OSS Capital Pt. 2

These past few years, open-source has been on the fast track to dominating the software market. It might sound contradictory to many people, but open-source is big business. A project’s source code may be free, but the products generated from that source code can certainly be commercialized. If you’re a project developer, therein lies the potential for economic benefit, if only you had the financial backing and business know-how. Heather Meeker, Founding Partner at OSS Capital returns to the show to talk about the different aspects of open-source business, from finding venture capital backing to navigating the intricacies of branding and product positioning. Tune in to this informative episode!

---

Welcome back, everybody. It has turned 2024 here. I put these dates in to try and encourage myself to get these episodes put into the live website as soon as possible, but to give everyone a frame of reference. We’re up to 59 episodes. The reason I mentioned that is with me, I have Heather Meeker who you may have remembered on episode five.

‐‐‐

Welcome back, Heather.

Thank you. I’m glad to be back. 

We were talking before that Heather’s the first repeat guest we’ve had on the show. It was super exciting and interesting to talk to you, which was almost exactly a few years ago. That was still quite COVID-y. I remember it very vividly. I was living way too many hours a day in a little room in the loft in my house while my family was trying to maintain their sanity and I was trying to build FlagSmith.

The world has changed quite a lot since COVID, but it feels like the commercial open source and everything is turning into this fairly large tanker that feels like it has an increasing velocity and momentum. In the context of that and before we talk about the book you’ve written, I was delighted to discover that you had a very non-technical early career, which I’m going to have to ask you about.

I took a zigzag path to get to where I am. In fact, it’s probably fair to say that at most points in my life, I never thought I’d end up where I am now. I started out after college. I was a computer programmer. That was so long ago that we called ourselves computer programmers. I then became a professional musician for a while. Anyone who has tried to do that knows it’s a tough job. You usually have to do some other stuff to make ends meet unless you’re some famous signed artist, of which there are very few.

I did that for a while. I ran a small business, which was a rehearsal studio. I then went to law school. Law was my third career. It was unexpectedly a good career. For me, it was a last resort to be a productive member of society. Yet, I have enjoyed it a lot. I’m on my fourth because I’m working at OSS Capital, which is a VC fund.

I found that photograph of you way back as a drummer. My godson has turned eighteen. He is about to go to college and study drumming. He’s in a Blondie tribute band. Funnily enough, two days after I discovered this fact about you, I saw him do his first paid gig. For those of you who are into Blondie, I didn’t quite realize how all of their songs are super fast.

Those are hard to play. They take a lot of stamina to play. 

He was in bits at the end of that gig. Physically, he was finding it hard to stand.

The disco era was hard on drummers. A lot of it is trying to sound like a drum machine, which is not much fun, but a lot of it is fast and sustained. A band I’m in does a disco song and it is exhausting. I have to play 16th notes for 6 minutes. I can barely do that. My rule for that is we only play the song once at every rehearsal because I can’t stand to play it twice. 

They’ve had a set list of 25 songs. They’re only 3 minutes each but they’re all 150 to 160 beats a minute.

I’m sorry if I’m going afield. I love to talk about drumming. You have to turn your mind off because the minute you start thinking about what you’re doing, you will mess up. You have to go in total automatic mode. That’s what I love about drumming. When you’re doing it, it’s a non-intellectual thing to do. I’m not saying you don’t have to be smart to do it. It’s that you have to turn off the conscious part of your mind and feel with your body and listen with your ears. That’s a wonderful experience, particularly if you’re one of these overthinking people like me. I really like the contrast between that and most other things I do.

Give your brain a rest. I had to bring that up because I thought it was super interesting, another side point. Do you want to set the scene a little bit? There’s a book you’ve written, but the last couple of years you’ve been working at OSS Capital. Is that right?

