The idea of BiModal or Two-Speed IT was born as a quick-fix for large enterprises to behave like startups and bring in agility to their business ideas without getting bogged down by the governance processes of their existing mission-critical IT. As Mode 2 (the sprint mode) gains higher momentum and enterprises learn their tricks to manage both agility and stability together, the line between the two modes starts to blur and enterprises begin to explore the possibility of going Mode 2 for the entire estate. There are multiple factors driving this change and we have started seeing the symptoms already.
Does the "cost of agility" justify its benefit?
BiModal IT would mean two teams, two processes, two sets of resources and two different ways of thinking. Achieving agility thus comes at a cost. Additionally, re-training people and re-aligning resources to achieve agile mode is also expensive. Either ways, there is a cost attached in achieving agility and enterprises need to assess whether it justifies the benefit. Enterprises need to plan and space it out but in the long term - aim at having one team, one common shared set of resources, and one way of thinking.
Are Mode 1 people feeling left out?
Today’s practical dynamics within IT organizations suggest that Mode 2 people are "trendy", "ahead in the curve' and doing interesting stuff that contributes to immediate business returns. There are two ways in which Mode 1 people are reacting. One, they are good with what they are doing and don't really care on how Mode 2 people are treated - "as long as I keep my systems' lights on, I continue to have my job. I'll care about learning something new when the existence of my systems (Mode 1 systems) get challenged". And two, they feel left out. They feel inferior. They aspire to be in Mode 2 - Innovators and Operators - how do they sound? In reality that's the emotion of BiModal IT.
Is leaving Mode 1 people in their comfort zones a good idea?
In the long run, most IT systems would be running in Mode 2. MicroService Architectures and technologies like Docker and its ecosystem are presenting a completely new way of building IT systems. Even newcomers out of college would directly leapfrog into building systems of this nature. Talent and skill set for Mode 1 will get irrelevant or the number of people doing Mode 1 work will continue to shrink. Would it be a good idea to leave them as they are and not think about transforming their skillset and prepare them for Mode 2?
Is Mode 2 completely independent and isolated, and escapes the IT complexity?
No. In reality, Mode 2 systems speak to Mode 1 systems in some form or other. There exists a cost to build interfaces to integrate with Mode 1 systems. It complicates the technology footprint further. Indirectly, enterprises would be growing their technology debt.
Governance in Mode 2
“Fail fast, Recover fast” mode assumes that new technologies are brought in, tested fast, and discarded if failed. In this process, enterprises are bound to build a set of technology inventory. Some may discard those technology components that are not useful and some may retain a few for use by some other BU. For example, in the process of evaluating and using emerging technologies for speed, one of the large banking organization had accumulated a huge list of Open Source components in their inventory. Mode 2 approach potentially negates the idea of standardization and the IT inventory would over time, have lot of elements that remain unattended and unutilized. Bi-Modal IT was a great idea for enterprises to have a quick start into the fast and furious world. However, sooner or later enterprises need to look at agility and stability holistically. Else, legacy will grow to be more expensive in the long run.