MLOps Zoomcamp: Free MLOps course. Register here!

DataTalks.Club

Articles Slack Events Podcast Books Courses

Investing in Open-Source Data Tools

Season 15, episode 2 of the DataTalks.Club podcast with Bela Wiertz

Links:

Did you like this episode? Check other episodes of the podcast, and register for new events.

Transcript

Alexey: This week we'll talk about investing, and investing specifically in open source tools, and in data and ML tools. We have a special guest today, Bela. Bela is working for a German family office that invests in VC funds and early-stage startups with a focus on open source data, AI, and developer tooling. In his work, he's sourcing, evaluating, and working with many different open source companies in the early stage of finding their product community fit. Actually, there is a funny story of how we met, because we have this thing called Open Source spotlight. It was a big pleasure for me to discover that people don't only use this for looking at tools, like developers, but sometimes investors also look at this. So this is how we met, and it's my pleasure to welcome you, finally. It's been a while since I was trying to invite you to this podcast. Welcome to our podcast! (1:15)

Bela: Yeah, happy to be here! It's a pleasure. (2:17)

Alexey: Great. The questions for today's interview were prepared by Elena Samuylova. Thanks, Elena, for your help. Elena has already been a guest with us multiple times. She works at an open source startup. Thanks to her, we have amazing questions for today's interview. (2:19)

Bela's background

Alexey: Before we go into our main topic of open source and investing in open source, let's start with your background. Can you tell us about your career journey so far? (2:40)

Bela: Yeah, for sure. I think it's maybe a little atypical for your show, but I have a classical business background. Right after my studies, I started working in the startup and tech ecosystem. Right after my business studies, I worked at a company builder, which was about commercializing under-commercialized open source communities. And right after that, I joined TKM Family Office, where I am today. (2:50)

Bela: Now, since pretty much two years, I'm helping to build up the investment activity. I started there right when the Family Office also started. As you said, correctly, [we're] looking for startups, talking to founders a lot, researching those topics, especially on the open source and developer tooling data side. That's also why I'm happy to be here today. (2:50)

Alexey: Yeah. Indeed, it's not a very typical background [for this podcast], which makes it even more interesting to have you on this show. But I'm wondering, what does it mean to have a business background? Or what do you actually study there? How to run a business or what? (3:46)

Bela: Yeah, it certainly differs from university. University, in general, it is about [ensuring] that you know... the most important stuff for doing what I do today is to understand how a company is structured, to understand how to read the balance sheet, to think about business models, to think about margins, marketing, and all this kind of stuff. So it's a very general type of study, but that's generally it. If you don't really know which way you want to go, it's definitely good to have some kind of orientation. But you can, for sure, specialize more later on into the studies. But I only did my Bachelor's, so I'm quite a very generalist type of guy, until I got into tech and startups and ventures. (4:02)

Alexey: So an MBA is also a business-related thing, but it's more like a Master's level education? (5:03)

Bela: Yes, it's Master level and MBA is also for non-business Bachelor's. If you did, for example, an engineering major in your... I think this is also for business Bachelors and Undergraduates, but mainly, also for non-business undergrads to do an MBA. You kind of go the executive way in your organization. (5:07)

Alexey: Okay. So in your studies, you learn how to run a business, how companies structure, and all that. With this background, you joined a company that was interested in commercializing open source communities. What does that actually mean “commercializing open source communities”? (5:37)

Bela: So it was a very small company builder, run by Philip, who is also our Chief Investment Officer today. The thesis of a company builder in general is, “Okay, we have the thesis and the machinery to build up more and more early-stage startups and spin them out, get founders there,” and so on and so forth. The specific thesis there was that there is a lot of value in these communities. Back then, we have seen, or we thought that there might be a lot of communities that don't have any access to any kind of commercial version of a product. There are a lot of open source products with a lot of traction in the open source world, but there's no one to build some additional features around it, host it, package it, make it a great experience, and sell it. (5:58)

Bela: It turns out that it's really cost-intensive to build a lot of startups. This company builder founded one startup, which is called crowd.def. I think some of you might have heard about it in the open source world. It's about community management and this kind of stuff. Jonathan, who runs it today as the CEO and founder, was also the second founder of the company builder. That was actually also the whole team of the company builder. Then, Jonathan went out and crowd full-time, Philip got the offer to start the Family Office, and he asked me to join. I took the opportunity and that's how we went there. So it was quite a short journey with the company builder. But really, it was my first touchpoint with open source. I actually haven't even really heard about it before. I just had a business background and I had not been in tech, really. (5:58)

Alexey: Interesting. Did I understand correctly that the [business] model for this company builder was to I look at [sites like] GitHub, find promising open source projects, and then reach out to the people that create them, asking, “Hey, do you want to start a company for this project?” (8:09)

Bela: I have to say that there is no blueprint on that because there was never a project that got so far. Crowd.dev was completely internally developed and is open source today. But this was kind of a thought and a strategy, but then the ways parted, and now we're doing what we're doing. But yeah, that's an idea that we had. (8:26)

