I came across an interesting article by Clint Boulton on CIO.com discussing the recent shift that Capital One has made to the DevOps methodology in order to keep pace with its customers’ evolving preferences for digital banking services. He quoted a statement made by the head of Capital One’s IT department, Rob Alexander, regarding this development.
“Winners in banking are going to be the ones that recognize that technology is really going to play a central role in how consumers want to bank in the future. We’ve got to be great at building software.”
In today’s financial market, many banks are working through slow mainframes and other legacy technologies and fighting through slow bureaucracy that limits innovation. However, instead of upgrading their outdated technology infrastructure, most banks are opting to focus on digital imperatives, resulting in CIO’s losing the budget battle and not allowing them to implement DevOps.
However, Capital One is not one of those banks. Since its founding in 1988, it has boasted less technical debt than its rivals and has prided itself on its low levels of bureaucracy. From the start, the executive team was determined to run the bank more like a technology company than a typical bank. For this reason, it has quickly embraced Agile, open source, cloud, and analytics.
In this vein, when Alexander realized in 2010 the trend of customers demanding more online and mobile options for their banking, he changed from purchasing the commercial, on-premises software packages that are so common among other banks. Instead, he hired engineers and co-located them with a line of business leaders in order to build software, release it early, and update it consistently.
This, he says, was the start of their eventual switch to DevOps. Now, instead of developers building software and then turning it over to the production teams to test it, fix bugs, and push it to production, they work together with them to complete this process using containers and microservices. They push and run apps in Amazon Web Services’ public cloud, which has helped the IT teams focus on their core competencies of building digital products.
What they have realized is that this change has made the development teams own their code more and have an accountability that works as a very powerful incentive for better work.
For this reason and others, researchers are now realizing that DevOps is not just for large cloud providers but is destined to become a mainstream strategy that will be employed by 25% of global organizations. Financial institutions are actually outpacing other industries in the adoption of Continuous Delivery practices. One survey showed that among banking and finance DBAs and managers, more than 70% utilize database automation for software and 64% of those also practice CD for the database.
I believe that those not including the database in their adoption of DevOps are making a mistake. Proper database automation and version control allow for peace of mind that the database can run itself and adapt on the fly, responding to the need for rapid change in the current landscape, while also saving 80-95% of the time it would normally take to manually process changes. Implementing Continuous Delivery best practices for the database, therefore, provides a clear return on investment for financial institutions today.
As consumers’ preferences for digital banking services evolve, banks would do wise to follow Capital One’s example and shifts to DevOps to keep pace with their customers.