Five years ago, I wrote about 10 companies that were killing it with DevOps. Today, all of those enterprises—including Amazon, Walmart, and Netflix—remain household names, not just for consumers, but in the tech community, too.

Since then, who else has emerged as a clear DevOps leader? These 10 organizations are all pushing the boundaries of DevOps, breaking down barriers between development and operations, transforming their companies and their industries—and continuing to do so during a time of global upheaval.

1. Adidas

In 2015, Adidas released the first of its Yeezy sneakers, which were designed in collaboration with Kanye West. Demand for the original shoe—and its many follow-ups—was instantaneous and enormous, with Adidas selling over 2,600 shoes in the space of five seconds. That was great for business, but it presented a huge challenge for IT, which was tasked with ensuring the Adidas’ tech infrastructure didn’t collapse under the strain.

At the start, Adidas wasn’t ready. The site crashed when new Yeezys were announced. Inside the company, developers moaned that they were helpless to fix things and that it could take up to a week even to get a simple virtual machine spun up.

Faced with upset customers and employees, Adidas underwent a massive transformation that embraced cloud-native architecture, Kubernetes, and DevOps and that involved a wholesale cultural shift. A DevOps Cup competition was instrumental in taking the company’s delivery cycle from as slow as six weeks to as fast as five times a day. Today, you can find Adidas’ DevOps Maturity Framework on GitHub.

2. Capital One

In 2010, Capital One was a stodgy financial firm with a dated waterfall development process that was heavily outsourced, a massive in-house data center, and manual build, deployment, and test operations. By 2018, the firm had completed a multi-year journey that culminated in becoming the first major bank to host its banking application on the public cloud, solidifying its reputation as a DevOps leader.

Since that time, the company has been fully reinvented. Today, Capital One is regarded as a leader in agile development and cloud-native infrastructure. It has an almost ruthless devotion to developing its own tools through open-source software—done almost entirely in-house.

The initial goal of Capital One’s DevOps transformation was speed, but, as a financial services firm, security and governance were equally important. The culture of collaboration created through DevOps has been critical to this move, as has a focus on accountability and “fearlessness.” Capital One’s homegrown DevOps dashboard, Hygieia, is highly regarded—and, of course, it’s open source and available to the public, too.

3. Verizon

As the mobile industry began lurching toward 5G, America’s largest wireless carrier embarked on a digital transformation of its own. Verizon 2.0, as it’s known internally, introduced major changes in the business model to make the company more nimble and flexible. These were key qualities needed to enable its successful technical transformation.

Verizon went “public cloud first” in 2016 and embraced immersive DevOps dojos a year after. From there the company worked to modernize its monolithic architecture, embrace open-source tools, and generally become a more agile organization.

Today (or, at least, pre-COVID 19), Verizon has flipped its dojo concept from a six-week-long offsite training session to one where coaches travel to the developers and embed themselves directly within their business units. This has improved the pace of knowledge transfer and allowed for a wider range of training types.

4. CSG International

CSG isn’t a household name, but as a back-end provider of software and services for the telecom industry, it’s a business that touches just about everyone. It primarily provides billing, payment processing, and customer service.

The company launched a broad DevOps reorganization effort in 2016 and integrated the product management function with DevOps in 2018. Modernizing the technology stack has been a goal of the company ever since.

The problem is one of sheer history: In 1994, CSG was spun off from its original owner, First Data, which itself was founded back in 1982. Shards of First Data’s legacy software have endured in the CSG codebase ever since.

But a monumental effort over the last few years—involving revamping its version-control procedures and embracing automated testing—has successfully transformed the enterprise. After doubling down on DevOps, the company is now seeing dramatically improved efficiency, with an 83% reduction in incidents caused by each release.

5. Disney

Disney has been pursuing DevOps aggressively for nearly a decade, revamping its org chart from functional teams to a matrix organization. The company embedded its engineers and managers across various business units and improved cross-functional communications, both of which have radically enhanced the business’s technical capabilities.

The last few years of those efforts were instrumental in powering one of the biggest streaming video launches in recent history, which concluded earlier this year. Almost overnight, Disney+ became a serious competitor to services from Netflix, Amazon, Hulu, and HBO—adding 10 million subscribers in one day, all during a time of incredible global crisis.

6. Northrop Grumman

When your company’s job is to make billion-dollar military aircraft, it’s safe to say that reliability is at the top of the to-do list. With an estimated tens of millions of lines of code powering a single plane, application development is a serious challenge. It can’t be easily addressed through the traditional tools of agile and DevOps, given the potential hazards of pushing a prospective update to aircraft in the field.

Sure enough, Northrop Grumman’s first experiences with DevOps were rocky, in part because it had to work with all manner of government software tools. But it achieved a breakthrough when it embraced “continu-ish integration” and digital twins, and it is now running 15,000 automated test processes each day.