Alexey: Okay. But right now you thought, “Okay, maybe instead of finding founders and all these people who can start a company, let's focus on companies that already exist and give them money.” Right? (8:52)

Bela: It wasn't really my decision, as you said. Philip got the offer, he asked me to join, and it's a great opportunity. Yeah, it is definitely easier with a smaller team to invest than to build your own companies. Definitely. Phillip and I are the only really full-time persons in the investment team. It wouldn't be so easy to build up [something like] one company per month with this team. But that's the story with that. (9:08)

Why startups even need investors

Alexey: Why are investors needed at all? Can't founders just start companies without your help? Why are you needed? (9:44)

Bela: Yeah. Venture capital, in specific. I think Investors are not needed for any type of business. I think you do not need to think that. But venture capital, and tech in specific, I would say, provide the money for you to build up the tech startups from the ground up. When you start, as a first time founder and not having that much money, you need to have some money to hire your first employees, to get your first setups, to do whatever you want to do. And to be able to do that and to not need to focus on profitability right away, you need some kind of investor to help you because they enable you to focus on growth and building instead of only selling and making a margin from the early days. (9:55)

Bela: They also make it possible for founders to pay themselves a salary from the beginning, if you're thinking about bootstrapping something. This can be definitely a great thing because, I have to say, not every startup – not every tech business even – is a venture case. But it definitely is something where you can have a salary, where you can make a living from your startup from day one – at least from day one when you raised [funds]. That's why I would say they're needed. (9:55)

Alexey: So investors give founders money to run the company (to build a startup) and the alternative, as you mentioned, is bootstrapping. This is when the founders do not look for external investment (into external money), but they try to earn money themselves. They focus on profitability, and from this money, they pay themselves a salary and hire their first employees. But this way is not always easy, I guess, especially in open source. In open source, you usually do not make money directly from the open source [tool] from day one. (11:31)

Bela: Yeah, it's slower and it also defocuses you from growing and building because you always have to think about, “Oh, how can I finance the next month?” and so on. But yeah, those are the two main ways. I mean, there are some people who already made some money, who can finance their own venture, but then, they're their own investors, I would say. Therefore, this also goes in the category of having an investor. (12:06)

Alexey: This means that, let's say I work at a usual full-time job, I set aside some money, and then use this money to just run a company. Then I don't need an investor. (12:36)

Bela: Yeah, you have something for a start. But it really depends on what kind of business you want to build. It really differs. If you have a bootstrap business in profitability from day one – maybe it's a little bit slower, but you definitely know what you're doing there. With venture, you also really have to know that it is a little bit different. If you take money from an investor the first time, you definitely need to have it a second, and a third time. You have to decide which way you want to go, especially in the data and open source ecosystem right now, I would say there are so many... You really have to see where you can find your spot in terms of competition and positioning. It's hard to compete without having any money in your pockets. (12:46)

Why open source is a viable go-to-market strategy

Alexey: It's not one of the questions we planned to ask you, but maybe you have some ideas about the following. Why do companies actually do business in open source? It's a very difficult business model, right? You open all the code from the very beginning and everyone can just go there, take the code and not pay you anything. Yet, there are startups that do that and there are investors who invest in them, saying, “Okay, work in the open – keep all the code open – and here's money to do that.” Why does such a thing even exist? (13:42)

Bela: Yeah. I mean, there are different ways that those companies get started, right? There are some that are just a hobby project from some engineers. This is why they open source it, because it was a hobby project, and that's the way you build normally – in public and so on. So that's one part. And once you're open source, and you also have a certain size community, it's hard to take it back, so then you're in it. But there are also a lot of companies, especially startups, that plan to go the open source route because it's a way of distribution and building up a moat against competition in the long run. (14:16)

Bela: It is not always the case that, even if a lot of people wouldn't consider that real open source, a lot of them don't have all of their code open sourced. They have proprietary features around it, they have proprietary versions, a lot of the customers they address wouldn't make the hustle to self-host it. That's the way. A lot of them see it as distribution channels, building a competitive moat through their community, selling through their community, having this kind of potential for early traction without revenue. (14:16)

Alexey: As a developer myself, I would add that I kind of have more trust in companies – in the code [if it's open source]. (15:36)

Bela: Yeah, that too. Good point. (15:42)

Alexey: I just take a package from GitHub, or do something like pip install the package name. Then I have it, I can experiment with this, and if something's not clear, I could just go to GitHub and look at the code. This way, I see that, “Okay, that makes sense.” The entry level is nothing – I can just install it and then play with it. Then I have trust in this thing. (15:45)

Bela: Yeah. This, essentially, is a part of this kind of distribution because if you want to go this community-driven, go-to-market, especially with a focus on developers, it is exactly how you said it is. They have more trust in open source. You can build a bottom-up community very much easier when you build it in the open, as opposed to when it's a closed free version. (16:12)

