Why Docker’s Business Model Struggles: Lessons from the Container Wars

Docker is still alive, but after a turbulent few years it is struggling to find commercial value. Last week Docker announced a change to its software license: professional users of Docker Desktop working for large enterprises will soon need to pay a subscription to continue using it.

At the same time Docker launched a new enterprise‑focused subscription plan called Docker Business. The new terms take effect on August 31, with a grace period until January 31 2022 for developers to comply with the new service agreement and purchase a paid subscription if needed.

The move sparked extensive discussion among developers. While many understand the rationale behind the subscription, others voiced strong dissatisfaction. Critics fall into two camps: some see the decision as a dangerous precedent that could lead to higher future costs, while others argue that the criticism ignores the higher expenses teams face when relying on poorly maintained open‑source software.

Another major concern is the difficulty of internal procurement processes in enterprises, which could deter developers from using Docker and push them toward alternatives such as Minikube or Canonical’s MicroK8s.

It remains to be seen whether developers will find these alternatives more convenient or will simply pay the modest subscription fee to keep their existing workflow.

Docker’s commercial attempts are understandable, but the company has never succeeded in turning its technological innovation into a sustainable business model, even selling its enterprise business to Mirantis in November 2019.

Today, the popular open‑source orchestration tool Kubernetes has supplanted Docker’s Swarm as the core profit driver. Docker now struggles to survive, and the story is far more complex than it appears. InfoWorld recently interviewed former and current Docker employees, open‑source contributors, customers, and industry analysts to uncover the truth behind the company’s decline.

“Every Thing Can Only Be Touched Lightly”

Heavy reliance on massive venture capital, a constantly shifting competitive landscape, and the desire of major cloud players to claim a piece of the market placed pressures on Docker far beyond most people’s imagination.

Since around 2014 Docker began seriously considering how to monetize its leading position in the container market. In 2014 and 2015 the company used part of its venture funding to acquire Koality and Tutum, and launched the first iteration of the Docker Enterprise Support program.

These investments later gave rise to products such as Docker Hub – essentially a GitHub for Docker images – and eventually Docker Enterprise. However, these products never gained true enterprise acceptance; customers tended to partner with more mature vendors or build their own solutions rather than purchase Docker’s offerings directly.

Founder Solomon Hykes, who left Docker in March 2018, said in an interview, “We never produced an outstanding commercial product because we could never focus. We could only skim the surface of everything. Maintaining community growth and building a great commercial product was already hard enough, and we were trying to develop three or four products at once. We poured a lot of time and money into this, but we never reached the finish line on any of them.”

Looking back, Hykes believes Docker should have slowed its product expansion and spent more time listening to customers, gathering insights from the community, and building a team capable of addressing real needs.

Some argue Docker released its best work too early for free. Earlier this year Google’s Kelsey Hightower said, “They essentially gave away their most powerful ace for free. They solved the problem end‑to‑end – building images, storing them, running them – and then there was nothing left.”

Hykes disagrees, stating that core open‑source products must first achieve massive growth before viable commercialization can follow, and many enterprises have successfully commercialized Docker, just not Docker itself.

Early partners such as Red Hat and Pivotal (now part of VMware) integrated Docker containers into their commercial PaaS offerings – OpenShift and Cloud Foundry – and contributed to the open‑source project.

Hykes later realized he had conflated “community” with “ecosystem”. Red Hat was never truly part of the Docker community, and expecting them to be was a mistake that never benefited Docker.

Consequently, early customers like Amadeus switched to Red Hat from 2015 onward to fill the enterprise gap left by Docker. As Edouard Hubin of Amadeus explained, “We moved from early adopters to a major Red Hat partnership, using Docker’s open‑source version with Red Hat providing container support. Containerization was the first major shift away from virtualization, but Kubernetes ultimately changed the game, and Docker could not suppress it.”

Former CEO Ben Golub described the situation as “a battle among cloud providers that used Docker as a pawn”. He noted that the constant need to scale quickly left Docker unable to balance community growth with commercial transformation.