That’s right. It’s interesting that we talked a few years ago because things were very different for me than personally as far as what I was doing for work. I was still at a big law firm and was mostly working as a lawyer, but I was working with OSS Capital as an advisor at that point. At the beginning of 2022, I went to OSS Capital full-time. I still have a law practice, but it’s less than it used to be. I still like doing both because they’re really synergistic with each other. Since then, I’ve gone full-time at the fund. Things have changed a lot for me. I work at home full-time and what I’m doing is a little different.

From Project To Profit

What’s the Genesis story of the book that you’ve written? Where did that idea come from?

I’m the kind of person who likes to write in order to organize my thoughts. I have three nonfiction books. It’s not coincidental they track my career. I wrote one on technology licensing when I was doing basic technology licensing. I wrote one on open source when I was doing open-source law mostly. This new one is about building businesses around open source projects, which is squarely what I am focusing on at the VC fund. That was the genesis of it. I was thinking more and learning more about how to help companies grow when they are based on open-source software. I thought, “I might as well get this down in a book,” because that’s the way I like to do things.

When you mentioned it to me, there are a lot of seminal software and business books around specific areas. I don’t know if you’re interested in car manufacturing. There’s a seminal book about Toyota. It’s a bit too early for something to be defined as the seminal or the thing that everyone refers to when they’re talking about this. It made me realize that a lot of the learnings that we’ve acquired at FlagSmith, and I’m sure that you guys at OSS have, have been fished out of the stream from little blog posts here, a little video there, a thread on Twitter, or something like that.

That’s right. I don’t think there’s another book precisely about this topic, and there’s a reason for that. There’s some tension philosophically between open source advocates and business people even after all these years. A lot of people are disinclined to go to the trouble to write a book about it. There are not so many people who are focusing on this day and night. In my book, you might find some new insights, but a lot of it is stuff that you could find if you looked around a lot. Putting it all together into one thing that says, “Here’s the basic stuff you need to know if you’re considering doing this,” I don’t think there’s another one.

Have you written it for a group in particular? Is it more for people who’ve started an open source project and are considering commercializing it or somebody who works in a large organization who’s looking to do something, or is it written for multiple sorts of people?

Mostly, the first group you mentioned, someone who is working on an open source project and is thinking, “Could this be a business for me?” is the basic audience for the book. Also, there are lots of people who work at big companies who are doing open-source projects and are thinking, “Should I go off on my own and do this?” It has some helpful hints for people who are in that situation as well.

In terms of the large trends that you’ve experienced and you’ve researched around this, are there a handful of trends that you think you’ve identified and are consistent in the last few years?

Yeah. There are a number of them. First overall, business is built around open source. In OSS Capital, we call it COSS businesses, Commercial Open Source Software businesses. They are getting more popular, and we think they’re going to continue to be more popular. We think they’re taking over the software industry. One interesting trend is that if you look years ago maybe at the COSS businesses that are out there, you would find that almost all of them were in the infrastructure area, operating systems, database, content management, security, etc. Now, if you look at COSS businesses, a lot more of them are in the application area.

Businesses built around open source are getting more and more popular. They're taking over the software industry.

The book talks about why that is and why applications work for open source. The traditional wisdom was that it didn’t work in the application space but we’re seeing more businesses that are working very well in the application space. That’s a sea change. It means that open source-based businesses are branching out into all areas of software. People who say, “You have to be a Red Hat-like company in order to be successful in commercial open-source software,” that’s not true anymore. They were fantastically successful, and the book explains why it might be hard to duplicate that model, but there are successful businesses in all sorts of areas.

What do you think has driven that move up into the application there?

People have begun to understand, and this is a core principle of the book, the difference between source code and a product. People buy products. They don’t buy source code. A lot of people say, “How can you possibly make money with open-source software?” If you look at it that way, it makes total sense because nobody buys source code. They buy products.

For companies who are in the commercial open-source software space, what they’re selling is the delta between the source code and the product. It’s that people have begun to understand better that a product is a product and it doesn’t matter whether it’s an application, infrastructure, or whatever. You can still sell products because people don’t buy source code. Why sell it or why keep it for that matter?