Building a bottom-up community

Alexey: And what does a “bottom-up community” actually mean? (16:40)

Bela: It's from the distribution side. You could go to market bottom-up, meaning you start with the smaller people using your product, like the developers. And on the other side would be the potential to make typical “top-down enterprise sales,” as you would call it. So it would start at the decision makers and making very, very big deals there. Community building the bottom-up way that we see also see a lot in open source, a lot of them also do top-down sales later on. But especially in the beginning, we see a lot of this, “Okay, how can you find the individual people using your product?” And you can find developers in the open source spaces. (16:45)

Alexey: As a developer, I do not necessarily have money to buy a tool [Bela agrees] but because it doesn't cost anything, I can just download and run it. Then I can come to my manager and say, “Look, there is this awesome tool. What about self-costing it?” And then my manager would say, “Yeah, I will talk to my manager about that.” The manager of my manager actually has the money and has the budget and then they say “Okay, it looks like a useful tool. Here's the money.” (17:32)

Bela: Yeah, it's a little bit of the evolution of the product-led sales that [companies like] Slack and Notion did. Some people used it, some people used it, and then you come in like, “Oh, now I have to pay.” But now 50 people are already using it as their main note-taking tool – now we really have to pay for it. It's the evolution in the developer space for the same kind of playbook. (18:01)

The investment thesis for the TKM Family Office and the blurriness of the funding round naming convention

Alexey: When preparing for this podcast episode, I learned a new word –“thesis”. Well, I did know this word, but in a different context. For me, it was something like a Master's thesis – the thing I wrote at university. But here, you already mentioned the word thesis for your company builder. The thesis was that you find founders and let them run a company. Then there is also a thing called an “investment thesis”. I guess this is for investors – what they invest in (which companies). So what I wanted to ask is, what is your investment thesis? Where do you invest, in which companies, at which stage? (18:33)

Bela: Yeah. We definitely, as I think already mentioned, do early stage. We consider [companies] as early as possible – not really later than the Seed round. In venture language, you have the first round, which is sometimes called an Angel round, sometimes a Pre-seed round – it depends. You can always frame it however you want. Sometimes you also reframe it in the future. [cross-talk] (19:14)

Alexey: Before we go to other rounds, what does it actually mean to have an “angel round” or “pre-seed round”? Why “angel”? (19:47)

Bela: Yeah, it's very blurry. You can actually call everything anything, I think. I've seen a lot of different words for a lot of the same stuff. But I think essentially, “angels” are non-institutional investors. What institutional investors are, we will, I think, come to that later. There are venture capital funds, also corporate venture capital investors. (19:53)

Alexey: Basically a company with money. (20:24)

Bela: Yeah, yeah. Angel rounds are pretty much only for individual investors, which include business angels and also family offices. Pre-seed rounds could be in the same size, but there are dedicated pre-seed venture capital funds as well. They also sometimes invest in Angel rounds, so the wording is not quite clear, I think. (20:26)

Alexey: When you already have an idea as a founder, or a potential founder, and then maybe you have some code, maybe you already have some GitHub repo. You see, “Okay, this is promising. Here's a bunch of money.” That's the Angel round or Pre-seed round? (20:51)

Bela: Actually, you can do it however you want. It also sometimes depends on the size. We have seen the French guys for Mistral raising over 100 million in a seed round as their first round, which some people would call a pre-seed round even. And 100 million is sometimes [called] a Series C – so there really are no rules for that. But I would definitely say that it's better to start with the smallest name possible. Always consider raising a Pre-seed instead of a Seed if you've not raised before, because it gives you an opportunity for the Seed round to also raise without revenue and so on. You will not really be able to raise a Series A without certain metrics, but you would with a Seed sometimes. But the names are really blurry, and you can call anything everything. There are no rules, I would say. But you would normally start with a Pre-seed then go to Seed, and then the Series A, B, C, D, and so on come after. This is the normal way of how it progresses. (21:07)

Alexey: Typically, a company, or a founder, says, “Hey, this is what I have. Give me some money.” They put in an open call, or maybe they reach out to different investors, and then the investors think, “Okay, this is promising. Here's some money.” Right? And they do this multiple times. (22:20)

Bela: Yes. In the best case, as a founder, you should think about what kind of money you need for the next 12 to 18 months, for example. It depends. Say I want to hire two-three developers – I want to hire whatever. I need this amount of money for that. I want to do some marketing experiments. I need some money for that. In total, I need x, y, z type of money. And then I go out with this. Because all the investors are gonna ask you, “What do you need that money for?” And if you don't have a good answer to that, that's not a good sign – if you don't know what you're raising for. Then the investors have their own processes. Normally, if you have a venture fund, with the money from one fund, you typically invest into 20--40 startups. It always depends on the round you're investing in. But in the early stage, I would say, most of them invest in like 20 to 30 to 40 different companies. (22:36)

Alexey: Okay. So you focus on Angel rounds, Pre-seed rounds, and Seed rounds – the early stage, when companies do not have much in terms of money (in terms of investment) yet. And you focus on open source. Right? (23:42)

