We still hear a lot about mode-1 and mode-2 in enterprise IT. Like many other large enterprises, you might have adopted the terminology introduced by Gartner and use it every day to shape the way you deliver and operate technology. However, I have always been uncomfortable with the idea of bimodal, as it ignores architectural reality of tightly coupled systems that we have today and the cultural factors. There is also the cultural dimension that is emerging a lot in the transformations that I help clients with.
The cultural problem with Gartner’s bimodal IT view is that it relegates legacy systems such as mainframe (Z series) or I-series to mode 1 and discourages the new generation of professionals from working on these platforms. If you are in mode 2, you get access to the latest tools and training opportunities, and if you are not, then you are relegated to the end of the queue!
What do you think? Is bimodal creating two classes of people? Or am I being too harsh in my judgment?
In a recent article by BMW CIO Klaus Straub, he went on record and said bimodal IT doesn’t work due to cultural issues. As he said in the article, “Until September of last year, I also had a bimodal world in my head—but then the concept was incomprehensible to me. However, it is not suitable for permanently structuring an IT department.”
BMW is not alone. Other enterprises are joining the voices, such as Peter Jacobs, the CIO of ING Bank Netherlands, who also has expressed similar views on bi-modal. To quote Jacobs: “I would rather work agile at my core bank system than at the channels.”
The full interview of Straub is below. Unfortunately, the article is in Dutch, so you will need to translate the article to read it: https://www.cio.de/a/amp/bmw-cio-haelt-bimodal-it-fuer-einen-irrweg,3562374
A few other relevant and interesting articles are below:
- https://www.cio.com/article/3193793/cio-role/in-the-digital-era-cios-not-buying-this-bimodal-crap.html
- https://www.cio.com/article/3191810/it-industry/why-bi-modal-it-doesnt-really-work.html
