This article is based on coverage of the 2019 Accelerate: State of DevOps Report, originally sourced from tprotoday.com, and highlights from a keynote by Nicole Forsgren at the Open Source Summit in San Diego.
Overview
Released on August 22, 2019, the 2019 Accelerate: State of DevOps Report drew on survey data from approximately 31,000 IT professionals worldwide to measure the state of software delivery performance across the industry. Nicole Forsgren, researcher and strategist at Google Cloud and a co-author of the report, presented the highlights at the Open Source Summit in San Diego.
“When I’m talking about DevOps, I’m talking about software development, software delivery, and I’m also talking about availability,” Forsgren said during her keynote.
The Four Key Metrics
The report measures software delivery performance using several core metrics that have become a de facto standard for evaluating DevOps maturity:
- Lead time for changes: The elapsed time from a code commit to that code running in production.
- Deployment frequency: How often a team successfully deploys code to production.
- Time to restore service: How quickly a team can recover from a production incident or outage.
- Change failure rate: The percentage of deployments that cause a failure requiring remediation (referenced in the broader report methodology, though not highlighted in the keynote summary).
For elite-performing organizations, the 2019 report found:
- Lead time for changes was less than one day.
- Deployment frequency reached multiple deployments per day.
- Time to restore service was less than one hour.
Elite Performers Nearly Tripled Year-Over-Year
One of the most notable findings in the 2019 edition was the significant growth in the proportion of teams classified as elite performers. Forsgren highlighted the shift directly during her keynote:
“Check out those elite performers last year; 7% of our survey sample were elite performers, optimizing on speed and stability. This year, check it out—the elite level almost tripled and hit 20%.”
Nicole Forsgren, Google Cloud
At the same time, the share of low performers—teams with a deployment frequency of between once per month and once every six months—declined from 15% to 12%.
“That means that more people are getting better,” Forsgren said, framing the shift as evidence that DevOps practices are maturing across the broader industry, not just at a handful of large technology companies.
Faster Code Is More Stable Code
A counterintuitive but consistent finding in the Accelerate research is that deploying more frequently is associated with greater system stability, not less. Forsgren explained the mechanism in practical terms: when teams delay releases, code accumulates. Larger, less frequent deployments increase the complexity of each change and expand what she calls the “blast radius”—the potential scope of impact when something goes wrong.
“When we introduce delays, we intend to create a more stable system, but we’ve never seen that. It actually creates instability.”
Nicole Forsgren, Google Cloud
This finding argues against the common instinct to slow down release cycles as a risk-reduction measure. Smaller, more frequent deployments reduce the blast radius of any individual change and make it far easier to isolate and fix problems quickly.
How to Improve DevOps Performance
The report identifies several practical levers that organizations can pull to improve their DevOps capabilities. Forsgren discussed two in particular during her keynote.
Addressing Technical Debt
Technical debt—code that is poorly maintained, disorganized, or inadequately documented—is a direct barrier to engineering productivity. Forsgren defined productivity in this context as “the ability to get complex, time-consuming tasks completed with minimal distraction and interruptions,” and was explicit that “technical debt reduces productivity.”
The report points to a loosely coupled architecture combined with continuous monitoring as practical means of keeping technical debt in check. A loosely coupled architecture allows teams to deploy and update individual services independently, reducing the coordination overhead and risk associated with tightly coupled systems.
Building a Culture of Trust and Psychological Safety
Beyond technical practices, organizational culture plays a significant role in DevOps performance. The report found that giving developers visibility into the organization’s goals—and fostering an environment where people feel safe to take risks and raise concerns—has measurable positive effects.
“Trust and psychological safety have a positive impact on software delivery performance, organizational performance and productivity.”
Nicole Forsgren, Google Cloud
This aligns with a long-standing theme in the Accelerate research series: high-performing DevOps teams are not simply a product of better tooling. They are a product of better communication, clearer goals, and a culture that treats experimentation and learning from failure as normal parts of the development process.
Key Takeaways for DevOps Teams
- Measure what matters: Lead time, deployment frequency, and time to restore are proven indicators of DevOps maturity and should be tracked consistently.
- Deploy smaller, deploy more often: Batching changes does not reduce risk—it amplifies it. Smaller, more frequent releases are demonstrably safer.
- Invest in architecture: Loosely coupled systems reduce both the cost of change and the accumulation of technical debt.
- Culture is not a soft metric: Psychological safety and organizational trust have a direct, measurable effect on software delivery performance.
- The industry is improving: The jump from 7% to 20% elite performers in a single year suggests that the practices advocated by the DevOps community are reaching mainstream adoption.