Collaboration with Kubernetes Failed: Arrogance and Hubris

Docker made many regrettable decisions, the most serious being its refusal to adopt Kubernetes as the preferred container orchestration tool, instead pushing its proprietary Docker Swarm.

In 2014 Docker had an opportunity to work closely with Google’s Kubernetes team and potentially shape the container ecosystem.

Long‑time Docker engineer Jérôme Petazzoni admitted, “Our biggest mistake was missing Kubernetes. We were in an inflated collective mindset, thinking Swarm would succeed and dismissing Kubernetes as too complex.”

Multiple witnesses confirm that tense technical discussions took place at Google’s San Francisco office, with both sides holding strong opinions on implementation details.

Kubernetes co‑founder and VMware VP Craig McLuckie said he once offered to donate Kubernetes to Docker, but the talks fell apart. He recalled, “There was a mutual lack of respect. Docker’s team seemed inexperienced, and we felt they didn’t understand the secrets of distributed‑system management.”

Hykes also admitted that Docker’s arrogance dominated the relationship. “We were blinded by our own hype, thinking we didn’t need Google’s expertise,” he said.

Nick Stinemates, an early Docker employee, reflected, “We could have made Kubernetes the flagship Docker project on GitHub, but Swarm’s late release doomed us.”

Stinemates added, “Our self‑confidence clashed with Kubernetes founders’ insistence on a robust service‑level API, while Docker cared little about a single API.”

In the end Kubernetes won the orchestration battle.

McLuckie emphasized, “No one could have predicted how far Kubernetes would go. It’s easy to analyze in hindsight, but at the time we couldn’t see it.”

Upper‑Management Rift

After raising $95 million in a D‑round in 2015 at a $1 billion “unicorn” valuation, Docker hit the peak of its hype cycle.

Stinemates explained, “That round set extremely high expectations and exposed fundamental problems. CEO Ben Golub and Hykes often disagreed. The board tried to mediate, giving the CEO room to push the company forward. Hykes wanted a community‑first approach, while Ben believed early commercial operations would give us control.”

This tension eventually created two Docker tracks: the popular open‑source Docker Community Edition for developers, and Docker Enterprise for large‑scale corporate customers. The company struggled to clearly separate and resource both.

By 2018 the split became evident. Docker could no longer satisfy both the demanding open‑source community and the rigorous needs of enterprise customers.

In March 2018 Hykes left the company he founded, writing in a blog post, “As a founder, I have mixed emotions. Every entrepreneur hopes their company succeeds after they’re gone. My day finally arrived, but not in the way I imagined.”

He later reflected, “I realized I no longer belonged there and could not contribute constructively, so I left.”

Facing mounting financial pressure, Docker rotated CEOs: Ben Golub handed over to former SAP CEO Steve Singh in May 2017, and in June 2019 former Hortonworks CEO Rob Bearden took the reins from Singh.

What Is Docker’s Position Today?

Backed by initial investors Insight Venture Partners and Benchmark Capital with $35 million in cash, Docker’s “remnant” now survives under the leadership of seven‑year veteran Scott Johnston, sustained by Docker Engine, Docker Hub, and Docker Desktop.

Johnston told an interview, “We are laser‑focused on real developer needs, aiming to bring Docker back to its roots. Our three core pillars are customer‑centricity, unified product planning, and an ecosystem‑friendly business model.”

He is determined not to repeat past mistakes and to provide value for core software developers inside enterprises.

Johnston believes “Docker 2.0” growth lies in offering secure, verified images, new developer tools, and trusted content, while powering emerging compute models such as serverless, machine learning, and IoT workloads.

Docker remains the industry‑standard container runtime; Docker Desktop boasts roughly three million installations, and 49 % of respondents in Stack Overflow’s 2021 Developer Survey reported frequent use of Docker Desktop.

Nevertheless, optimism is limited. Stinemates bluntly said, “I question whether Docker truly exists anymore. From a career perspective, it’s sad – I haven’t seen a company as exciting and spark‑filled as Docker in years.”

Hykes concluded, “Fairly speaking, Docker has not yet turned its potential into real value… but I’m glad Docker has another chance to build a business after all these hardships. Its brand and foundation are still strong.”