Northrop frequently presents at conferences with its industry mate Lockheed Martin, which has developed and adopted many of the same operating principles.

7. BMW

Now four years into a major DevOps overhaul, BMW has pursued a digital transformation journey as it has migrated its entire firm from a waterfall methodology to a fully agile one. A stodgy organizational culture revamped to support new ways of working.

In 2019, BMW’s “100% BizDevOps” approach began tying project funding to the business value that the project would potentially provide, while bringing microservices and cloud-first architecture into the mix. With 30,000 servers running 5,000 different applications, BMW found the road to success by starting small and taking baby steps toward change rather than attempting to make a wholesale transformation all at once.

8. Optum

Optum is the sister company of UnitedHealthcare, the largest healthcare company in the world. Optum is heavily focused on data management and analytics in the healthcare space. It faced a problem familiar to many large enterprises: Development wanted to move quickly by implementing new features and applications, while operations wanted stability in the apps it was already managing.

Optum embraced DevOps in earnest in 2014 to break down these silos, and the company’s recent metrics have shown that its efforts have proved very successful, with millions of dollars saved every year by reducing mainframe usage alone.

The amount of data managed by Optum and its parent company presented its greatest challenge, with unstandardized data coming at the organization from dozens of sources, all courtesy of thousands of APIs. A massive standardization and modernization effort has seen the creation of a scalable, centralized data platform that allows information to be seamlessly shared across multiple consumers. The approach was borrowed from successful digital transformations in the retail and financial industries.

That’s been a foundational part of a cultural shift that has been effective at finally getting some 160 tech teams on the same page.

9. Nationwide Mutual Insurance

Nationwide began its digital transformation journey in 2009, and since pulling out the stops in 2017, it’s become a poster child for DevOps evangelism and the business successes it can foment. After a wholesale move to agile and distributed development that took three years, the insurer improved code quality by 50% and reduced downtime by a whopping 70%—despite continuing to actively build for its legacy mainframe platforms.

The key has been a bimodal process that lets different groups iterate at different speeds, allowing Nationwide to “move slow and fast at the same time.” Large transactional systems with many dependencies can percolate a bit more methodically, while more visible, customer-facing services (such as mobile apps) are updated far more frequently. This ultimately lets Nationwide more effectively prioritize its tech efforts while giving staff the freedom to experiment.

10. Starbucks

Making coffee isn’t so easy when you have more than 300,000 people involved in the process. Starbucks’ bold move to DevOps began in 2015, and it learned quickly that it needed to balance the desire to constantly push code forward with the need to ensure that code is appropriately tested and ready for deployment. Organizational tweaks within the company were hugely successful, ultimately reducing cycle time by a massive 74%.

One of the most challenging parts of the move to DevOps at Starbucks was the transition of the company’s product leadership to an environment composed of self-managed teams. Working carefully with employees to understand how jobs and responsibilities change after DevOps is adopted is an essential step to smoothing the path forward, the company learned.

Looking ahead

Digital transformation efforts continue to drive enterprises in all industries toward DevOps and agile methodologies, and the payoff from those efforts hasn’t slowed. Even large and very old businesses are finding ways to integrate DevOps thinking with old ways of doing business, integrating legacy operations with cloud-first and open-source initiatives to create innovative operational structures.

As the Covid-19 pandemic continues to challenge traditional ways of operating, watch for these trends to accelerate in the next few years.

Why DevOps Implementations Fail

DevOps is not a product you install—it is a cultural and operational shift that touches people, processes, and tooling simultaneously. When organizations underestimate this scope, they set themselves up for partial adoption at best and costly failure at worst. Below are five of the most common and consequential mistakes.

1. Overlooking Organizational Change Management

One of the most frequent root causes of DevOps failure is treating it as a purely technical initiative while ignoring the human side of the transformation. Introducing new pipelines, toolchains, and workflows without preparing the people who must use them almost always results in resistance, low adoption, and eventually rollback to old habits.

Effective DevOps adoption requires deliberate change management:

• Communicate the “why” in terms of customer value, not just technical efficiency. Engineers and managers alike are more motivated by outcomes that matter to end users than by abstract methodology labels.
• Give teams adequate time to learn new practices before measuring them against new performance expectations.
• Address cultural transformation explicitly. Shared ownership, blameless post-mortems, and cross-functional accountability must be modeled by leadership, not just mandated in policy documents.

Without cultural buy-in, even the best tooling will sit unused or be worked around.

2. Treating Speed as the Primary Goal

Faster delivery is one of the most visible benefits of DevOps, but organizations that optimize exclusively for speed often sacrifice the code quality and reliability that DevOps is equally designed to improve.