Bela: Yes. Our general focus is everything that is B2B software, but our niche is developer stack (developer tooling) and a lot of this is open source because open source is a great go-to-market model for those kinds of companies. But not everything we do is open source and not everything we do is even developer tooling. Most of it is. As you said, we're doing Pre-seed/Seed rounds, mainly in Europe. We would also do [other places] but mainly in Europe. We really like, as I said in this niche, founders and companies with a strong focus on community-driven go-to-markets, as for example, open source. This includes... We did stuff in AI infrastructure, in cybersecurity and cloud, in data streaming. But we also did, to mention here – maybe it’s not interesting for everyone watching this – but we also did investments in other areas of b2b software, like intralogistics, data privacy, robotics, and so on and so forth. The main focus, I would say, 80% of what we do nowadays, is in this niche. (23:59)

Angel investors vs VC Funds vs family offices

Alexey: We already talked about Angels – they are individuals who have some money and they invest. Maybe not a lot of money, but they have some money that they say, “Okay, this startup's promising. Here's a small investment.” There are also VC funds, which are larger organizations with more money. They typically invest (give more money) to companies. And there are also family funds (family offices). This is where you work, right? What is the difference? Or what does it actually mean to have a family office? What is a family office? And what's the difference between Angels and VC funds? (25:19)

Bela: Yeah. Maybe I'll start with the [VC] funds, because that's the most prominent, I would say, type of investor that you also see in the news and so on. In the end, I would say they are like money managers. They raise capital from outside investors, like big pension funds. In the US, they also [get capital] from university funds, government programs in Europe and Germany. Also, there are individual corporate investors – individual investors could be Angels, or it could be family offices. Corporate investors could be classic corporations, big companies, who have a financial interest in working with their money. So they raise funds from these kinds of people and organizations. They have a certain amount of time, and they try to maximize what they can give back to them. That's the model of a VC fund. (26:00)

Bela: Then you have the ones who work with their own money, which are business Angels and family offices, for example. I would also say that there, the line is a little bit blurry. I think historically, family offices are more connected to businesses. But nowadays, family office also just means that you, as an Angel, invest a little bit more and a little bit more professionally. An Angel and a family office doesn't really have to be a lot different. There are, for sure, also family offices that are quite closely connected to an operational business because they then also invest with not only financial interest, but also they want to have some overlap with the operation of the business. But we, for example, don't have any operational business, we are pretty much a VC fund, but only with our own money that we have in the company. I don't know if this answers your question. (26:00)

Alexey: More or less, but maybe let me try to summarize what I understood. Business Angels, or Angel investors, are people with their own money, and they manage this money themselves. For example, I have a bunch of money, I see a promising startup that is trying to raising funds (they need some money) and I say, “Okay, here's $10,000 (or whatever).” Then they respond with, “Oh, thank you. With this money, we can pay our developer for one month (or whatever).” [chuckles] So these are Angel investors. From what I know, they typically don't give a ton of money to startups. (28:05)

Alexey: Then there are, from what I understood, family offices. Here, the difference between a business Angel and a family office is that the person with the money hires somebody to make these investments. They don't do these calculations themselves or think about, “Should I invest there or not?” Instead, they hire somebody like you, for example, who does that. Right? That's the main difference? (28:05)

Bela: Yeah. As I said, there are also some Angels who work with employees. It also may be different in terms of how you market it, because all the Angels are also working like entities for their investments. Most of the time, they're not investing as a private person. But yeah, that's pretty much it. It's a little bit blurry. I would say family offices are a little bit more of a professional setup. But family offices also, in the past, did a lot of other things than venture capitals. We are focusing on venture capital – a lot of family offices are also more like general money managers – doing a lot of real estate, doing public equities, doing private equity, in more senior companies. It is quite blurry. I would say that in general, family offices make a little bit larger checks. And you really have to differentiate between family offices who have an operational business behind it and industry specifics, and the ones who operate like a venture capital investor, like we do. (29:16)

Alexey: I was going to ask you about these checks. You mentioned that family offices typically give larger checks. I know a few people who do angel Investing, and from what they told me, the investments they make are not large – it can be 10,000, it can be 20,000. But for a startup, just this investment alone does not make much difference. But when they have like 10-20 Angel investors, and each of them gives 10-20K, then they have some money for some time. Meanwhile, a family office can give a larger check than that because they have more money, I guess, than an individual who can invest in a startup. Right? (30:28)

Bela: Yeah. I would say that business Angels are also typically more focused on early stage [startups] and there are also a lot of family offices that do later rounds. So the check sizes that family offices write are, I think, pretty much uncapped. There are ones doing serious, A Series, B Series investments – there are ones like us who are doing early stage investments – but, yeah, that's pretty much it. There are also angels who write 100K checks, so it really differs. It's not hard to really... I'm sorry to say this to a lot of questions today, but there's no rule for that. (31:13)

