BiModal IT key to exploiting data for business insights

The combination of an accelerating rate if data generation and drastic changes in IT infrastructure, software and development environments leaves most organizations unable to keep up with, much less exploit new technologies while also maintaining critical IT business processes. Indeed, an organization’s IT legacy can leave it at a disadvantage to nimble, disruptive new competitors building digital businesses around cloud services and using agile, DevOps development processes. One approach IT organizations can take to resolve the paradox of trying to be both a business utility and technology innovator is through what’s popularly known as Bimodal IT.

Coined by Gartner, the term is often interpreted as a proscriptive edict for how IT should operate, however as Gartner Managing VP Sheila Childs described in a presentation at Veritas Vision 2016, it’s actually a reflection of how many progressive organizations have already organized themselves to foster innovation without disturbing ongoing operations. Specifically, bimodal entails splitting IT into a so-called Mode 1 organization focused on legacy, mission-critical, inherently conservative activities and a Mode 2 counterpart working with emergent, experimental and risk-tolerant projects. Indeed, Gartner surveys show that 40% of CIOs claim to have bimodal organizations with most of the rest expecting to within a couple of years. Still, separate doesn’t mean equal and given their criticality to ongoing business operations, 75% of IT spending in bimodal organizations typically goes to mode 1 activities.

Given their importance to business operations, mode 1 organizations and practitioners focus on reliability, security, stringent IT controls, heavy governance with detailed planning and approval processes using linear, waterfall development methodologies. In contrast, mode 2 organizations value speed and agility, using small teams, DevOps processes, new technologies and cloud services and minimal process overhead, preferring to fail fast and adapt rather than overplan. Mode 2 is epitomized by Jeff Bezos’ rule for ‘two pizza teams’: if you can’t feed a team with two pizzas, it’s too large.

The bimodal concept has been around for several years and often is applied by using mode 2 organizations to build new digital business products and services. Indeed, the digitalization of business fueled by the explosion of data and catalyzed by the nexus of cloud services, mobile devices, social networks and big data analysis software is reshaping most industries and government agencies. According to Childs, 16% of business revenue now comes from digital products and services, a figure Gartner predicts will hit 37% in five years. Likewise, 42% of public sector processes are now digitalized, on the way to 77% by 2021.

Although often associated with agile, mode 2 DevOps teams using public cloud services to build new products, Childs notes that bimodal is increasing appropriate for data-related activities as organizations seek novel ways to exploit the flood of business information that often doubles in volume every couple years. According to Gartner surveys, the top priority for CIOs for three years running is business intelligence and analytics. She notes that data storage and analysis tools are becoming more sophisticated through the addition of features like enterprise content management systems, distributed big data repositories and processing systems like Hadoop and cloud data storage services for file sharing, archiving and disaster recovery. Introducing these new capabilities into an organization without disrupting existing mission critical processes is best done via a bimodal organization.

Given the learning curve for big data and analytics software like Hadoop, Spark, R, Kafka and Storm and the complexity of building and deploying the distributed systems to run them, Mode 2 teams are the ideal vehicle for introducing data analytics to an organization, building aggregated data repositories and experimenting with new analysis techniques. Likewise, cloud infrastructure is the perfect target for big data implementations due their inherent scalability and ease of software orchestration. Gartner estimates that by 2018, one-third of businesses will be monetizing their information: namely, treating data as a business asset. Says Childs, “information must be the second language for IT.” Bimodal organizations using cloud-based services are a great way of building data analysis capability without risking mission critical IT processes.