be. Some vertical specific thing could be particular type of engineering domain, whatever.

[00:47:53] Um, and so what I, what I find as the, the number one challenge is sort of navigating this, You know, we've [00:48:00] launched the project on GitHub. We've gotten some trending activity, we've gotten a community, It's really exciting. Now how do we go and connect all the missing data from that environment to basically domains and assets that we control, like on our own web properties or on our, on our own distribution mechanisms, and get closer to the customer and collect more information that would be helpful, helpful to them without, without becoming too tempted towards going and saying Open source is so great.

[00:48:28] It's given us so much enormous leverage and EF. , but we really hate the fact that we don't have all the telemetry and information on how people are using our software. And in fact, not just how, but who is using our software and the identity characteristics of all of those people and how many people are they, or how many instances are they and how many machines are they and how many API calls are they?

[00:48:47] And so on. So there's this like giant, um, space of complexity between taking an open source project and building a lot of activation, energy and momentum around it. And then at the same time, [00:49:00] building a similar set of momentum and activation energy and data and engagement with potential end user customer that will maybe over time want to pay you money for your product experience.

[00:49:12] Uh, but it's actually, it's a perfect incentive structure aligning. They're paying you money. They want you to actually have a lot of insight into how to make that experience better. But yet, if they're just paying you money for the open source through a support contract implicitly, there's, there's no incentive system that allows you to connect to their.

[00:49:30] Uh, you know, to connect to their environment and get data from them. So this is why you have a lot of open source companies going and saying, Let's just build a file service from day one. Personally, I actually think that that's a flawed conclusion in all cases, because you could have some situations where there's gonna be a huge market for you to build the enterprise product version of that and sell a subscription on premise or on the colo facility or in some vpc and there's a better version.

[00:49:53] There's an enterprise or a pro version or whatever of your open source that you could ship just as a binary. Maybe it's source available, maybe it's totally [00:50:00] proprietary. That's also totally fine. And like you can charge money for it and then you can get, uh, a lot of telemetry and data from your customers that way.

[00:50:07] Um, so it's a very, a lot of this is case specific. It's very, very complex. Um, you know, there are answers to a lot of these things that it just tends to be very bespoke and custom and how these, these sort of problems are addressed. And, you know, we have 20 odd major investments in companies that we're very close to that are go going through, uh, these types of problems.

[00:50:26] In different ways and shapes and forms and, and they're sort of solving them in different ways. And I, and I think that the next kind of iteration of this, this, um, category is gonna be where people start to standardize and sort of tools and services like scarf emerge that, that contribute to kind of making these problems like less painful and less complex, but as well, like just kind of better, you know, patterns and practices and observations in the, in the industry that, that become better, better understood, and more widely, um, appreciated.

[00:50:52] And then, so maybe some, some more common patterns and solutions, um, emerge as well.

[00:50:57] Matt Yonkovit: Yeah, it, it's interesting because you mentioned [00:51:00] like the, the enterprise versions and there's a lot of companies who still don't want to invest in the as a service, especially for their infrastructure components. When they've got other infrastructure components, they're building their own private clouds internally.

[00:51:12] They want to implement that and that telemetry data, um, is so important for the end users, but it. Almost a, a taboo thing to ask for in some open source, you know, communities, I mean, Avi, you know this because, you've experienced

[00:51:28] Avi Press: Very well . Yeah. Yeah, yeah.

[00:51:31] Matt Yonkovit: Um, but yeah.

[00:51:33] Avi Press: and I think, um, that's why. Opensourcing, these exact initiatives as well is going to be so important. And, uh, we'll tease, look for some stuff from scarf in this exact vein, uh, coming up soon that I'm very excited about. But can't, I won't say more yet. Um, but yeah, I mean the reason all the stuff is just so important is because, you know, when we talk about, more, although be it small comp, small number of companies using source [00:52:00] available kinds of license, we just need to do as much as we can to best set up open source companies for success, um, systemically in, in many kind of different ways.