Alexey: Every case is different. Some people don't have money to invest, but some people do. (31:59)

Bela: Yeah. In the end, you can call yourself whatever you want. (32:05)

Alexey: Yeah, as long as you have money. [chuckles] Right? (32:11)

Bela: Yeah, as long as you do good work and do your stuff, people won't really care if you're calling yourself an Angel investor or if you call it a family office – whatever. I think it doesn't really make a big difference. (32:14)

Bela's investment criteria and GitHub stars as a metric

Alexey: When you invest in open source companies, what do you actually look at? What are the criteria you use to make a decision whether you want to give them money or not, and how much you want to give them? (32:31)

Bela: How much is a question that is more related to how much they're raising, and our typical ticket sizes. If you have a fund – if you want to invest in 20-30 (whatever number of) companies, then you need a typical ticket size that you write, or at least the typical amount of shares that you want to own after you invest. It's not like I'm a little bit more committed to this company, so I invest more in this and I think that they're cool, but not so cool, so I invest a little bit less. I think that it doesn't work this way. You really have to be fully committed in every startup that you invest in. In the end, there is some amount of finance in there, so you really have to make the calculations whether this works, so there's this amount. That's where the amount of money comes from. (32:46)

Bela: But in terms of which company you invest in, I would say it also differs a lot from case to case. For sure, in the open source world, for example, there's no minimum amount of stars that you need. We also see a lot of investments that are made pre-launch. They call themselves an open source, because they want to be open source – they want to see it as their go-to-market – but they're not even launched. There's no metrics on the open source side. So I would say that stars are a great discovery metric – for example, for me to discover projects – because it's hard to discover something if it's just a private GitHub repo. That's quite hard. But you can really tell that a startup that has twice as many GitHub stars can raise twice as much money. It really has to stay with, “Okay, the team must have an understanding of the value of the community. They must at least have a plan on how to grow it, and how to capture the created value in the future.” (32:46)

Bela: I think this is a very important part for investors to understand – if they think that this team knows how to close this kind of gap that you sometimes have in open source between value creation and value capturing. So I would say that there is no magic recipe. As an investor, you definitely have to go deep into this kind of thing like, “How do they think about their commercialization?” And this kind of stuff. “Do they have an idea about it? Or do you yourself have an idea about it?” You don't really have to have everything figured out, but you have to think about whether you can make money out of this product. Besides that, I would say that the most important criteria are the same as a non-open source company. You look at the team, you look at the need in the market, how big the problem is, and so on, and so forth. (32:46)

Bela: Especially in Pre-seed and Seed rounds, the team is a very big factor, because the product can also change a lot, market conditions can change a lot. Also, for the founders, I really have to say that picking the right investors from a human kind of way is very important, because you're working together for 10 years, minimum (most of the time). So that is really important. I would say, those commercialization ideas, and so on, are a little bit different for open source. You have to think about them as an investor. But most criteria are the same as in non-open source startups. (32:46)

Alexey: So you look at the founders and the team, then you look at product market fit – whether there is a need in the market and how large it is. Right? (36:27)

Bela: There is no product market fit, most of the time, if you're a Pre-seed or Seed. You have a product market fit when you have a lot of traction, and people are just ripping your product out of your hands. But you have to understand if there's a potential... Is there really a problem? Are you solving a problem or are you just building anything to build a solution? And you are just looking for a problem somewhere [that may not exist]? (36:37)

Alexey: Okay. So it's not the product market fit yet? It's more like, “Okay, the problem does exist and this is a solution to it. Potentially, in the future, there will be product market fit.” (37:02)

Bela: Yeah. And “That's the team that I trust to solve that problem.” I think this is one of the most important parts. (37:13)

Alexey: Do you have an interview with them where you ask these questions and they reply? Or do they do a presentation? How does it usually work? (37:19)

Bela: It really depends on how professional you are, but we talk with the founders multiple times before we invest. We have several calls with them, do interviews with potential and also current users, and talk with other investors about it. I mean, there are certainly some Angels who just see a pitch deck and put money on it. But we certainly talk to the founders, 2, 3, 4, 5 (depends) times. It's definitely a very important factor to get to know them – to get to know how they think. (37:26)

Alexey: Okay, this is the kind of feed between you and founders, then you also look at the team, you look at the problem size and then when it specifically comes to open source, you look if the team has an understanding of how the community is going to develop and how large it is. You look at how they are going to capture the value they create. It isn't just the product that people will use, that there is potential for making money out of that product. (38:10)

Bela: Yeah. What you said first, it's more about this, “Okay, what is my idea for growing my community? Do I have a plan on technical marketing? Have I tried out different ways of growing the community? Do I have a plan on how to do it?” But yeah, what you said is correct. (38:40)

Alexey: We did talk a bit about GitHub stars. You said that this is not the most important metric, but you still look at this sometimes because it gives you an idea of how discoverable [the product] is. At the end, you do look, but, maybe the important part for you is not the amount of stars, but if there are any stars at all. Is that right? (39:01)

