Creating an Open Source Culture

Creating an Open Source Culture

Armory is commercializing the open source project Spinnaker, making it a "COSS" company. As Joseph Jacks puts it:

COSS companies like Armory create and capture value differently than traditional proprietary SaaS companies due to a Flywheel Effect, which my CTO Isaac Mosquera shared in his Spinnaker Keynote:

I call this interplay between community, platform & enterprise consumption the "Flywheel of Inevitability" – and it's what makes open source projects like Kubernetes, Linux, Git and Elastic so powerful.

Balancing this Flywheel of Inevitability with Armory's growth means that every decision Armory makes happens differently than it would in a closed SaaS company. For example:

  • Should Armory open or close source any given new feature? Open sourcing it makes the Spinnaker project better, which spins the Community side of the flywheel, growing the size, importance, and relevance of the Spinnaker project, which sits at the core of Armory's software delivery platform. Close sourcing it helps ensure that Armory captures enough of the value we're creating to enable us to build a successful, rapidly growing, and enduring company, which we need to do in order to continue investing in the Flywheel.
  • Where should Armory put any given marketing dollar? By promoting Spinnaker, Armory helps ensure the world's largest Global 2,000 companies know there is a better and more modern way to deliver software into the cloud. By promoting Armory, we help ensure companies understand how Armory is extending the open source project to meet their enterprise needs.

And so it goes – just about every decision at our company, every day, has a COSS lens put up to it that requires Armory to do additional thinking about whether we are going to invest in the Flywheel itself, or invest in our ability to invest in the Flywheel by growing the company.

The short history of open source is littered with examples of companies that didn't get it quite right:

  • Docker: Did a phenomenal job of investing in the container Flywheel, which then started spinning faster than its ability to commercialize the Flywheel, and it got swallowed by Kubernetes.
  • Mesosphere: Did not invest enough in getting the Flywheel of Inevitability spinning. Although Mesos was more scalable than Kubernetes, it was also harder to adopt and use, and the company had to pivot to adapt.

It's also true that the Flywheel of Inevitability is not required for a company to successfully commercialize a project. For example:

  • GitHub: Git was (and still is) hard to use before Github. Github commercialized Git by wrapping a proprietary platform around an open source project to make it accessible.
  • Mulesoft: While Mule is open source, the project never had meaningful escape velocity. However, Mulesoft built a very successful company around it.

Both those companies had ~$7 Billion exits.

I believe, however, that there is a class of open source project that can have many multiples of that level of value creation, and therefore, opportunity for value capture. Linux and Kubernetes are two excellent examples. These are projects that are literally changing the very fabric of technology and society. And successfully commercializing these projects is orders of magnitude harder, because the Flywheel of Inevitability spins so fast that, like a black hole, it consumes and ejects the companies that try to commercialize it.

I also believe that Spinnaker is one of these projects. The world just doesn't realize it yet. Spinnaker is a multi-cloud software delivery orchestration engine. It is changing the way the world's largest companies deliver their ideas (aka software) to users, which, as software eats the world, is effectively the same thing as saying that Spinnaker is enabling them to out-innovate and out-compete their peers over the coming decades.

As Isaac describes in his excellent Spinnaker Summit keynote above, what makes Spinnaker special is the community of companies that are positioning it "above the clouds" – including the cloud vendors themselves. AWS, Google, Microsoft, Pivotal, Oracle, Alibaba, Hauwei, Tencent, and many enterprises are coming together as a community to support Spinnaker as the de-facto standard for cloud-native software delivery.

Think about the power of that for a second. I just put the world's largest cloud vendors together the same a sentence and bound them together with word "community." That is the power of Spinnaker.

Spinnaker already has over 2,400 contributors, including many of the world's largest Global 2,000 enterprises such as The Home Depot, Nike, Adobe, SAP, Samsung, IBM, Salesforce, HP, NTT, Volvo and JP Morgan Chase.

The Spinnaker Flywheel of Inevitability is starting to spin... hard. In order for Armory to successfully commercialize this project, without swallowed by it, I believe we need to construct a new type of organizational operating system for our company. To put it simply: Armory is focused on building an Open Source Culture.