There will always be particularly large customers who will look at your source code and say, “I can set that up on my own and run it on my own” Honestly, that’s happening less because of the vertical disintegration of software people. A CTO, when they’re buying a product, they want to solve a problem. They don’t want to get some source code and create a lot of headcount, trouble, bill of materials, and everything for them. That was also part and parcel of the move to the Cloud.

People, when they’re buying software, are most of the time looking at buying SaaS instead. For that, open source is a fine business model because you’re not giving people software at all. It’s pretty easy to open up the software and give people the benefits that they get from the open-source software without leaving too much on the table in terms of business.

I’m conscious of thinking about how projects and companies apply in this way. The thing that you haven’t mentioned there is the data as well. There’s an interesting play there as well in that, and I’m fairly naive in this area, you can have this AI that the runtime is implicitly open source, as in the actual code that defines the neural network and runs the actual execution and the runtime. With the actual models, it has taken the actual source code for these neural networks. I still call it neural networks because that’s what I studied many years ago.

That’s pretty accurate.

It’s open. There are companies figuring out there are some very popular LLM models that are open source and then there are some that are private.

There are a few that are open source, but there are very few that are truly open source. What constitutes open source in the AI space is a huge definitional issue for the industry. There are lots of people thinking about that. I’m personally working on at least two different projects to try to define what open-source AI might mean because it’s different. There’s no source code for it. The data may be freely available, but that’s not usually the case. The software that you use to build it is usually freely available, but that is called comfort for a user who is getting a model that’s a black box.

What constitutes open source in the AI space is a huge definitional issue for the industry right now.

Also, a lot of the LLMs that are being provided to people are only being provided by a SaaS. They’re not being distributed at all. Some are, and some aren’t. That’s going to be a big change in the software business because if we had this conversation in another five years, we would be talking about AI and software almost interchangeably. The software landscape is going to be a mixture of traditional procedural code and AI.

A Few Pieces Of Advice

A lot of the guests that I’ve had on the show have started their projects and businesses like this. They’re sweating away on evenings and weekends. In the current climate of early 2024, what would be the three most important pieces of advice you could give to those people?

There are a few things, and these are outlined in the book. The first thing you have to do is ask whether you really want to run a business, and if so, what kind? People who run open-source projects sometimes have big dreams for businesses. Sometimes, they have modest dreams for businesses. Sometimes, they have dreams that have nothing to do with business but they need some funding. Those are all different situations. They’re all perfectly fine situations, but you need to know generally what category you fall into.

If you are not interested in running a big business, you need to be realistic about getting outside investment. That’s going to be difficult. If you don’t really want to run a business and you want to run a community project, and you’re looking for funding, those are different sources of funding. One source of funding is also your sweat equity. It’s important that people consider that as a kind of funding too which costs the developer. It’s not magic and free. That’s one thing. What kind of business do you want to run?

‍

If you decide that you want to run a business, particularly if it’s something other than a maintenance and support business, which is more like professional services, if you want to run a product business, then you need a very clear business model to get from where you are now to where you want to be. The clearer and more critical you are about thinking about your business model, the better off you’re going to be.

There is a magical thinking that goes on in open source-based businesses. Sometimes, it’s like the old South Park underpants gnomes thing. It’s like, “Step 1, collect underpants, step 3, profit, and then step 2 is a question mark.” There’s a lot of that in open source. You need to know what the question mark is. How are you going to get customers who will pay you? What’s your customer base? Think about pricing, although you can work that out later. You have to think critically about all those things before you can really start a business successfully.

The last thing I’ll say is that the one thing that founders often miss about preparing to start a business is their trademark and branding. It’s enormously important. The reason they forget about it is that in most open-source projects, the people running them don’t consider the name of the project to be a brand in the traditional sense, so they don’t manage it as a brand. The problem is that if you don’t, then later on down the line, you’re going to have some challenges in managing the brand. The earlier you start thinking about it as a brand, the better. 