Bela: No, what I said was more about if there are no stars, and it's a private repo, I cannot find it, so I cannot invest at all. It's more like stars are a great discovery metric for me to discover projects myself. But you really also have to set this in perspective with the project they're working on. For example, if you're working on something in the generative AI space within the last months, you would have a good chance to have a lot of stars of people just being curious and putting a star there as a bookmark – and that's it. If you have a project and have 5000 stars but no active people, it's not worth anything. (39:28)

Bela: If you have 100 stars, but 20 people actively engaging on your Discord every day, I would say that this could be more valuable as an early stage company. So yeah, the stars are a metric – you have to set them in perspective. It can show how people adopt your product, but it cannot show everything. It's often a kind of vanity metric as well. (39:28)

Alexey: But having a good number of stars is actually a good thing. I know that when you have many stars in one day, then you get into the GitHub's Trending page. This is how your project can be found by other people. Is this how you discover projects yourself? Do you go to GitHub Trending and look at the projects there? (40:41)

Bela: The problem there is that those projects are also most of the time too late for us as an investment. That's another thing – if they have 5000 stars, for example, and more, and they have a great team, and they have a great idea of how to commercialize it, and it's a big problem – most of the time, they already have an investment, because it is quite transparent. It is great to discover the projects that are still smaller, but have an engaging community – that have this spike in stars after you invested. It's better than before you invested. Right? It also depends on the case and who you're building for but, in general, stars are good and bad. You can track it, you can look at it, but it shouldn't be like, “Oh, I invest in only above 2000 stars and with this kind of star growth rate.” (41:03)

Alexey: So if a project is on the Trending page, it's most likely already too late for you to invest in this project. How do you actually find projects to invest in? (42:18)

Inbound sourcing, outbound sourcing, and investor networking

Bela: Networking is a large part of venture capital in general – knowing people who know people who know people. But for us, it's also more and more about looking and screening GitHub – I have my search queries where I look for startups. A lot of VCs also do data-driven sourcing where they have automated... I think it's clear what data-driven sourcing means. Then it's also inbound, right? People who are getting in contact with you, who are texting you, “Hey, I'm building this and that.” I think inbound/outbound are the other two most important [things]. (42:26)

Bela: And then you have a network, which is also kind of an inbound channel, but more qualified and a little bit less [broad] – the people know you, they know what you're looking at. Those, I think, are most important. For sure, you look at [things like] GitHub, for example, Product Hunt, different other newsletters that send out tools that are in beta. You're also at conferences, you're talking with accelerators – a lot of different ways. There's also no recipe there. I would say that there is a lot of stuff coming from your network, but you cannot rely on that, so a lot of people build up other ways to find cool stuff in the market. (42:26)

Alexey: So how much time do you spend on discovering projects and products? (44:09)

Bela: I really have to say that this has a little bit to do with what kind of things I have on my desk, whether I work closely with some of our portfolio companies, or working on some internal projects, or I'm really digging deep into one or two startups that we're looking at right now. Then I don't have that much time, but I try to allocate a minimum of two hours in the morning, normally, to sourcing. We have quite a narrow scope with this niche, so it's quite easy to have a good overview there. For sure, it's not everything, especially the pre-launch stuff that is made by people who are just thinking about leaving their big tech company and know some VC investors from their lunch break and then just make the deal before it's public. Therefore, you need a network for those people. But for outbound sourcing, I try to allocate some hours every morning to stay on top of that. (44:18)

Alexey: So outbound is when you find something interesting, and then you reach out to [the people who run the project]. (45:30)

Bela: I have to reach out, yeah because they don't know me. (45:33)

Alexey: And inbound is when somebody you don't know someone, but they found you somehow. For example, by watching this interview, they say, “Okay, I want to write to this person and pitch my product/my idea.” (45:37)

Bela: Yeah, or, for example, if I know someone who's getting in touch with someone else – investors are not only investing alone. We also co-invest with a lot of bigger funds and smaller funds and Angels. So if they see something cool that we would fit with, they also say, “Hey, have you seen this? Do you want me to connect you to them?” Those are the ways that we get in contact with cool stuff. (45:55)

Making a good impression on an investor

Alexey: Speaking of inbound leads, or whatever the right term is – when somebody contacts you, how can they make a good impression? How can they pitch their project in such a way that you would actually decide to talk to them? (46:23)

Bela: I mean, the bare minimum would be to look at what we are doing. If you're building something in this vertical, especially in the developer stack – B2B software in general, but especially the developer stack – I'm happy to have any kind of request. You can just say, “Hey, I saw you there. I just discovered you. I heard from you about whatever. I'm building in this and that space. Do you want to talk? Here's my calendar.” Whatever. It doesn't need to be this huge amount [of text] because we only look at [things] in our niche. There are not 200 people a day outreaching to me. Or if they don't even look at what I'm interested in, for example, I would say it's a no-go. (46:42)

