Episode 92
· 30:20
I thought that I would learn almost like a formula in a sense of here's how you successfully build an open source company, but it it actually has turned out to be very different than that. There's so many different flavors of open source companies. It's kind of a fragile way to build a company, and you have to be really kind of thoughtful of how you're working with your community because once you put something out in open source, it's free. It kind of one of the unspoken rules is that it's gonna stay there and it's gonna stay free.
Jack Bridger:Yeah. Hi, everyone. You're listening to scaling DevTools. I'm joined today by Robbie, who is, well, firstly, I'm a big fan of Robbie's podcast, Open Source Startup Podcast, and Robbie has been an investor in DevTools infrastructure for 8 years.
Robby:Yeah. Long time.
Jack Bridger:Yeah. So thanks so much for joining.
Robby:Thanks so much for having me on, and huge shout out to my cohost of the Open Source Startup Podcast, Tim Chen, who I admire so much. And I probably would never have lasted three and a half years doing a podcast without him. So he's the best.
Jack Bridger:Yeah. Just my small Twitter interactions with Tim
Robby:can attest
Jack Bridger:that he seems like an absolute legend. Yeah. But, so, firstly, as I said, big fan of the show. And what I wanted to ask you is about, like, what kind of things have you learned about dev tools and about growing dev tools from the show? Because you've done over a 100 episodes now.
Robby:Yeah. We've done about a 150 episodes now, which is pretty crazy because time has just kinda gone by as we've continued to do these. But maybe go back to the beginning on why we started the podcast to begin with because I think it informs a lot of how, at least, I've thought about dev tools and how to grow a dev tool company, especially using open source. Tim and I started the podcast initially because him and I were spending a lot of time digging into go to market strategies for dev tool companies, and we wrote this blog post a number of years ago that interviewed the founders of HashiCorp, LaunchDarkly, LightStep, about some of their early go to market thinking and strategies and how they got their first enterprise customer. And through the course of those interviews, we learned a ton and then we consolidated all of our learnings into a blog post.
Robby:And then we realized after that the actual conversations we had, like the recorded ones, were the most valuable part of that work. And we had wished that we had taken those recordings and posted them. And so that's what got us starting to think about a podcast. And then when we were looking at what was on the market, there were a few really awesome DevTool focused podcasts, but not a lot that were specific to open source and how to build an open source company, not just a project. So, so we kinda started it as, like a little beta project, and we didn't really know what we were doing.
Robby:I was doing the editing myself early on. So if you listen to the earliest episodes, they're quite painful. But, we've graduated to an awesome editor who's way more professional now who works with us. And, the funny thing is, at least for me, doing the podcast, I thought that I would learn almost like a a formula in a sense of here's how you successfully build an open source company. Mhmm.
Robby:And there kinda be, like, a set of pretty much typical steps and things that founders go through. But it it actually has turned out to be very different than that. There's so many different flavors of open source companies and open source business models, and they change over time, and they change with the community that you're serving. And so I actually have, I've kind of changed my thinking, and now I have just, like, felt like every situation is so different depending on what the user community wants, depending on what the opportunity set is. And the last thing I'll say, we were talking about this earlier, but one of the things that, I have completely changed my tune on when I when we first started doing the podcast, in my mind, the most successful way that you would generally build an open source company is having a project's creators go out and start a company around it.
Robby:And there's so many examples now of new open source standards that come out, like something like OpenTelemetry, where a number of other companies are created around it, and a number of those companies are not tied at all to the project's creation. They've just seen its rise, and it's more of a new standard that a lot of folks are using. So it's just one example of how there's just many different flavors of open source companies, and everything is very kinda specific to the situation, the point in time, and the user base.
Jack Bridger:Yeah. That's, that's really good learning. And I know that when when you were talking about that, one of the things that episodes that reminded me of is the Century episode with David Kramer and how he was talking about, like, how it was just a Django library initially and that they were working on it for a long time before it became a company.
Robby:Yeah. Yeah. It's a so David was one of he says that as one of my favorite episodes actually because he is such a down to earth and honest person despite kinda how successful and big the company's gotten. Because usually, we we spoke about this earlier, when a company is young and founders are still figuring things out, they'll be very kinda open, honest, unfiltered about what's going on. And then as they get media training and get more senior, then it's very, very polished and they're amazing speakers, but you don't generally get as much authenticity in a way.