[00:52:10] And I think if we do that, we will see a lot more success with these open source businesses and ultimately just a lot better software for everybody, for sure. Yeah. Yeah. Um, yeah. Matt, did you have any other questions to,

[00:52:24] Matt Yonkovit: let, let me, let me ask you this final question because this is interesting. This gets back to not only these growth metrics, but it also talks a little bit about, maybe one of the things that I've seen, um, as a challenge, right? Adoption of open source in the open source space does not necessarily equate to commercial success.

[00:52:42] Getting the right commercial playbook in place is critical. Uh, just because you have millions of downloads or thousands and thousands of contributors doesn't mean that you have the capability to monetize that without having the right systems in place. And I think that's really a critical [00:53:00] thing.

[00:53:00] You know, JJ, you had mentioned earlier how you're starting to look at those, you know, contributors as an early indicator, which is great. I was curious, what advice do you give to people who have those yearly contributors to turn that into a commercial success?

[00:53:16] JJ (OSS Capital): I think a lot of it is not rocket science and it's not some new magical playbook. Um, by the way, I think generally the idea of commercial go to market playbooks, The world of commercial open source companies is, um, extremely dangerous, uh, for really one main reason, which is that as, uh, as a company goes from an open source project to a product or a set of product experiences, which are, you know, things that you charge, uh, for, and you have customers paying you money for in, in order to use them, as opposed to open source where there's this completely permissionless dynamic where you know everyone's doing anything at any time and you have no information.

[00:53:56] Um, you, you, you often go through a highly [00:54:00] evolutionary and iterative way of, of experimenting with monetization strategies. And I know that sounds like. Complex ball of wax, but like, it, it, it just, it just is what it is. Every very successful open source business that I've studied, uh, commercial open source company that I've studied over the last, you know, several years, went through many different ways of iterating on how to, um, you know, attract and acquire their first customers through charging, uh, for plugins, you know, services, uh, enterprise version, some version of a cloud service or a managed api.

[00:54:33] You know, it could have been, you know, something more specific like what GitLab does, like this buyer based open core approach where you have like three or four different ways of monetizing based on user personas inside of an organization. Um, you know, it could be just like, here's the a la carte enterprise or pro or premium version of each open source tool.

[00:54:50] And we go and basically identify, uh, has, she corp. Did an investor day recently and is really excellent articulating this for just public market investors today, which is [00:55:00] they have this like practitioner end user or audience. They build and nurture continuously forever. It's not one, one time static thing.

[00:55:07] And then that feeds into accessing a buyer customer persona, which is a different per buyer per persona, and different constituency than the, per the persona of a practitioner or an end user. And the buyer is often a decision maker in an enterprise, or it's a CTO or CIO or a VP of IT, or some person with, you know, budget and or access to a budget inside of an organization.

[00:55:31] And they have different sensibilities, different ways of, you know, making technology purchase decisions, different ways of understanding the, uh, the value proposition of pro of products and technology and the impact that those technologies have on their organizations. And so I think that, um, you know, it's very incremental and iterative.

[00:55:48] Um, each one of these companies is, is quite unique in terms of how they monetize. And it, it takes a lot of humil humility to kind of understand, like, you know, exactly as you said, some projects have massive amounts of adoption. [00:56:00] And they're really not able to, to, uh, to be, to be monetized or, or, or commercialized in different ways.

[00:56:04] Um, there's one kind of structural thing that I would say to kind of wrap, wrap that up, which is, um, you know, there's, there's actually hundreds of millions of, of open source projects out there. Like, uh, on, on, on GitHub, maybe there's now 50 million or so that are like actual projects. But last I checked, there's a few hundred million, you know, Rebos on GitHub, um, alone.