The result is a pipeline that ships broken or insecure code quickly—a worse outcome than shipping good code slowly. A balanced DevOps approach treats speed and quality as complementary, not competing:

• Invest in automated testing at every stage of the pipeline—unit, integration, end-to-end, and security scanning.
• Use feature flags and canary deployments to reduce the blast radius of any given release.
• Track quality metrics (defect escape rate, mean time to recovery) alongside throughput metrics (deployment frequency, lead time).

Speed without quality is technical debt accumulation at scale.

3. Continuing to Rely on Manual Processes

Automation is foundational to DevOps. Manual build, test, deployment, and infrastructure management processes introduce inconsistency, slow feedback loops, and human error—the exact problems DevOps is designed to eliminate.

Many teams are reluctant to automate because it feels like ceding control. In practice, well-designed automation enforces consistency and makes processes auditable and repeatable in ways that manual steps never can be.

Key areas where automation delivers immediate value:

• CI/CD pipelines: Automate build, test, and deployment on every code commit.
• Infrastructure as Code (IaC): Use tools like Terraform or Ansible to provision and manage infrastructure consistently across environments.
• Configuration management: Eliminate environment drift by codifying all configuration.
• Security scanning: Integrate static analysis and dependency vulnerability checks into the pipeline automatically.

Any process that is repeated more than once is a candidate for automation. Manual processes are slower, costlier, and a significant source of production incidents.

4. Neglecting the Right Toolset

Tools do not make DevOps, but DevOps cannot function without the right tools. Attempting a DevOps transformation without investing in an appropriate toolchain leaves teams without the visibility, automation, and collaboration infrastructure they need.

Core categories of tooling to evaluate:

• Version control and collaboration: Git-based workflows (GitHub, GitLab, Bitbucket) are the foundation of modern software delivery.
• CI/CD: Jenkins, GitHub Actions, GitLab CI, CircleCI, and similar tools automate the build-test-deploy pipeline.
• Containerization and orchestration: Docker for packaging applications and Kubernetes for orchestrating them at scale are now de facto standards in cloud-native DevOps.
• Observability: Logging, metrics, and distributed tracing tools (such as Prometheus, Grafana, and OpenTelemetry-compatible backends) are essential for understanding system behavior in production.
• Security: Shift-left security tools integrate vulnerability scanning and policy enforcement into the pipeline itself.

Tool selection should be driven by team needs and existing workflows—not by industry hype. Consolidating too many tools too quickly creates complexity without benefit.

5. Maintaining Silos Instead of Breaking Them Down

Breaking down organizational silos is a central premise of DevOps. When development, operations, security, and QA teams work in isolation with separate goals and separate tooling, the friction and handoff delays that DevOps is designed to remove simply persist.

A common anti-pattern is treating DevOps as something to be implemented within a single team—often an infrastructure or platform team—while leaving development teams unchanged. This produces a “DevOps team” silo, which is itself a contradiction of DevOps principles.

Practical steps to break down silos:

• Train all team members—not just engineers—on DevOps principles, shared ownership, and the value of continuous delivery.
• Establish shared on-call rotations between development and operations so that the teams responsible for building software also experience its production behavior.
• Align team incentives around product outcomes rather than functional output (lines of code written, tickets closed).
• Use cross-functional product teams where developers, QA, and operations collaborate from the start of a feature, not at the end.

Bonus: Micromanagement Undermines Autonomy

DevOps is built on the principle that empowered, autonomous teams produce better outcomes than closely supervised ones. Micromanagement—prescribing exactly which tools to use, monitoring individual work at a granular level, or requiring approval gates at every step—introduces the same friction and bottlenecks that DevOps aims to remove.

Organizations that succeed with DevOps give teams clear goals and constraints, then trust them to find the best path. This is especially important in an automated, pipeline-driven environment where the system itself enforces standards—removing the need for constant human oversight of individual tasks.

Conclusion: Intent Matters as Much as Implementation

DevOps adoption driven by trend-following rather than genuine process improvement is unlikely to deliver meaningful results. The organizations that succeed are those that approach DevOps as a long-term investment in people, culture, and engineering practice—not a one-time tooling upgrade.

Before embarking on or continuing a DevOps transformation, ask honestly:

• Do our teams understand why we are making this change, not just what we are changing?
• Are we measuring quality and reliability alongside speed?
• Have we automated the most error-prone and repetitive parts of our workflow?
• Do we have the tooling infrastructure to support continuous delivery?
• Are development, operations, and security genuinely collaborating, or working in adjacent silos?

Avoiding the mistakes that have derailed other organizations is the most direct path to a DevOps implementation that delivers on its promise.

This article was informed by original reporting from jaxenter.com.