Robby:And David was just super open and honest about everything that went on, but it's such a good example of, the project was around for a long time. Like, he did not try to turn that into a company for a number of years, and, one of the things that we've seen over the last, like, 5, 6 years, but particularly in 2020 and 2021 when there was a lot of capital available, investors pushing project creators to create companies and giving them a ton of money really early, which then means those companies have to go and try and monetize in order to meet their valuations. And a lot of these projects just take so long to really resonate with the community. And so trying to force a business model too quickly and trying to, you know, grow a community maybe in a way that's not super organic, it can really hurt the opportunity. So it's it's kind of a fragile way to build a company, and you have to be really kind of thoughtful of how you're working with your community when you try to focus on monetization, and kind of how you're gonna set that up because once you put something out in open source, it's free.
Robby:It kind of one of the unspoken rules is that it's gonna stay there and it's gonna stay free. Yeah. We've had a few episodes talking about situations that have gone wrong where something that was freed up and sourced then went to paid, But it's kinda the open source model of building a company, in a way, can be at odds with venture timelines and scaling. Yeah. So it's it's quite tricky.
Robby:And, and I think David, but a number of other examples, didn't raise, like, right out the gate and took some time to really nurture and grow their community before really identifying the company potential behind it.
Jack Bridger:Yeah. When you're talking to founders, if they want to build like, I guess people must come to you and say, like, I want to build an open source startup. Yeah. What kind of stuff do you say to them?
Robby:So it's really funny. I I hear this a lot where founders will say, I wanna use open source, and, and they kinda look at it as especially if they don't have a lot of experience with open source. They look at it as this tool that they can use, kind of like a go to market strategy. Yeah. They're like, oh, yeah.
Robby:It's it's freemium. It's but it's not at all the same. Yeah. An open source community is very different. It's essentially a a stand alone product.
Robby:And, oftentimes, depending on how it's constructed, it can be almost a competitor in a way to what you're building. So it's really and and there's also, like, a community that may be helping you build out your projects that might be adding functionality, adding to your documentation, answering questions with one another, and where the community is hosted, whether like a Slack or Discord. So it's a very kind of sensitive thing that encompasses much more than a go to market strategy. And so I when I talk to founders that want to use open source, my first question is why. Yeah.
Robby:Like, why does this why does this help you? And, like, why is it right for this kind of tool and community? Because there are certain types of tools where whoever the user profile is, they want it to be open source because they want the transparency. That's where other tools how other tools are being built. But then I also wanna understand how they're planning on, like, building a community alongside building a product and how much thought has gone into that because it is a a hard road to go down to build an open source company.
Robby:And so doing it and not fully thinking through the the cons, I think folks think through the pros a lot where they're like, oh, yeah. Distribution. Everyone will use it. Everyone's gonna contribute to it. But then the downside of needing to maintain this community and needing to contribute to it, needing to add functionality to it, while at the same time you're trying to make money, especially if you're raising venture capital early on.
Robby:It's it's quite challenging. It's really hard.
Jack Bridger:Yeah. And so is it something that you think do do you work with founders that come at it with, like, I think that we'll grow we'll build a bigger business if we are open source? Mhmm. Or do you kind of, like, focus on people that have already built this, like, kind of open source project, and they're, like, now I want to build a company.
Robby:Oh, that's a good question. I mean, I don't think I'm so formulaic about it to say that if someone wanted to open source part of their solution or thought there was huge benefits to that in the community, but they hadn't done it already, then I wouldn't fund them because I'm also investing super early generally. Like, pre seed, seed, very early. I would say for the most part, the open source companies that I have invested in, they already have a project up and running usually when they're thinking about company potential, but it's not always the case. I have a number of companies that will build products on top of an open source standard, like an open telemetry, and and or they might have an open periphery approach where they open source part of what they're building, but maybe not the core of it.
