This article draws on reporting originally published at InformationWeek, citing findings from the DORA State of DevOps 2018 report and a LogicMonitor cloud survey.
Cloud Infrastructure as a DevOps Accelerator
The relationship between cloud adoption and software delivery performance is well-established, but the depth of that relationship is more nuanced than simply “move to the cloud, ship faster.” What actually matters is how completely an organization implements cloud capabilities.
According to a LogicMonitor survey cited in the original reporting, 58% of respondents said DevOps was somewhat or significantly driving public cloud engagement at their company — ranking below only digital transformation (63%) and IT agility (62%). This alignment makes sense: teams running customer-facing applications in the cloud need rapid feedback loops, elastic capacity, and safe experimentation environments — all hallmarks of mature DevOps practice.
The NIST Five Characteristics and Their Impact on SDO Performance
DORA’s State of DevOps 2018 report quantified just how much cloud maturity matters. Using the five essential cloud characteristics defined by NIST, the report found that teams whose organizations met all five characteristics were 23 times more likely to fall into the elite tier of Software Delivery and Operations (SDO) performance.
The five NIST cloud characteristics are:
- On-demand self-service — developers can provision compute resources without requiring human intervention from IT or operations staff.
- Broad network access — resources are available over the network and accessible from standard client devices.
- Resource pooling — computing resources are pooled to serve multiple consumers dynamically, with location abstracted from the user.
- Rapid elasticity — capacity can be scaled up or down quickly, appearing unlimited to the consumer.
- Measured service — resource usage is monitored, controlled, and reported, providing transparency for both provider and consumer.
Despite this 23x performance correlation, only 22% of survey respondents reported using all five characteristics — a figure that surprised practitioners in the field. The gap between what is known to be effective and what is actually implemented remains stubbornly wide.
Cloud Adoption Alone Is Not Enough
Lynda Stadtmueller, Vice President of Cloud Computing Services at Stratecast/Frost & Sullivan, was quoted in the original reporting as connecting cloud strategy to broader business goals: speed to market, agility, innovation, and always-on availability. But she also cautioned that technology strategy alone doesn’t close the loop.
Stadtmueller noted that the fast-paced environment enabled by cloud — alongside Platform as a Service (PaaS), Infrastructure as Code, cloud-native architectures built on microservices and APIs, and open-source tooling — creates a more engaging and creative environment for developers and engineers. That environment, in turn, attracts and retains high performers. In her framing, it becomes a self-reinforcing cycle: companies that genuinely embrace cloud characteristics are better positioned to attract and develop elite talent.
Pipelines Are the Prerequisite for Microservices to Deliver Value
Mandy Hubbard, software engineer and QA architect at CS Disco, offered a pointed observation about the limits of architectural decisions without supporting automation. Speaking from experience with DevOps, cloud, and continuous delivery, Hubbard argued that microservices architectures fail to deliver on their promises unless a robust CI/CD pipeline is in place.
“All of those promises we have been given if we move to microservices are all completely dependent on having a continuous integration and delivery pipeline. It doesn’t really matter if you’ve broken down your huge application into multiple tiny services if you don’t have an automatic way to build, test, and deploy those services.”
— Mandy Hubbard, Software Engineer and QA Architect, CS Disco
This is a practical constraint that organizations frequently underestimate. Decomposing a monolith into microservices increases deployment surface area dramatically. Without automated build, test, and deploy workflows, teams end up with more complexity and slower delivery — the opposite of the intended outcome.
Removing Gatekeepers: The Access and Trust Problem
Hubbard identified a cultural pattern that limits even technically sophisticated organizations: placing gatekeepers between developers and infrastructure. Many teams have mature delivery pipelines but still require a manual “push to live” handoff — often controlled by a separate operations team or manager.
This isn’t a technical limitation. It’s an organizational one. Solving it requires building a culture of trust in which managers and executives are confident that developers and operators will use their access responsibly. Without that confidence, automation stalls at the most critical stage.
What DORA Says About Autonomy, Trust, and Psychological Safety
The DORA 2018 report reinforces Hubbard’s point with research findings on team dynamics. The report found that when leaders give teams meaningful autonomy over their work, it generates feelings of trust and psychological safety — and those feelings directly predict high-performing team behavior.
“We found that when leaders give their teams autonomy in their work it leads to feelings of trust and voice… Trust and voice, in turn, positively affect organizational culture.”
— DORA State of DevOps 2018 Report
The DORA report also references a Google study identifying “psychological safety” — the feeling that it is safe to take risks within a team — as the single most important factor in predicting team performance. This aligns with years of organizational behavior research, and it applies directly to DevOps contexts: teams that fear blame for failures will hide problems, avoid experimentation, and slow down.
Practical Takeaways for DevOps Teams
The combined findings from DORA, practitioner experience, and industry analysis point to a consistent set of priorities:
- Audit your cloud implementation against all five NIST characteristics. Partial adoption limits the compounding benefits. On-demand self-service and rapid elasticity are especially high-leverage for development workflows.
- Build the CI/CD pipeline before — or alongside — migrating to microservices. Architectural decomposition without automation increases operational burden rather than reducing it.
- Identify where manual gatekeeping is blocking automated delivery. Map your pipeline end-to-end and find the handoff points that reintroduce human latency unnecessarily.
- Invest in the cultural conditions for trust. Blameless post-mortems, developer access to production tooling, and leadership that explicitly signals confidence in its teams are all mechanisms for building the psychological safety that correlates with elite performance.
- Treat talent strategy as part of infrastructure strategy. The cloud environment you build determines what kind of engineers you attract and retain.
Bottom Line
Fully implementing cloud capabilities — not just adopting cloud in name — has a measurable, significant impact on software delivery and operational performance. But infrastructure alone is insufficient. The organizations that reach elite SDO performance combine mature cloud architecture with automated delivery pipelines, genuine developer access to infrastructure, and a leadership culture grounded in autonomy and trust.
Technology enables; culture determines whether that enablement is actually used.