In order for organizations and teams to move toward DevOps maturity, they need to understand their DevOps capabilities. An Assessment of DevOps Capabilities determines a baseline DevOps state, measures and accelerates continuous improvement.
When we designed our Assessment of DevOps Capabilities (ADOC), the ambassador cohort brought a ton of ideas to the table around the high-level topics based on their experience performing this work for years. We talked about how important it was to cover the three aspects we have classically referred to in our industry; people over process over tools and many referred to the CALMS (culture, automation, lean, measurement and sharing) model widely known in the DevOps world and originally coined by Willis, Edwards and Humble.
Our chief of research, Eveline Oehrlich, who led the 40-strong ambassador cohort who crowdsourced the model along with myself, had been conducting research for our Upskilling Report for several years. We looked at the models she’d been using there to drive comprehensive coverage of the different dimensions of DevOps. They aligned with the other models we were considering, resonated with the team and we all agreed that they gave us the best chance of ensuring the organizations that would use the assessment considered all the elements of these complex and sophisticated ways of working we call DevOps.
The 5 Dimensions of the Assessment of DevOps Capabilities (ADOC)
- Human aspects
- Process and frameworks
- Functional composition
- Intelligent automation
- Technology ecosystem
Our next job was to define the topics in each dimension and then to define Likert statements for each of the topics. These statements allow for data to be collected that provides teams with insights into where their capabilities are strong and should be shared, and where we recommend they experiment with improvements. Let’s look at each dimension in more detail:
1. Human Aspects Dimension
At DevOps Institute, we’re all about the human aspects. We know that culture is the biggest challenge for most organizations on a DevOps journey. Helping individuals and teams to understand and change behavior is a core purpose of ours.
In this dimension, we address psychological safety and a culture of trust. We explore how leadership behaves, using transformational leadership as a model. We ask questions that explore how the team feels the organization is performing in terms of alignment, accountability, autonomy, empowerment and if and how they feel supported. We look at dynamic learning and constructive collaboration.
Some topics focus on The Three Ways; we consider flow through the lens of customer focus and value stream thinking. We ask about how feedback is generated, received and reviewed in retrospectives. We explore how innovation and experimentation happens. There are some statements that help the teams learn about organizational design and team topologies. And we also include happiness and joy at work, and diversity and inclusion topics.
2. Process and Frameworks Dimension
Many years ago, I borrowed our esteemed leader Jayne’s phrase, “DevOps is the happy, polygamous marriage of ITSM, lean and agile.” Naturally, these frameworks are in this dimension – specifically, how they connect with DevOps activities.
We also look at how the teams are experiencing scaling their activities and whether it’s working for them. We want teams to understand how to achieve high performance in continuous compliance, so we help them explore how they incorporate governance, risk and compliance (GRC) in all that they do.
We take another look at value stream ways of working from a process perspective, and how well teams are transitioning from project to product. We look at a couple of frameworks very closely related to DevOps; site reliability engineering (SRE) and DevSecOps.
3. Functional Composition Dimension
In this dimension, we are looking at all the steps from ideation to value realization.
We start with looking at how the teams collaborate and then where and how ideas are conceived, nurtured and refined through portfolio, backlog and product management and ownership.
We look at change, whether the teams are using lower-level change advisory/approval board (CAB) patterns, and how far they have progressed towards peer-review. We explore with them how loosely coupled their architecture has transitioned, and how they build, develop and integrate. We ask them to rate their capabilities around test and validation, and deploy and release.
Other capabilities we explore include operations and support, security, infosec and cybersecurity – helping them understand the patterns that create the safest environments to reduce business risk. We look at data lakes and how they manage data generally; its statefulness has always caused problems with DevOps at speed; for remediation and test data in particular. And we help them understand how to build in reliability and recovery patterns to counteract the risks associated with high-velocity working.
4. Intelligent Automation Dimension
This dimension is all about the DevOps toolchain; from artifact management and source control through CI/CD and into live.
We also look at how tools are used to create a unified backlog, and how service and support happens post-release. We explore environment management and how well the teams are achieving continuous compliance and what this means for customer delight.
We consider some of the key tool categories that help obtain feedback for improving customer experience; observability and monitoring and tie it all together by exploring how the team is using value stream management to gain visibility, traceability and insights into the end-to-end flow through the DevOps toolchain.
5. Technology Ecosystem
Our final dimension is designed to help teams understand how the infrastructure that surrounds their product, service or value stream influences their ability to work better, faster, safer and happier. This means understanding their capabilities for cloud, elasticity, serverless, containers, microservices and APIs.
They rate their attitude toward open source, and how inner source does or doesn’t help them work better. It requires investigation into how they construct their route-to-live and how they architect the toolchain within these contexts.
We look at security elements, such as secrets management, and some of the technology concepts that are currently emerging and that promise to drive us further forward; robotic process automation (RPA), blockchain and virtual and augmented reality.
What to Expect From the Assessment of DevOps Capabilities
This five-dimensional model allows teams and organizations to assess their current state on their DevOps journey. The data collected allows teams to self-discover improvements and accelerate progress, with the guidance of a qualified ADOC consulting partner. Organizations can compare team capabilities across their enterprise, pinpoint high levels of capability and turn local discoveries into global improvements. ADOC is a tool to measure and accelerate a DevOps journey.