We’ve seen all sorts of examples like this. The quintessential one is Linux where the name was not registered for a long time. They had to backpedal to figure out how to manage that brand. Even now, it’s not really managed as a brand the way other brands are managed. Don’t make that mistake. Think about your branding and your naming early on because you’ll be glad you did down the road.

Think about your branding and your naming early on because you'll be glad you did down the road.

It’s interesting because the exact same thing happened with FlagSmith. FlagSmith wasn’t called FlagSmith when it started life on GitHub. It was called Bullet Train, which was a name that I’d given about 30 seconds of thought to. I’d come from running a business and working with customers who had started a company and hadn’t done the right searches for trademarks and collisions with things like that and I made exactly that mistake. I was like, “It’s some random project on GitHub. It seems like a good name.”

It turns out it’s a terrible name not only because there was another open-source project called Bullet Train that was four months older than us, but also because you are competing with actual trains in search engine optimization. We made exactly that mistake. It’s fascinating. I’ve been helping people start businesses and building their software for them and stuff for 25 years, so I should have known much better.

The guy that owned the other Bullet Train open-source project got in touch with us quite a while ago and was like, “We need to sort something out here because it’s my name,” and it was. He launched it four months before us. I’m not an expert in this area of law, but I learned quite a lot about it. It was like, “We need to change our name.”

It was a very valuable moment because we then got to go, “Let’s figure out what we want the name to be, what we want the right brand to be, and what we stand for.” It’s not sitting at the new project prompt on GitHub and coming up with something. I’m always surprised how often on Hacker News, you get a new open source project blah, and then the first 25 comments are like, “This is the same name as this project or that project.” They’re always like Greek gods, drinks, or things. It’s hard.

It may be completely generic stuff. I don’t even want to say one because it might be an actual name like MySoftware or something. That’s really what most people do. They don’t pay any attention to it. It’s in part because you don’t know at that moment when you’re creating the project whether it’s going to be a big deal for you or not. It might be a fun project, so you never care what it’s called, but it’s important to pick a name that works from a trademark point of view and then do your filings and stuff to protect it.

As you learned, you can rebrand, but it’s not much fun. The better you have done in getting your community together, the more painful it is. You don’t want to be making the decision a year or two down the road after you have successfully built a community around your project that you have to rebrand. It happens all the time. It is a very common problem. That one piece of advice to treat the name of your project like a brand is probably one of the most useful ones particularly early on in your project. You can fix a lot of things later, but rebranding is painful.

We were very lucky because the company was very small at that point and we didn’t have a huge amount of capital value invested in that name. The name’s super important. I talk about this a lot. I want us to be 1 of the 3 company names that engineers think of when they’re looking for a feature flagging tool. If you change your name, then you are doing a tremendous amount of damage to that global mind share of engineers. 

Another thing to fill in the gap is that you have an open-source project. It’s got a license on it, like MIT, Apache, GPL, or whatever. Those licenses really don’t do anything about branding. All of the laws about branding are outside of those licenses. There are some licenses that say some minor things about it, but they don’t even have any legal effect. This is completely orthogonal to the license. Every business has to protect its branding. That’s the nature of a business. If you’re going to have a business around a project, you need to take the business part seriously. That’s number 1, 2, and 3. It is figuring out your branding. 

Talking about your background in law, one of the challenges around this trademark dispute was that by definition, almost all of these open-source projects are international and have to trade across international law. That was another problem we had. I can’t recall. He might have been Canadian. He was in Canada, the guy with the original Bullet Train. I had some personal experience of this issue in the UK and I was on the other side of the table in that situation, so I understood it from the UK, but I had no experience of how to deal with that in an international manner.

The last thing I wanted to do, sorry to say, was to phone my lawyer and get them on the clock to try and deal with this issue because you don’t want to spend that money. How much does the book talk about the global aspect of this? It’s not a legal issue. It’s a team issue, a time zone support issue, and a customer issue, right?