Bela: So if you're looking at some investor who is also an Angel, please look at what kind of deals they're normally doing and if you fit that and not just spamming all people you find on LinkedIn that have “investor” somewhere in their profile. Look at them and maybe also write one line about why you two could match together, “I saw that you focus on this and that, and I saw that you invest in XYZ. That fits perfectly because I'm building whatever.” Then I'm always happy to talk with everyone because it's always great to have an exchange even if it doesn't lead to an investment in the end. But yes, I'm happy for everyone who's building in this space to message me. (46:42)

Alexey: LinkedIn, right? (48:19)

Bela: It doesn't have to be a [giant] paragraph like this. It's good to have a short wrap-up and maybe a link to the project or the website or deck or whatever. Yeah, LinkedIn is the main way. I mean, you can theoretically contact me on Twitter, but I'm not really active there. I think you can also find me via GitHub – there's my mail there. But I think LinkedIn is the easiest way. (48:20)

Alexey: That's how we got in touch, I think, through LinkedIn. (48:53)

Bela: Yeah. I think LinkedIn. Because I think I messaged you because of the Open Source Spotlight, as you mentioned before, right? A lot of companies where we already invested in are also featured there. It's a great way to present your project to the community. That's amazing. (48:56)

Alexey: That was a very nice, unexpected surprise because I did not think it would be used in such a way. It was very pleasant to know that it's not just developers who look at Open Source Spotlight. (49:15)

Balancing open and closed source parts of a product

Alexey: We have a question which is quite related to what we just talked about. The question is, “We are building a conversational data tool. Some parts will be open source, while some other parts will remain closed source. Would you consider this model?” (49:28)

Bela: Yeah, definitely. This open-core, I would call it, startup model is something that is really familiar. A lot of people work with this. For me, like I'm not the religious open source guy, I have to say. For me, it's not about that you have to have everything open source MIT license and don't make money out of it. It's great that you have an idea of, “I put this and that in open source, and it helps me with this and that.” So if you have reasoning behind it, and if you also have reasoning behind doing something not open source – that's amazing. If you're building something totally proprietary, and you have very good reasoning for it, that's also fine. If you think about what you're doing and if you have good reasoning behind it, you can do open-core, you can have full open source and only sell enterprise licenses – you can do whatever you want, if you reason it in a good way, I would say. Definitely, I think commercial open source and this kind of open core model, where you have proprietary features, and a hosted version of that, also I would say the most is the prominent model that I see with newer and younger open source startups. (49:46)

Alexey: We have open core, which is a model where there are some features that are closed, and they offer something like a “cloud version” of the tool – a hosted version. You don't have to host this thing yourself. What are the other models of monetizing/commercializing an open source project in addition to open core? (51:09)

Bela: Yeah. One of the most classic ways would be to just sell enterprise licenses, right? There's also “on-premise” licenses. It depends on what license you have to determine whether you are able to really host it yourself. But the hosted versions are really the number one [way in terms of popularity], I would say. There, you have the difference between the ones who have proprietary features versus the ones who just have a better user experience and do not have the hassle with the data, and so on and so forth. You don't need some DevOps – you don't need anyone to configure it. This would be just hosted open source. But then you really have to look at the license because you don't want any big cloud providers to just host it and make money out of what you build. (51:30)

Bela: You should think about the license twice. Selling enterprise on-premise licenses is possible. In general, on-prem license, also with proprietary features – you can mix these different licenses, different features – you can mix and match that as you want. You really have to think about who your users are, what problem you are solving for them, and what kind of business model or what kind of service they want. If you're really in a really regulated industry, selling to fintechs or financial industry players, handling a lot of data, maybe it's better for you to work with a proprietary enterprise license or an enterprise license, if your open source version with additional features or without or with support. This is not really the most liked model for investors, because it's not that scalable. There are a lot of different ways – you just have to find the best way for your customers and your financial interest. (51:30)

Alexey: So when it comes to these enterprise licenses, I guess the idea behind it is that I, as an individual, can use this software for free, but if I want to use it for a commercial activity – for activity that generates some money, let's say at work – then I have to pay. (53:34)

Bela: Yeah. There are models without additional features for these licenses, with them, without – whatever. But yeah, that's the model, I would say. (53:57)

Alexey: And the last one you mentioned – with support – is let's say I have... (54:06)

Bela: I would say [that's] the oldest model, right? I haven't been in the industry that long, but I would say that's the old school model – supporting your open source model, which is running on-prem on your users' instances. (54:10)

Alexey: So the idea is that – there is an open source project, but sometimes you need help to run this project... (54:26)

Bela: Right. To configure it, build up the infrastructure for it and so on. (54:37)

Alexey: Then you talk to the project team and they help you. (54:42)

Bela: So as I said, this is maybe great for bootstrapping, but I would say this is not really liked by investors because it's not scalable. You always have to think about... if you want to go the investment route, you have to figure out an investment model that is scalable – one where you don't need one more employee for more revenue, but you can scale the revenue without scaling up your employees at some point for the same amount. This kind of leverage in the long term – I think a lot of investors want to see that. (54:47)