Robby:So I'm I'm comfortable investing, in kinda all different flavors, but I wanna understand the thought that's gone into it and the user research that's gone into it too. One thing that I I spend a lot more time on than I think other other folks that do really early stage investing is I care a lot about the prework, that you that founders do, understanding their users and customers before they before they build or while they're building. So, usually, even with open source, like, before they've decided that part of their solution or the core of their solution should be open source, I wanna know why and, like, what's informed that and who they've talked to. Because even if they themselves are the user profile, they don't represent all developers or all data engineers or whoever their user profile is. So I wanna know how much thought and work has gone into understanding why parts of their solution should be open source.
Jack Bridger:Mhmm.
Robby:And then if they have a preexisting project, then a lot of that work has kind of naturally been done by virtue of that community existing. But that that work upfront is something that I spent a ton of time, trying to understand with founders.
Jack Bridger:Have you given them advice about, like, where they should go and talk to these people? Because sometimes it's it can be quite tricky, I feel like, if, like, if it doesn't exist yet, do you just go and kind of hang out in other open source projects or, like, related or, like
Robby:Yeah. Yeah. It's a really good question. So a lot of times, it'll be, talking to folks who would be potential users, or if it's going to be like, I I talked to a number of founders who are gonna be focused primarily initially on, say, like, the Next. Js community.
Robby:So it's like, okay. Yeah. How well do they know that community? Yeah. And, like, how much time have they spent in it?
Robby:How much do they know about what that kind of profile engineer wants and, like, what kind of functionality? Where are their challenges? Because even if they don't have something to necessarily show them yet, it's more about understanding, okay, what do they want me to build versus, like, kind of building something, coming out and saying, like, okay, this is this is this is what I'm gonna try and give to you. Yeah. And the other thing that, I've seen happen a a number of times is folks will if they don't kinda spend a good amount of time understanding what a community wants and, say, they have an open source, either the core of their solution or part of their solution, they open source it, it just doesn't get adopted, and they don't really understand why.
Robby:And a lot of it is because it's not necessarily what that kind of group of users needs. Sometimes it can you can get lucky, where you kind of release something and it, like, takes off, and that happens with a number of open source companies where there's a project that just kinda catches heat and they were, like, not necessarily expecting it. And then as they understand who's in that community and they understand the company building potential, then they really dig in. But, but yeah. I wanna know kind of what kind of work and thought has been put into, like, understanding what the what the company potential is.
Jack Bridger:Yeah. That that makes a lot of sense. And I I'm building something myself and, like, I definitely, you know, I it was like, okay. This is a challenge I've got. Put it out there, and then just see.
Jack Bridger:But it's, like, the rationale of, like, why open source and stuff is like yeah. It's it's tricky, I guess.
Robby:Yeah. And actually so this is a thing too where I think it is a totally valid strategy to put different things out in open source and kinda see what see what hits, but the the problem is once you raise venture funding, you're kinda sticking yourself onto a timeline Yeah. Of execution to be able to raise the next round, which generally the the way that, at least, I think about it is every round that you raise, you want to be able to 3 x by the like, that should be your target to 3 x your valuation by the next round. So what that means is whatever you're raising at, you have to achieve usually, like, metrics that will get you 3 x evaluation by the next round. And so that's that it's totally fine if you wanna, like, put a bunch of stuff out there and, like, see what sticks, but that is kind of at odds with having this venture timeline of Mhmm.
Robby:You need to perform in a certain way in order to kinda keep growing. So that's where it's really hard because Yeah. I think some founders should do that more when they're kinda, like, in discovery, but they really shouldn't raise venture funding at that point.
Jack Bridger:Yeah. That makes so much sense. So it's like you can do all the kind of, like, throwing stuff out there, but, like, before you look to raise money.
Robby:Yeah.
Jack Bridger:Yeah. Okay. And so is that, like, maybe that transitions on to quite a well, like, quite a good topic of, like, when should you look to raise money on an open source project?
Robby:Yeah. So I think, first and foremost, there are some projects that will take off, but they don't necessarily have company potential behind them. Like, say they're a and I've seen a number of projects like this where they're super widely used. Like, I saw that there was an open source project that was really cool, really widely adopted. That was an open source version of QuickBooks, which is, like, super useful.