Yeah. The book does talk about it. I also talk about this in my Open (Source) for Business book maybe in a little more depth, but you’re exactly right. Trademarks are jurisdictional. If you have the trademark protected in one nation, it doesn’t mean it’s protected in another nation. When you do branding protection, you have to have a plan for where you’re going to seek protection. You don’t seek protection everywhere because that would be ridiculous and expensive.

‍

The thing I’d like to emphasize is that if you do your branding protection work early on, it’s not at all expensive. You can file applications and do some trademark searches. You’re talking a few thousand dollars in legal fees, maybe $10,000 if you do a couple of applications. You usually do 2, 1 for Word and 2 for the logo. You then do counterparts in the US, EU, and maybe a few other countries where you expect to do your most business. It’s cheap and easy and you’ll be glad you did it.

The other thing that some people don’t tend to understand is they think that if they’ve got a domain name, then they can use the trademark. They’re different things. When you use a domain name, that does require the use of a trademark, but it doesn’t go the other way if the converse is not true. If you have the domain name, it doesn’t mean you have rights to the trademark. When most people start a company, the first thing they do is go look for a domain name. They think that if they have cleared that, then they’ve cleared the trademark rights. That’s not quite true. You also have to do separate trademark clearances. If you do it early, it’s not so difficult.

One more piece of advice for people who are thinking about this is to think about 3 different brands you like, not just 1. That’s because if you think of one and then it doesn’t work, what you’re going to end up doing is paying a bunch of money for a domain name or paying a bunch of money to solve branding problems. Whereas if you pick 3 and the 1st one doesn’t work, you move on to the 2nd one. It’s important mentally not to get married to one brand when you’re starting a business and to be a little bit more creative about it.

You can also hire branding consultants. They’re very good, but they can get a little expensive. If you’re going to do it yourself, you think of 3 brands you like. You go and look on Google and so forth. If they look like they’re not occupied already, then you go and talk to the lawyers. You’ve then done a lot of the preclearance work on your own.

How OSS Does Its Job

In terms of with your OSS hat on maybe a little bit more, do you guys have a framework? How’d you go about evaluating an open-source business? If someone mentions a project to you or a founder connects with you guys and says, “I’m Ben. This is my project,” what sort of things do you look for in terms of signals for you as an investor?

What we look for is probably very similar to what any investor works for, but it has a little bit of a spin with open source. First is technology. Most of the projects that we find, we find either because we happen to find them and we like the project, and it looks good. We’re looking at TechCrunch and all that stuff. We’re looking at all the online sources of news like Hacker News and so forth that everybody's looking at. We’re looking at, “What are the hottest open source projects?”

Sometimes, we then reach out to the founders of those projects and say, “Do you want to start a business?” For VCs, that’s a little unusual. Most VCs have companies come to them with a proposal. With reasonable frequency, go to the founders. Since we’ve gotten better known, people are more likely to come to us. We were a startup fund for the first three years or so. We’ve been operating for about five years. People come to us more. Plus, the economic climate has changed, so they’re coming to us more. That’s number one, technology.

For us, the open-source spin is we want to see a release that’s been made. I have a lot of people come to me and say, “I’m thinking of starting a company and I’m going to release the code open source.” Usually, we’re not interested in that. Usually, we want to see that the release has already been made. We’re starting with technology that’s open to look at. As an investor, that’s great because we can do diligence on that without doing NDAs, starting a process, or anything.

The next thing we look at is the people. Part of what we’re looking for in founders is people who really want to run a hugely successful business. We’re venture investors. Venture investors invest in high-risk situations and want big returns because they’re high-risk situations. In order to make the economics of it work, you have to have some real home runs.

We’re looking for the kind of people who really want to run a successful business. It doesn’t mean that you have to have done it before. Many of our founders are first-time founders. You have to have that dream, and that’s part of what I talk about in the book. If you don’t have that dream, don’t go to a venture investor because it’s not the right fit. There are other things that you can do. Honestly, we’ve met some founders who had great projects and were great people, but we could tell they didn’t want to run a conventional business. They’re not the right fit for us.