Alexey: Okay. So if you want to bootstrap, then fine. But if you want to look for investments, you should probably think of other ways of commercializing your products, right? (55:18)

Bela: Yeah, yeah. (55:30)

The future of open source

Alexey: Another question. “How do you think open source will look in the future? Will it become more popular or less popular? What's your opinion on that?” (55:33)

Bela: I think we are just at the start of... I also have the feeling that there are more and more and more open source players, and also some doing the same [thing] but if you look at the proprietary startup ecosystem, you see 10–20 companies doing exactly the same boring stuff. I'm sorry to say it that way. [chuckles] We are not there – not near there. There are only a few niche applications where open source is the status quo and I think there is a possibility that we can get there, where open source players will be the dominant players in a lot of those markets – not only the challengers, but the ones who are dominating [these spaces]. (55:44)

Bela: I think we have to work on a lot of things to get there. But you've seen where we are today with databases – with a lot of open source databases. I mean, in the end, they're hosted a lot of times by the big cloud providers. But I think there is a chance that open source will go further, and we will see some big market leaders there. I don't really think that we have them yet. I look at a lot of open source, so I also have the feeling that there is already a lot. But if you then look into other areas, you will see that it's still small. (55:44)

Recent successes of open source companies

Alexey: Do you know any more or less recent success stories of companies doing open source? Of course, there is Linux, Red Hat – companies like that, which have been around for a very long time. I know MySQL, Oracle, and Java have been around for years. I also know, personally, Elasticsearch, even despite the difficulties with AWS – they had a problem with AWS hosting Elasticsearch and not paying anything. It's still, I think, kind of a success story because they are a successful startup – they do make money. But it's also a relatively old company. Do you know any recent successful open source companies? (57:26)

Bela: I would say that for [companies like Elasticsearch] it can only be said if they're successful after a few more years. Even ones like Mongo, like Elastic, like HashiCorp – right now they're all on the public markets, they're all stock-listed, and they're still fighting with the capital markets about their profitability. So what is their success? It's always hard to tell. But for sure, there are some promising companies that closed larger funding rounds in the recent past. So far you see this as a success – it is at least a sign of success. There are ones such as, for example, Superbase, which is building a platform as a service. Hugging Face, for sure. Kong, the API management platform. Apollo, the GraphQL platform. I think they've raised big rounds. (58:16)

Bela: I cannot tell if they're really going to be multi-billion dollar companies, but we see signs that there are a couple who could be the next Mongo, the next Elastic, the next HashiCorp. There are also some small projects in Germany, where we are both located, which are a little earlier, but still very promising. You see Gitpod, which has quite good traction. In the vector space, you see Qdrant, you see Deepset – they are not the biggest projects yet, but [a part of] a great wave of open source companies from Germany and from Europe. It's not the amount of funding that companies like Superbase have, but we see the different stages of those companies that are getting filled. (58:16)

Alexey: There is a comment from Alexandro saying that, “Neo4j is a good example. They have a good open source project with an enterprise package.” (1:00:10)

Bela: Yeah. (1:00:17)

Bela's resource recommendations

Alexey: Last question for you – we should be wrapping up. The question is, “Do you know any good book or other resource that you can recommend to the listeners or somebody who wants to learn more about investing and about things that we talked about today?” (1:00:18)

Bela: Okay. It depends. On investing, I would say, in the community-building kind of way... in this open source community building, there's an article (I can later send you some documents about it) which is from Crowd.dev, actually, about their top six books on developed community building. I think this is really important for early stage founders that are building open source to understand. This is not the investment part. There's also a great article from Bessemer Ventures – it's a big VC fund. It's already from 2020, but it's about open source metrics in investing. I think this could be interesting. Books about investing in general? I can maybe send you something afterwards. I would say that despite everything that he is doing personally, Zero to One from Peter Thiel is a good book to start with startups. (1:00:35)

Alexey: Peter Thiel is an investor? (1:01:48)

Bela: He founded PayPal, I think. And he's one of the first Investors in Facebook and has the Founders Fund and he's an investor. He's quite, I would say, a difficult person in terms of his political thoughts. But despite that, the book Zero to One is recommendable. (1:01:52)

Alexey: Okay. On that note, I want to thank you for investing your time (one hour) to talk to us and share everything you learned, or at least everything I asked you. Thanks for that. And thanks, everyone, for joining us today and asking your questions. (1:02:19)

Bela: Yeah, it was a pleasure. Thank you for having me. (1:02:37)

Alexey: Yeah. Well, it was a pleasure talking to you. Have a great rest of your day and your week. And thanks, everyone, for joining us today. I guess we'll speak with you, Bela – we will meet in Berlin. See you next week for the next interview. (1:02:41)

Subscribe to our weekly newsletter and join our Slack.
We'll keep you informed about our events, articles, courses, and everything else happening in the Club.

DataTalks.Club. Hosted on GitHub Pages. We use cookies.