Robby:A lot of engineers really liked it. They could kinda customize it more than QuickBooks. But then when you dug into who the users were, it was a lot of hobbyists, a lot of folks who were not going to pay necessarily for a solution. So when you really looked at that kind of, like, wild growth trajectory and you dug into who is in that community, you really don't see, like, company potential where, like, venture scale company potential, in my mind, is the ability to generate a 100,000,000 ARR in, like, 5 to 10 years. Like, there's potential there.
Robby:And so with a a project that's taking off, but it doesn't necessarily fit in a category where there's enough budget tied to it or the users aren't ones that will ever really be able to pay, then it's really challenging. Now some projects will start with a lot of hobbyists as users and over time move up market, but then there really has to be a vision there. Yeah. And it's not just, oh, look at all this usage. Like, I think again, we saw this a lot in 2020, 2021 where project has, like, thousands of GitHub stars that, like, came up overnight, and, you know, everyone is excited about it, and all these hobbies are using it.
Robby:We're actually seeing this a lot in Gen AI right now because there's a number of projects that are being used for what I call, like, the experimental AI economy.
Jack Bridger:Yeah.
Robby:And they're taking off super quick, and the usage numbers look amazing. Then when you dig into who's using them, it's a lot of hobbyists Yeah. Who are never really gonna pay, or it's other AI startups that may or may not have the ability to pay, but then they're, like, getting funded at a rate that doesn't really match to the company potential there. I think that's where it gets really messy where, like, I think the the growth in a project can kinda, glaze, like, over what's really happening. Yeah.
Robby:There are some projects that aren't, you know, super widely adopted, but the users in that community come from, like, really high quality companies with the potential to pay, and they're in a category that has a lot of budget allocated to it. And so, like, that might have a lot more to it than some others that are just super widely adopted, but none of those users will ever pay.
Jack Bridger:Yeah. So you mentioned that users who are from kind of, like, very legit companies Mhmm. Plus, in a category that's, like, got budget.
Robby:Yeah.
Jack Bridger:So I guess it's, like, infrastructure and, like, these kind of, like, examples, I guess, databases, infrastructure.
Robby:Yeah. Yeah. Yeah. Yeah. And it's but it's not always, it's not always, like, a replacement necessarily, but it's there's budget in the sense that companies will it's automating some activity or it's a better version of a tool where there could be budget allocated.
Robby:So to give you an example, I have a portfolio company that I invested in from Cowboy called Mobile Dev. They have an open source project called Maestro, and it's eliminating a lot of manual work to do with mobile testing. Yeah. So they budget in the sense that it's clear it's clearing up a bunch of time for these engineers, saving them time so they'd be willing to pay for kind of, like, paid products on top that add automations. So it but it's it's real companies using it.
Robby:It's, like, mid market enterprise, like, folks who have access to to capital. It's not just hobbyists. And that's where things can get really, really tough for open source companies is if, you know, they try to come up with a paid module. And then if you really dig into using the community, it's a bunch of folks who have no no access to a real budget. The other thing I'll say on this, it's really hard to understand sometimes who's in your community.
Robby:You can kinda go through in a more manual ish fashion, like your Slack community or Discord and kind of understand that way, but then now there's tools like Common Room, which are great for this, which will give you a sense of what companies are in there, who's really engaged, how are they engaged, so you can almost use that as more of a pipeline than you used to be able to for an open source project.
Jack Bridger:Yeah. That that makes sense. So you can kind of see who's actually using it. And if you're like if if I if I was, like, mobile dev, like, on the day on day 1, like, pitching to you, like, how do how are you thinking is this a want can this achieve, like, 100,000,000 ARR?
Robby:Yeah. So it's funny. They didn't start open source. When I invested in, they weren't an open source company. But what was really compelling is the mobile testing space generally is really large.
Robby:There's a lot of mobile test engineers, a lot of work that's very manual that would provide much better app experiences. If it if there was better test coverage, you could test more cases. You could make sure things wouldn't fail every time you did an update. And also, it would just automate away a good amount of work that mobile engineers don't really wanna do. The founder was at Uber, and he was, like, a super elite mobile engineer there.
