Adopting DevOps: How To Make It Happen

Problem

I was engaged by a large bank to assess its DevOps strategy. They've been working on it for two years and asked me to see if they were on the right track. Although they had spent a considerable amount of money, deployed tools and made organizational changes, they had yet to see the benefits they were expecting.

Actions taken

I looked at all the paperwork, went back through everything they've done and interviewed the teams involved. I created a model of what was missing, where the gaps were, what things they should be focusing on and where they should go next. The report I completed identified far more gaps than expected and my initial two months-long engagement turned into becoming a trusted advisor to the executive team in charge of the transformation to guide them in closing the gaps.

To accelerate the DevOps program I started by setting up and creating particular metrics for measuring what the organization was doing and measuring the improvement over time. Also, I helped find and put together frameworks and architecture for the management of the centralized tools, processes, and standards so that we could get out of the way up the delivery team and create enabled capabilities for them. Essentially, I helped with the mindset change within the teams that were largely centralized as often happens in the banks. There was a lot of work done helping the central team understand that their role wasn't to be involved in the delivery of every little piece but to create capabilities that would enable other teams to solve problems. It involved a significant amount of work in the security and governance space bringing security and compliance to the table and having them understand what are the things that we need to do in order to be able to push code all the way into production.

To help with this last part of creating alignment between security, compliance, delivery and the platform team, I put up a straw man and brought all the teams to the table to help them pick apart what’s missing and what else needs to be in there. They could then use that as a way of bringing teams to the same way of thinking about the delivery.

The basic practices would remain the same for a different type of organization in different industries. The nuances are really around what is required from a regulatory and organizational standards perspective. For example, in healthcare data protection is paramount and you need to make sure the management of that data is handled appropriately. These practices are independent of the type of organization as it is more about the taxonomy, the language that we use to talk about it.

All of these actions and the work we did together over the course of a year resulted in wider adoption of the capabilities within the bank. This, in turn, allowed us to track an average reduction in lead time for a change of 3 months to 2 weeks. Continually developing these capabilities has allowed the bank to be able to rapidly respond to changes, including pushing out a new loan processing platform for the COVID-19 pandemic in a matter of weeks.

Lessons learned

• Building out high-performance teams requires creating a common alignment and ensuring a common vision across the organization as well as continually communicating key messages into the different groups.
• Successful transformation requires the right people and it is essential to identify the skills you need and grow your teams correctly.
• A framework tailored to the organization that would entail its own taxonomy will smooth out the communication across all the different groups and ensure clarity.
• Understanding how you deliver value and putting the right people, practices, and tools into place to accelerate it allows you to do amazing things.