Ifyou don't have a big dream for your project, don't go to a venture investorbecause it's just not the right fit.

It’s not like these are in order of importance. These are the top three things. The last is market fit. We want to look at something that’s going to have a good market. It’s great if there are no products in that area that fulfill the needs of that market yet. I have to say honestly that one of our strategies is to find the proprietary software product that everyone uses and everyone hates and see if there’s somebody developing an open-source version because we think categorically the open-source version will be better. 

There are a lot of products out there that have what I don’t know if I’d call them legacy products exactly, but they’ve been around for a while. They’re used by everybody, but they have serious usability issues. They’re not getting properly developed by the company that’s developing them. That’s where an open-source project has a massive ability to disrupt and make money. If you do something really well in that area, you’ve got a huge market and a motivated market and you’re going to take advantage of the complacency of the existing players to win in the market.

It’s one of my soapbox issues. I am really focused on software usability. I don’t think our world has focused on it enough. It’s hard. Making software usable takes a lot of work. One of the great things about open source is that some open source projects tend to be a little do-it-your-self and difficult to use, but the good ones are easy to use and they don’t tend to suffer from the same feature bloat issues that the proprietary suffers from. When you’re selling proprietary, how do you make new sales to existing customers? You add a feature and say, “You need that feature,” or, “We’re going to make you buy that feature,” because they’re in the situation of dictating terms.

I am a lawyer. Lawyers are power users of Word processors. We do lots of formatting and so forth. It has so many features that nobody understands. When I worked at law firms, they mostly didn’t support most of the features of the software they had chosen and were making you use. We had to use that software. You couldn’t go and use LibreOffice if you wanted to or Google Docs or whatever. They have tech policies about this. When you ask tech support, they would say, “We don’t support that feature.”

Why does that happen? It happens because a feature gets added to the software in order to sell the software to maybe 1% of users, but then everybody else is saddled with it. That doesn’t tend to happen with open source because if 1% of users want to feature in open source, it doesn’t get put into the product. It may be put into a custom version of the product or something, but no one’s going to code up a feature that only 1% of the community wants. It ends up being cleaner, easier to use, and better software in the long run.

You talk about how there’s the experience of the software, the experience from the user’s point of view, and then also the experience from the infrastructure point of view. One of the things that we’ve found really successful is if I find a new open-source project, I want to be able to have it up and running in a few minutes on my laptop and hack around with it and kick the tires on it.

Technologies like Docker make that much easier. It used to be a lot harder. This was for the benefit of the open source project, not with an enterprise customer in mind to make the experience of installing, deploying, upgrading, and maintaining the infrastructure around that software really good. That’s what I want to see if I’m coming to a project. I want the documentation to be good. I don’t want there to be any weird, wacky errors coming out of the log that make you think, “What the hell is going on?”

It turns out that that means that you’ve then got an amazing experience when you’re selling your software into the enterprise market. We’ve had enterprise customers who’ve been really surprised. They’ve put aside two weeks to deploy the software, and then they get it done in an afternoon or whatever and are like, “Wow.” I’m not saying it’s as simple as that. There are a lot more things I have to consider about security, scalability, and stuff.

That’s a really great point.

I didn’t have those people in mind when we did all that work to do that.

That underscore is an important thing about open source-based businesses. One of the things that they do very well is sell through the developers instead of selling through procurement. In fact, that’s the main way they usually sell, particularly at the beginning. The easier you make things for the users on the technical side, the more successful you’re going to be. That’s a great point and an example of a different kind of usability that is way more likely to happen in open source. If you don’t do that with open source, nobody’s going to bother using your code, right?

‍

Yeah. It is true. That’s very much the case with FlagSmith. We call it very much a bottom-up sales approach where an engineering team will contact us. We virtually never get contacted by a CIO or CTO of a large organization that wants to roll it out, unfortunately for us, that’s like, “We’re going to deploy this to 10,000 engineers or whatever.” As a result, for those developers, the first thing they’re going to do is to evaluate your project based on the quality of the deployment experience.