Robby:Knew the space really, really well. He was kind of thought to be, like, a leader in understanding kind of, like, next generation testing trends, and, and so the market overall was quite big. They didn't start with an open source strategy though, which was interesting because we at the time, it wasn't super clear that that was necessary. And then over the course of it was probably the 1st year where we really started to see or audit like, to the to the founder, Leland, to his credit, really started to see the potential for there to be this new standard with Maestro around a testing framework for mobile. And that strategy ended up making a ton of sense, and it got it was, I think within the 1st year for, like, 1 quarter of that year, it was the fastest growing open source project, like, period, but with real users.
Robby:Like, users that Yeah. You know, could eventually and we're seeing this now turn into customers of our paid products that sit on top of Maestro. Yeah. But yeah. And and by the way, I'd say that this is more common than not for super early stage founders where there's a slight shift in how they're bringing their products to market or what the product direction is, that's totally fine.
Robby:But when I invest, it's behind the fact that I think the founders have the right process of iterating with the community. Like, they really have, like, those nodes and know what they want, and they have the ability to iterate and move depending on what the community is telling them.
Jack Bridger:Yeah. That makes sense. And how do you, like, assess that? It's like you can just
Robby:A lot of it is just spending time with the founders and understanding their process of, like, how they get user feedback, how they're like, even how coachable they are in conversations I'm having with them is a really good tell sign of how well they kind of, like, input output. And the thing is, like, all the folks that I back, all the founders I back, are product visionaries, but they don't operate in a vacuum. They, like, take in feedback of, like, what kind of community wants. So they have a direction that it kinda shifts go to market strategy. They shift their positioning.
Robby:They shift parts of their product depending on what's gonna work for the community. Another company that I was really lucky to work with, that's done really well, not an open source company, but a security company called Drata. They automate security and compliance.
Jack Bridger:Is it like it's like Vantaa competitor.
Robby:It's a competitor, but my, you know Yeah. Yeah. Yeah. Bias opinion much better. Yeah.
Robby:Yeah. 100¢. 100¢. But, but one of the things work at. Yeah.
Robby:One of the things that was so compelling about that team was how much prework they'd done before they'd even, like, written a line of code. They had I think when I first met with the team, it was, like, a Figma design of what they wanted to build that at different points during the meeting, I was, like, oh, what happens when you click there? They're, like, this is literally a design. There's, like, nothing nothing in existence yet. But they've done so much prework, especially since they were coming from an ed tech background.
Robby:Like, they'd gone through, TalkTo in compliance processes with their previous company, but they hadn't actually, like, built a security product before, and they had done so much really sharp interviews so many really sharp interviews with customers or potential customers to figure out exactly what they want in a solution, and it led them to some really interesting product insights around being continuous. So instead of a company automating their audit once a year, they would know at any point during the year if they were in or out of compliance Yeah. And also building in an auditor portal to make it even easier for their customers to automate and not have to email auditors separately from working with Drata. So there were so many points of insight that they translated, like pain points with other products and pain points within their users to what they were going to build. And that's kinda like the the really deep kinda customer centricity and, like, user informed process I like to see.
Jack Bridger:Yeah. That's that's really interesting. I feel like it also is, like, the same methods that you kind of book these interviews and, like, meet people is, like, the same probably method that you'll use to actually, like, market it as well. And, like, it's like if you kind of shows that you understand where these people hang out and, like
Robby:Yeah. Yeah. It that's actually a really good point. I haven't framed it like that before, but, yeah, it does. It demonstrates your ability to get in front of those users as well.
Robby:Because one of the things that's really challenging for, especially, first time founders that are engineers, they really want to build because that's what they know how to do really well. And it's more a regular than regular to find an engineer founder who really wants to work alongside their community and get that feedback and, you know, like, figure out how to get in front of the right folks and market to them. So I try to figure out if a founder kinda has it in them early on
Jack Bridger:Yeah.
Robby:By seeing how they work, seeing their process. One of the biggest legs to me is when I meet founders at, like, the pre seeder seed stage, and they're like, no. This is just the right way to build it because I know it's right, and I, like, don't really need to get feedback, which there are some cases where that can work, but the problem is that's not really a sustainable strategy because if you operate in a vacuum, then suddenly, like, you're on your own echo chamber and you don't really know, like, what's kinda going on. So I think it's just really important to have, like, big ears when it comes to customer and user feedback.
