As you’re aware, speed of software development is critical in the financial services sector, but you have an added level of risk and complexity in your users’ data that many industries don’t have. The margin for error is very slim, and yet the systems are growing increasingly complex as modern banking moves more and more online and mobile due to shifting customer preferences. And despite the best efforts of the industry, some banks suffer customer-facing errors five times a week. These glitches can range from poor customer experience and in turn poor customer retention due to customers not able to access their money, or the errors can be much more catastrophic, as seen in the famous case of Knight Capital, who lost nearly have a billion in value in a matter of minutes due to an error in their code.
Firms that fall behind the competition in regularly bringing out new, innovative features that improve the customer experience are likely to miss out on tremendous opportunities for business growth in the hyper-competitive personal banking market. But rushing features out that aren’t ready or adequately validated by customer risks losing current customers due to dissatisfaction. What are firms to do?
Enter feature flags. You’re likely no stranger to the buzzwords like DevOps, CI/CD, and agile to speed up software delivery and improve code quality, but scaling feature flag (also known as feature toggles or flippers) usage and process at the enterprise level is becoming an increasingly common practice to easily speed up delivery while also reducing risk. And even better, they provide the ability to easily test out new features with end customers in production to ensure that the feature will be well received by the user base at full release.
What are feature flags?
At the simplest level, they are an “if” statement in your code that allow a feature to be remotely updated in runtime without having to redeploy the code.
Feature Flags provide a variety of benefits to financial companies looking to amplify innovation while reducing risk:
Reduce Risk by testing in production
Feature flags allow you to decouple the technical step of deploying code from the business decision of releasing a feature. This means that you can deploy a feature and gradually release it to internal or external users in production to ensure that it performs as intended in production environments before releasing to the entire user base. No more risky “big bang” releases to all users! And if the feature doesn’t perform as intended? Immediately switch it off with a “kill switch” to limit the blast radius of the bug. No emergency patching and redeploying! The end result is a confidence in the performance of your features not as easily or as quickly possible using traditional CI/CD practices, and assurance that your company will not be in the headlines next for glitches that negatively affect customers.
Only create features you know users will love
By testing in production, you can gradually release features to real, select end customers either by predetermined criteria (such as language, geography, version, etc) or through an automated gradual percentage rollout to all users based on performance criteria established in your analytics tools. By testing features with real customers and analyzing data beforehand, you can be sure that the new feature will be well accepted by your entire user base as a whole when they receive it. This practice is becoming increasingly common, and companies such as Netflix have pioneered the practice for software developers everywhere.
Improve developer productivity
Getting more features released faster with fewer developers required is something that any Director of Engineering would ask for, and it’s doubly true in the current global economic environment. Feature flags improve developer productivity in a wide variety of ways, such as allowing them to eliminate “merge hell” commonly found in CI processes as now feature branches can be flagged and easily merged back into the trunk. Further they remove developers from the release process as with feature releases decoupled from deployment they no longer have to work lock-step with the business’ schedule, allowing them to dive back into the development of the next feature. And flags provide a level of release consistency: no longer will a release be held up by a specific slow or buggy feature when everything else is ready to go: simply flag that feature off and release the rest on-schedule.
Feature Flags are more common than you think, but not at the scale financial institutions need
Your developers may already be using feature flags today, but this could be a problem when you try to scale. Feature flags themselves are not brand new technology. After all, they are just a type of if statement in your code. But the enterprise management, scaling, and integration of feature flags across CI/CD toolchains are an emerging practice and one that should be handled with careful consideration and care.
Back to the Black Knight example from above, where they lost $440M in value in 45 minutes. Did you know the root cause of their error was poor feature flag management? It’s true, read all about it here. While this may be an extreme example, it highlights the need for high amounts of visibility, governance, and auditability in feature flag management as you scale the practice across your company. Most “homegrown” systems can’t keep up with the needs of a large financial institution as it scales in areas such as security, infrastructure, and visibility across the organization. Items all financial companies need more than perhaps any other vertical. CloudBees has been trusted by 42% financial institutions in the Fortune 100 to help them controller their software delivery, and we’re bringing this expertise to the world of enterprise feature flag management.
CloudBees Feature Management is the most developer-friendly and secure feature flag management platform today. We are the industry’s only self-managed feature flag management platform available to financial companies wanting to keep their user data secure in on-premise environments, and have a SaaS-version as well. Our unique architecture ensures no personal identifiable information (PII) from your end customers can be collected on our platform, ensuring user data security. And better yet, we offer this solution with tight integrations between our other CloudBees tools such as CloudBees CD and CloudBees SDM for even greater overall value across your CI/CD toolchain.
Use CloudBees Feature Management to Give Your Company an Edge
The demands on IT in financial services will only continue to grow, and the risk associated with changing technology and increased complexity across devices and systems will not go away anytime soon. CloudBees Feature Management offers the most secure, enterprise-ready feature flag platform today that will allow you to continue innovating at high velocity while also decreasing risk. And, it’s a feature flag platform from the company that is already widely trusted by the world’s leading financial services institutions.
Give CloudBees Feature Management a try to see the difference they can make in your development today, with the following resources:
Stay up to date
We'll never share your email address and you can opt out at any time, we promise.