Konstantin Sutyagin - Fotolia
It's easy to think of a bimodal IT strategy as little more than a way to segregate vastly different parts of an IT organization into two mutually exclusive sections. Mode 1 is for the legacy, "keep the lights on" activities while Mode 2 is where the exciting, new cloud work happens. This is a simplistic view on several levels: It understates the amount of innovation and service improvement that needs to happen in Mode 1, business-critical systems; it creates a false dichotomy concerning cloud usage within IT; and it romanticizes the nature of Mode 2 work. In reality, most Mode 1 applications aren't on life support and should be rearchitected for the cloud to improve efficiency, scalability and resilience, all of which leaves many places for important, rewarding work. Still, Mode 2 is where IT's experimentation and risk-taking should occur and that's our focus here: the Mode 2 opportunities for IT rejuvenation, modernization and innovation.
The impetus for a bimodal IT strategy is to create space within a traditionally conservative, methodical, risk-averse and, yes, bureaucratic IT organization for some disruptive, experimental, startup-like activities. However, the goal isn't to attract a bunch of scruffy, young, 10x developers that can code mobile apps and cloud back ends while binge-watching Netflix, although that might be a beneficial side effect. Instead, Mode 2 provides the structure, or lack thereof, for IT and developers to learn, inculcate and perfect the behaviors and technologies required to attack fast-changing prospects for digital business through new applications and services.
The following are the key areas of Mode 2 development and experimentation.
Cloud application and infrastructure design
All of IT is becoming service-oriented, virtualized and cloud-like; however, bimodal IT Mode 2 is where organizations have the freedom to start with a blank slate, unencumbered by legacy requirements, and build cloud-native applications. This is a far cry from deploying a conventional, monolithic code on some AWS instances with the data sitting in S3; rather, it requires designing systems around high-level platform services and making them more granular, distributed, scalable and stateless.
Popular cloud services providers, including AWS, Microsoft and Google, keep adding to their portfolio of application services with platform-as-a-service (PaaS) features like mobile back ends that include data synchronization, user profile management and database integration, or advanced data services like predictive analytics using Hadoop/MapReduce, programmable data pipelines and real-time stream processing. These platform services obviate the need for DIY implementation of complex software; however, using them properly requires thinking about application design in a different way, like decomposing tasks into microservices instead of macro systems, and using stateless, RESTful APIs instead of persistent network connections.
Agile development processes and DevOps collaboration
Today's application lifecycle is measured in weeks, not years, meaning neither customers nor employees have the patience for a lengthy software development process. Organizations that are too slow to capitalize on an emerging digital business opportunity lose out to competitors that move quickly. But such a quick process requires using Agile development practices, fostering close cooperation between developers and IT operations, heavily instrumenting applications to measure performance, feature usage and errors, and employing continuous delivery processes that facilitate a steady stream of bug fixes and feature enhancements.
These result in significant cultural, governance and process changes that would disrupt a well-functioning Mode 1 organization but are perfect for Mode 2 experimentation. Indeed, Mode 2 is the perfect place to break down silos within IT and between IT and business units by building multidisciplinary, project-focused teams that are unencumbered by sclerotic bureaucracy and empowered to make quick decisions.
Digital business platforms
Mode 2 is a place to incorporate entirely new application platforms. Today's most compelling clients are mobile, yet mobile development is sufficiently different that the learning curve is best tackled on new projects that don't require integrating legacy code or achieving mission-critical reliability. The same could be said for other projects requiring entirely new platforms like IoT and big data. By combining smart devices with back-end data services and compelling client applications, IoT promises to unleash new revenue-producing and cost-saving services. Likewise, data-driven analytics and dashboards improve executive decision making and expose trends, preferences and behaviors that can increase customer satisfaction and inspire new products and services.
Ideas for new business are fraught with risk, with unknown prospects of success. The key to exploiting them is to minimize the cost of failure, or more rightly, learning, by attacking them quickly, iterating aggressively as new information comes in and being unafraid to kill a failing project and move on. Mode 2 IT provides a structure where this can happen without threatening the mission-critical IT activities that keep a business running.
If you have questions or thoughts around the value of a bimodal IT strategy, email Sue Troy at [email protected].
Corporate culture and the IT identity crisis
Getting ready for the 10 biggest IT trends headed your way
Bimodal IT discussion makes its way into IT channel