Whereas if you’re procuring a bit of software for 10,000 developers, you might not even consider those big matrices of, “Does your software have this feature? Do you support this security standard? They almost never ask you about your installation and deployment. They might ask you about your upgrade like, “Can you do zero downtime upgrades?” or whatever, but they never ask you any sanitization questions about how you install and deploy the software, which is a huge deal, especially with a lot of context.

That’s another example of the difference between source code and a product. Having source code is one thing, but in order to get it installed and usable, it is another thing. The more that you can provide for people there, the better. It underscores the value of software versus software usability, which is huge.

In terms of the business models that OSS is choosing to invest in, are there patterns emerging that you can share?

I can tell you maybe the business model that’s working really well. I’d like to emphasize that there are lots of models that work, not just one, and every product is a little different. It’s a bespoke process to figure out what works for your company. There are two that are very common. One is in the infrastructure area. You release the code under a permissive license like MIT or Apache, and then what you upsell are enterprise features that will make the software work with other software. It’s not that it won’t to begin with, but it takes a lot of money to develop connectors, glue, and stuff like that as opposed to basic functionality.

You release a basic functionality and then you sell things like connectors, glue, services, maintenance, and all the accouterment that make it useful within a large enterprise. The one thing that does is it segregates your community into the small users and the big users. You’re going to end up charging the big users, which makes a lot of sense. There’s too much trouble to charge the small users. Plus, let them have it for free. It’s good branding.

The other is more common in the application space, which is to release the software under a Faro GPL which is the strongest copyleft license that is in common use, and then you backend that with Cloud services. You’re like, “Here’s a GPL version if you want to run it on your own, but then if you use it for a Cloud service, you have to share your changes.” People who don’t want to do that will come and negotiate licenses with you, and then you sell Cloud services to back that up.

Those are both what I would call open-core models. The term open core means you have a core of software that’s open source and then you have some other stuff that isn’t. That other stuff, if you look at it from a 50,000-foot view, can be software, services, and lots of things. It can be hosting. As an overarching matter, the open source software is not a commercial product. The other stuff is a commercial product. It doesn’t mean you don’t professionally support the open-source product. It means that you’re not selling it.

Both of those models work very well if you do a little thinking about them. Taking a look at the book would help you plan out how to do that. You have to fine-tune these things as you go along. Things change. The business model that works for you on day 1 might not work for you 5 years from now. The important thing is to think about it and be realistic about it.

That reflects our experience. An important learning for FlagSmith was exactly what you said. The non-core stuff can be all sorts of things. A lot of engineers assume that that means code. You said that delta isn’t necessarily lines of code, right?

Sometimes, it’s code, and sometimes, it’s not. It can be all sorts of things. It can be content. It can be data. It can be software. It can be hardware. It can be all sorts of things. That’s how you build out an open-source model. You say, “I’ve got one thing,” which is a community product, and then I’ve got another thing, which is a commercial product. How do those interact?”

One of the things that the book spends some time on is, from an economic point of view, how to think about those because most people have a tendency to think of them as substitutes. They’re like, “You use the community version and then move to the enterprise version.” They start thinking about conversion rates and are like, “How do we convert more people?” All that’s fine, but what you have to understand is that the two are complementary and not substitutes.

You want to have a community product that’s robust. You also want to have an enterprise product that’s robust. You don’t want to get rid of the community product. This is the thinking mistake people make in planning open-source businesses. They think the community thing is a cost or something like that. It’s not. It’s an important asset of the business that is going to drive sales for whatever people are selling.

It’s not because they’re not going to use open source anymore necessarily. It’s because it’s creating trust. It’s creating brand awareness. It’s doing all sorts of great things for your company. You have to think of that as an asset. That’s the fundamental thing that people have to think about, particularly investors when they’re thinking about open-source businesses.