[00:56:22] Uh, but let's just call, you know, 50 million or so, or are more than just, you know, a single in IT file or like a readme, like they actually have some code that does something. Um, we're talking about tens of millions of, you know, quote unquote ideas out there. Um, the things to, to us that, that, that could constitute the basis for a large potential business over time fall into one or two or both categories, which is that the software touches, uh, reads or writes data, meaning it's a system of record at the sort of data persistence tier.

[00:56:54] Um, Or it's a system of intermediation and it [00:57:00] proxies or, um, uh, intermediates, uh, data at the network. Persistence tier, meaning it's sort of like, uh, a proxy or, uh, a message bus, um, or, you know, some, some kind of middleware, uh, uh, uh, abstraction of some kind. Um, but it's still, uh, interfacing with or touching critical data of some kind.

[00:57:21] If the project is not in either one of those two types of areas, and it's in the much more ephemeral domain where it's not really touching mission critical data, but it's sort of like a library or an ancillary thing. Um, the likelihood and the propensity for that project to be the, the sort of core basis for a large business is very, very low.

[00:57:39] Um, it's not impossible, but it's very, very low. Whereas if the project instead is reading or writing data in some, in some kind of way, shape, or form, meaning it's application or it's a database or something, or if it's intermediating or, or, or proxy data, At the network tier, uh, or both. Uh, it has a higher probability or a higher propensity potentially for [00:58:00] being the basis for a potentially very large and valuable commercial open source company.

[00:58:04] That is a technical distinction. Both of those abstractions I mentioned are completely orthogonal to vertical, to layer in the stack to any of these things. Those are very like technical kind of properties, and so we pay very close attention to those. Um, every company that I've like looked at that we've invested in either maps to those, one or both, and every large and successful open source business in, in the public markets today are the ones that are growing really fast in the private markets that are, you know, well onto a hundred million plus in revenue and so on.

[00:58:32] Also map to that, um, it's just really important to be, you know, controlling and managing, um, the way you intermediate and, and or generate and or, uh, persistent store data. Um, I think the final, final, final thing that I would say, which is. I think something I'm thinking a lot about is this world of, of neural networks and transformer models and attention networks and, you know, everything coming out of stable diffusion.

[00:58:57] And, and you know, frankly, the, the, [00:59:00] the oracle Microsoft that is open ai, which is not really open ai. Um, and, you know, frankly the world of neural networks is getting so sophisticated, so fast that we will very likely see software engineering disciplines become automated to the extent that, um, software developers.

[00:59:17] Start to resemble more, more, more, more like artists and and creators in the graphic design world than they are, um, you know, actually, uh, sort of translating natural language instructions from product managers down to, you know, computer source code and writing source code themselves. Um, I think the, the engineers and architects of the future are just gonna be really, really good at translating abstract, kind of high level concepts of what some software needs to do in terms of language.

[00:59:44] The computer understands using natural English, not c plus plus, or Python, or Java or JavaScript code, uh, or, or rust code if you're a, if you're a AVI or high, um,

[00:59:55] Avi Press: Yep.

[00:59:56] JJ (OSS Capital): programming. Um, there's semantics and ideas and those languages that are really [01:00:00] valuable, but to express them in natural English is actually very counterintuitive and difficult.

[01:00:03] That is the domain that that engineers will. For a very long time, cuz this is difficult to communicate the, the, the, the, the abstraction boundaries. So I think that the, I think, I think a lot about the, the, the machine learning and the neural networking stuff and that's probably where a lot of the, the software development practices are gonna be headed in the future.

[01:00:20] Matt Yonkovit: Awesome. Awesome. Well, uh, jj, I want to thank you for hanging out with us. Um, you know, uh, it's been a pleasure chatting with you about this and we really do appreciate you hanging out with us.

[01:00:31] JJ (OSS Capital): My pleasure. Thank you so much for having me guys.

[01:00:33] Matt Yonkovit: Uh, All right.

[01:00:34] Avi Press: thanks JJ..