We recently hired Kate MacAleavey as our Head of Culture and Leadership Development to help us build this Open Source Culture at Armory, and together we spoke to Joseph Jacks about this:

Building an Open Source Culture is a new term with deep implications. We are still exploring the boundaries of what it means for the organizational structure and working style of Armory. It's a journey I'd like to share with other companies that are also focused on building strong cultures that compliment rapidly spinning Flywheel open source projects. I would describe the basic tenants as:

1) We trust and empower those at the edge: Building software features with strong product market fit – and even more importantly, knowing when to invest in spinning the OSS Flywheel vs. proprietary – is an iterative process that requires those with the most context– those at the edge of our organization– to be empowered to innovate. The traditional command + control model of management breaks down in this environment. As the CEO of the company, my job is to empower and enable our tribals to do their best work. We are doing this by creating a "Roundabout, not Stoplight" culture. We are creating a culture that defaults to transparency and enables context over control.

"To put it bluntly, servant-leaders have the humility, courage, and insight to admit that they can benefit from the expertise of others who have less power than them. They actively seek the ideas and unique contributions of the employees that they serve. This is how servant leaders create a culture of learning, and an atmosphere that encourages followers to become the very best they can." - Harvard Business Review

2) We are a Learning Organization: Learning organizations prioritize psychological safety, appreciation of differences, openness to new ideas and time for reflection. This HBR article describes those basic tenants, and this one dives into more depth on how humans learn. We believe in Minimum Viable Policy, we prioritize rapid, iterative experimentation, and we retrospect on what we learn when those experiments aren't successful.

Many people talk about these things as "culture." But as Niels Pflaeging says, culture is "read-only."

Niels' work on Org Physics, and the "Value Creation Structure" - realm of reputation, is where the importance of creating an Open Source Culture to successfully spin an open source Flywheel of Inevitability (without it swallowing the commercializing company) really shines. As Neils writes:

"From this structure arises a particular third kind of power. We call this power Reputation. You have seen power of those with mastery, or Reputation, happening. It is when people who have a work problem which they cannot solve on their own, turn to someone else, asking: "Who knows about this?" or "Who is the expert on this particular matter who I can ask about it?" They are looking for mastery, and they can find it by hooking up to the network of power that is Value Creation Structure."

A COSS company must have Reputation within the open source community it is commercializing. A COSS company is beholden to stakeholders that are not on its payroll. The cellular membrane between those who work at Armory and those who are part of the broader Spinnaker community is permeable, and we are working to make it even more so. In a COSS company, the value of the contributors to the open source project starts with the individuals making the contributions, not just the company paying their salary. Traditional command + control proprietary business models completely miss this fact. Employees are no longer replaceable cogs in a proprietary value creation & capture machine. Instead, they are at the core of the company's value creation, and they have to decide on a daily basis whether to make decisions to capture that value, or spin the larger open source Flywheel.

I'm not a fan of using the world "culture" to describe all of this complexity – it's severely lacking as a term broad and comprehensive enough to describe the complexity above. I prefer the term "organizational operating system" as does Aaron Dignan of TheReady, who writes:

"The Organizational Operating System: Every team has at its core a set of assumptions, beliefs, principles, practices, processes, and policies that act as the foundation upon which the day-to-day work unfolds. I refer to this as an operating system (OS)—a collection of implicit and explicit constraints that shape how we operate. This organizational “DNA” is so pervasive, unquestioned, and deeply held that we don’t even notice it."

However, since the phrase "operating system" is commonly (mis)understood in our industry to only apply to code, calling what we are doing "Open Source Culture" is a strong proxy, because it describes how we are empowering the people at the edges of our company and defaulting to transparency so they have as much context as possible to make the best decisions for our ecosystem.

Armory is intensely focused on building our Open Source Culture in an intentional way that successfully spins the Spinnaker Flywheel of Inevitability and captures value without being swallowed & ejected by that Flywheel. If you'd like to join our tribe, I invite you to read Armory's Tribal Handbook, which more richly describes the way we are constructing this open source culture together.


A special thanks to Armory's Director of BizOps, Sue Ko, who originally pointed Armory on this journey early in our company's growth by having our exec team read Brave New Work and continues to invest deeply to optimize our Org OS as we learn, as well as the entire Armory tribe for investing in each other with such intentionality.