Jack Bridger:Yeah. Totally. That totally makes sense. So I had one last question, which is, like, why you I mean, you've, like, kind of, like, tied yourself to the mast of, like, open source. Like, this is, like, your so you've made a big, like, commitment that you believe in.
Jack Bridger:So, like, why why do you think that open source is something to I guess, why did you tie yourself to open source? Like, what you must believe that, like, it's gonna be a source of, like, great businesses. Like, how yeah.
Robby:Yeah. It's a good question. I don't think I was as, like, I there wasn't as much thought put towards I'm gonna time myself to open source. It really just came out of curiosity and interest when Tim and I started the podcast because I think both of us were like, Tim had a lot of experience as a founder and engineer building with open source. I'm a little bit more of, like, a self taught builder, and open source is really where I've been able to learn a lot, and I love the community aspect of it.
Robby:Like, I'm a pretty community centric person. So there it more just started out of intriguing curiosity. And, and also as I dug into it and met founders like Arman and Hashicorp, and really could under started to understand the power of building in a community centric way, it just really resonated with, oh, like, this is this makes sense when you're trying to build a new standard in particular, like, something really big to do it in this format where you are allowing the community to participate in it in a way that's just not, you know, available if it's a closed source company or project or product. And so just the power of it, I thought was really, really interesting. And for a number of categories, I think that it is the right way to build the like, kind of, like, the biggest company in the sense that there's, like, such a large potential to build a new standard in a way that you can't do if something's closed source.
Robby:And I I know there's a lot of debates going on about this right now, but I think with a lot of foundation models that are going to be, by definition, foundation tech foundational technology, it's so important for them to be open source in a way because it just creates so much more impact and also the ability for others to participate more, trust what's kind of going into the models. Like, it it just creates kind of a different level of of, like, trust and transparency and community involvement. Mhmm. So that's that's really where, like, I I've gotten more and more infatuated with it over time Yeah. The more I've seen the power of it, but then also appreciated how hard it can be to build a company.
Robby:Yeah. In an open source fashion, so become more kinda skeptical of founders that are a little bit too, like, oh, yeah, and then we're just gonna open source this, and it's gonna make it successful. It's like, no. There's a lot that goes into that. Yeah.
Jack Bridger:That's really cool, and I think it's like I I yeah. It's great to hear that kind of passion and, like, it's, like, kind of a beautiful thing, really, when you think, like, humans are kind of, like, benevolently collaborating. It's like I feel like it's one of the most not to be too dramatic, but, like, one of the coolest things that humanity's ever done. It's just, like, collaborating on software where people that don't know each other across the world are just, like, helping
Robby:Yeah.
Jack Bridger:Build things that everyone uses. It's it's super cool.
Robby:Yeah. I totally agree, and I also think that the concept of building cool new technology and letting folks use it, fork it, like, do what's do with it what they want, and just being super open about, like, here's what I've built, it it's kinda it's like, just like very it makes you very optimistic on technology and how it can bring people together and how it doesn't need to be about making money and greed. It can really just be about sharing innovation. Yeah. So there is there is something that is, not to get too, like, you know, whatever about it, but, like, soulful about it where you don't get that with most software that's closed and very much just there for commercial ends.
Robby:So I think that it's just really it's really powerful.
Jack Bridger:Yeah. Yeah. That's amazing. And so thank you so much for coming, Robbie. I know you're working on something very new.
Jack Bridger:There's something
Robby:coming down
Jack Bridger:in the pipeline.
Robby:It's new. It's done. She's building.
Jack Bridger:Robbie's building. Maybe by the time it comes out, maybe there'll be announcements. I don't know. But definitely follow Robbie on Twitter, and I'm sure that's where they'll hear first about what you're
Robby:Oh, yes. And listen to the officer's sort of podcast. Yeah. To listen to.
Jack Bridger:Recommend the Superbase Paul Coppenstein episode to get started. Started.
Robby:Oh, yes. That's a really good one.
Jack Bridger:Yeah. Awesome. Yeah. Is there anything else that you wanted to share before we
Robby:No. This is fantastic. I really enjoyed it.
Jack Bridger:Amazing. Yeah. Thanks so much, Robbie.
Robby:Thank you.
Listen to Scaling DevTools using one of many popular podcasting apps or directories.
