Tracie Berardi, Program Manager, CISQ
Prabha Anand, Assistant Vice President of Excellence Delivery at Cognizant, was guest speaker at the CISQ Forum: Modern Metrics for Productivity and Continuous Improvement on June 18, 2019 during the OMG Technical Meeting in Amsterdam, the Netherlands. Her session, "Measuring Quality and Velocity in DevOps - A Practitioner's View," inspired me to write a blog summarizing key points as her story will help teams on their DevOps journey refine their approach and definition of success.
Photo taken by me at CISQ's event in Amsterdam
Prabha explained Cognizant's DevOps journey started four years ago and has enabled the company to achieve faster time to market, lower total cost of ownership, and reduced IT costs. According to Forrester Research, 63% of organizations have implemented DevOps and that number is rising. Dr. Bill Curtis, Founding Executive Director of CISQ, has also observed Indian IT has adopted DevOps very fast. Cognizant tackled multiple issues during DevOps industrialization, such as lack of common definition, cultural issues, technology spread, organizational resistance, divergent tools, and architectural differences. Her session clearly outlined Cognizant's successful approach towards DevOps maturity and how they measure performance.
Cognizant focuses on DevOps maturity in three areas:
- Engineering excellence
Engineering excellence includes end-to-end DevOps platforms along with tool installation recipes, assets and add-ons.
Process includes a DevOps reference architecture, patterns, playbooks, and KPIs.
People includes DevOps training and certification, clarity of roles, and effective communication.
Prabha explains that process is integral to shifting to new work patterns in DevOps. Cognizant developed a DevOps maturity model with five vectors and a five point scale to assess a team's maturity. Levels 1-5 in the DevOps maturity model include Embracing, Adapting, Established, Measured, and Effective. This framework is a tool to identify gaps in process, tools or culture and helps Cognizant measure KPIs to benchmark performance and maintain best-in-class standards. The KPIs in the DevOps maturity model are tied to business outcomes including:
- Product or Service Quality
- Cost Savings
- Wait Time Reduction
The baseline for each KPI is set to Best-in-class, Performing, Functional and Inception.
This slide includes implementation best practices based on Cognizant's DevOps journey.
Prabha concluded the session with a summary of Cognizant's results on their DevOps journey:
- Velocity: Performance improved from 2017 to 2018 and is baselined better than the industry average.
- Product/service quality: Post-production defects were zero in 2017 and 2018, which is 1.5x better than the industry average.
- Time to market: Performance is better than industry in increased automation and reduced wait time.
How is your organization looking at DevOps maturity? Please comment below or head over to CISQ's LinkedIn group to start a discussion.