The Business of Open Source Emily Omier

This week on The Business of Open Source, I spoke with Brian Fox, co-founder and CTO of Sonatype. In addition to having a really interesting discussion about the usual topic of how to build a business around open source software, we also had a good conversation about security — it was hard to avoid, because we recorded this right after the xz backdoor discovery, and software supply chain security is kind of Brian’s thing. 

Business-wise, though, we also covered some really cool topics. Including: 

The tension between an open source project that’s “too good” and yet the need for the sales team to close dealsIn some ways, the fully commercial, closed-source products in Sonatype’s product line are more straightforward… but there are challenges that go along with a pure closed-source approach, too, especially for a DevTool company. Choosing your relationship with open source depending on who your target user / target buyer isPivoting to a top-down sales motion because the bottoms-up motion just didn’t work; and how that means the features that sell aren’t always the features that get usedWhat Sonatype gets out of it’s relationship with Apache Maven and open source NexusHow do we solve real problems, and how do we solve them for real? Keeping in mind that no one buys what they need; they only buy what they want. 

Check out the full episode, and come to Open Source Founders Summit if you want more opportunities to talk about about business and open source. 

This week on The Business of Open Source I had Rod Johnson, founder/CEO of Spring Source and creator of the Spring Framework (as well as board member of many other open source companies) on to talk about Spring, monetizing open source and what’s changed in the open source ecosystem since 2008. 


Key takeaways:

Consulting was burning the entire team out, and that threatened the health not just of the consulting business, but of the open source project as wellAn amazing salesperson can often sell anything, but that doesn’t mean that you’ll be able scale, because your entire sales team is not likely to be incredibly brilliant Spring Source ended up not monetizing Spring at all — but rather worked on monetizing with products that were complementary to Spring. “We monetized Spring by not monetizing Spring, by using it to open the door” The moment that the company really started to see success as a product company was when the team stopped thinking about what they wanted to build and instead focused on what customers where telling them that they wanted.The risk of having a bunch of very good engineers on your team is that they’re excited about solving hard technical problems — but your customers might want something that is not very technically challenging or interesting. A major part of the job of a company leader is to talk to your team and get them on board with your plans The environment around monetizing open source projects has changed — there are things that worked in 2008 that wouldn’t work today, and things that didn’t work then that would be fine nowIf you love (insert your favorite open source project here), it has to have a sustainable economic modelIt’s really critical to have a rationale behind what functionality goes in your product and what goes into your open source project
At the end we talked briefly about Open Source Founders Summit, a conference for leaders in open source businesses happening this May 27th and 28th in Paris. 

This week on The Business of Open Source, I spoke with BoxyHQ co-founder and CEO Deepak Prabhakara. We talked about a number of things, from BoxyHQ’s relationship with its open source project, called SAML Jackson to how to build a growth flywheel and how that flywheel does and does not depend on a community. 


Is BoxyHQ a security company? Does it matter either way? Starting the open source project at the same time as the company, and why they did it that wayThe relationship between the user community and the customer communityBoxyHQ as the anti-platform — instead of trying to build a platform, which is the default goal for a lot of companies I speak with — they are explicitly trying to build a more a la carte experience for usersThe challenges of community building around a project that isn’t sexy and how to build community that isn’t project-focused, but rather that’s focused around a problem spaceMaking the mistake of assuming your startup is completely unique and unlike any others! 

We talked about much more as well, and it’s definitely an episode you should check out. 

In the second episode that I recorded on-site at KubeCon EU in Paris, I spoke with Alex Olivier, CPO and co-founder of Cerbos. This was not a general discussion: It was focused on the process that Cerbos went through to figure out pricing. 

Here’s what we talked about:

The first step of figuring out your pricing is not the number, but rather what you’re charging for. Is it API calls, or amount of data you’re processing, or monthly active users, or monthly active principles… that last one is what Cerbos is charging forWhy it’s important to have a pricing system that allows potential users to be able to roughly estimate for themselves how much using your software is going to cost themYou also want to avoid pricing models that encourage people to look for ways to hack around to find ways to lower their monthly costsWhy your pricing model should be about the value you’re providing, not about how much it costs you to run your systemDiscovering what your price anchors are / what your customers are comparing you to
Check out the full episode for more details! 

And join us at Open Source Founders Summit for more discussions about the specifics of pricing for open source companies. 

This week, I had a dilemma: should I prioritize the episode where I spoke with one of the MariaDB co-founders, in which we discuss setting up a foundation as a way to ensure that the project continues to be open source in the future, no matter what (relevant given the Redis announcement); or should I prioritize the conversation with one of the founders of Sonatype, one of the oldest companies in the software supply chain security space, in which we talk about the xz debacle. 

I went with Patrick Backman, general partner at OpenOcean and co-founder of MariaDB, because it’s a little more in my lane. (The conversation with Brian Fox will have to wait for next week!). 

One of the main things we discussed was the relationship between the MariaDB foundation and the MariaDB company. Including: 

Why they decided to put MariaDB open source in a foundation, and why they created a separate foundation instead of putting it in an existing foundation The relationship between MariaDB foundation and company today, including the financial relationship
MariaDB was founded by the founders (and some key employees) at MySQL; we also discussed the lessons learned at MySQL that the team then applied at MariaDB. And we talked about customer acquisition, one of the things that Patrick thinks the team had learned at MySQL and therefore had pretty well figured it out at MariaDB. 

Patrick’s co-founder Monty Widenius is one of the speakers at Open Source Founders Summit — if you want to go into more details on with the lessons from MySQL and MariaDB, as well as lessons from being an investor at OpenOcean, join us in Paris May 27th and 28th at Open source Founders Summit. 

This week on The Business of Open Source, I have an episode recorded on site at KubeCon EU in Paris with William Morgan, CEO of Buoyant. We had a fabulous conversation, which touched on some touchy subjects, including Buoyant’s slightly changing relationship with Linkerd. But we talked about:

Being an open source mercenary, but also being dedicated to making Linkerd a ‘proper’ open source projectFeeling like open source was table stakes for a company in the space Buoyant plays in. This is an under-appreciated reason for being an open source company — you feel like it’s just expected in the market you play in, so you do. Waiting too long (or is it too long?) to commercializeStarting out by selling support, but the problem with that because Linkerd worked well and people kept saying that they didn’t need support because they never had problemsCompeting against Istio, which was backed by the Google engine and how that made Linkerd / Buoyant an underdog (or cockroach). For those of you who haven’t been following Linkerd / Buoyant… Buoyant recently announced that they would be doing edge releases for Linkerd, but not stable releases. We talked about why they made this change and how the ecosystem responded. 
Check out the full episode!