For us, the open-source FlagSmith is probably our best inbound lead generator there.

It does its own work. It works for itself. 

Exactly. A lot of those leads come in there almost at the bottom of the funnel anyway. They’re like, “We’ve been running your open-source version for six months.” They’re the sales team. They’re tapping into those goals.

By the time someone comes to you in that situation, you don’t have to prove yourself to them. They already trust you and like you, so then the sale is much easier too as opposed to going out and trying to push sales to people. Its brand value is what it is.

Future Outlook

Before we go, I’m curious. It’s the start of the year. Maybe we can chat in three years’ time. How do you think the landscape’s going to change and evolve over the next year or two as far as commercial open-source companies are concerned?

First of all, in OSS Capital, we have been fortunate that during the past couple of years, we’ve been going full bore like we were before. Most VCs were experiencing real difficulties. First of all, we’re early-stage and our funds are not huge. We were able to fundraise during that time, which, believe me, was not easy during those years but we were able to do it. My partner, JJ, is the sin master of that.

Also, we did, over the past couple of years, 45 investments maybe. If you look at what’s going on in VC during those years, there wasn’t a lot being funded other than AI companies. We were able to keep going like business as usual despite the economic times. Also, of our portfolio companies, we had only maybe a few that really had economic issues. It’s nearly impossible for software companies to get Series A rounds. We had a few during that time, but it’s tough now.

The thing is, the open source companies are so efficient with the money we give them. We talk to them and are like, “How much runway do you have?” They’re like, “We have about three years.” It’s like, “That’s good.” Most of them are virtual companies. I’m hard-pressed to think of one of them that doesn’t have people outside the US working on it. They’re really international companies.

That’s a long way of saying that economic times will get better in 2024. We’re all hoping for that. Open source weathers economic problems really well, so we’re happy about that. That’s what we thought. I wrote a couple of articles on that back when times were good. During those times, I’m always thinking, “Something bad is coming.” I wrote a couple of articles about why open source is good in recessionary times and inflationary times. I have to say I was very pleased that it looks like all that was born out because things have been going well for our companies. It’s going to keep going. I see no end in the momentum of people moving to commercial open-source businesses.

It accumulates as well, doesn’t it? Several years ago, if you were in a large blue-chip company, they’re like, “The source code is what?”

It’s getting better slowly.

We’ve noticed that, especially in procurement teams or answering those big security questionnaires. They’re slowly changing, but you can see where they understand that it might be an open-source business also. Everyone’s ships are going up on the rising tide in that regard. We’ve experienced that.

That’s right. The developers already understood it. The business people and the lawyers have been playing catch-up for a long time, but they’re getting more accustomed to understanding that having open-source software in your bill of materials is not a negative. In fact, it could be a big positive. They’re going to be seeing vendors who are primarily open-source.

Thank you so much for your time. Before we go, you’ve made it to the end. I’m more than happy for you to give your book a plug in terms of where you can buy it, what it’s called, etc.

It’s called From Project to Profit. It’s available on Amazon.com in either paperback or Kindle formats. I’m working on an audiobook. With luck, that should be up in the next couple of months or so. If you want more information about it, you can look at my website at HeatherMeeker.com. There’s a page for the book on it which has some more information. It’s a short book. I’m told it can be read in a day. I did it that way on purpose because there’s no point in padding books out. It should be very accessible and comprehensible, I hope. I hope people will be interested and enjoy it.

There’s nothing worse than a business book that gives you the goods in the first chapter.

I hate those. There are 200 more pages to read.

Thanks again for your time. I’m going to put a reminder in two and a half years to hit you up. Thank you.

Thanks so much.

Important Links

• Heather Meeker
• OSS Capital
• Interview with Heather Meeker: Founding Partner, OSS Capital
• Open (Source) for Business
• From Project to Profit
• https://HeatherMeeker.com/From-Project-To-Profit/
• https://a.co/d/aGkorOJ

‍