"Fail Fast," They Cried
By Paul Ayers, Enterprise Architect, Oracle
On the surface it makes sense - try something new and innovative with the hope it may lead a business to the promised land of efficiency and/or stakeholder value. In today’s fast paced business world, innovation can be a key differentiator but you need to succeed or fail quickly. But you only want to invest enough time and money to get a point where an informed decision can be made to either drop it and move on (‘fail fast’) or invest further to operationalisation. In the good old days, we used to call this process ‘seed funding’!
We now have the concept of a ‘bimodal’ approach, which was introduced as a way to help support an innovation process while also trying to run the complex business of IT. The Gartner group, which coined the term, defines bimodal (also known as two-speed or pace-layered IT) as: “the practice of managing two separate, coherent modes of IT delivery, one focused on stability and the other on agility”.
The emergence of new technologies like the Cloud has facilitated a greater ability to innovate without touching the sacrosanct on-premise IT assets. But the trouble is it’s in these on-premises assets where the real value lays. Integration is certainly a valid way to explore their precious bounties but there is a level of nervousness usually due to security and other risks.
So, with this perceived level of risk, it has resulted in most of the ‘fail fast’ activities undertaken being focused squarely on user interface and/or customer experience (CX) only. This work generates a lot of buzz as the business sees activity benefit so the performance pressure valve for the IT group is relieved (albeit slightly). This is a positive but, of course, it gets a lot harder once you delve into the complex backend of an enterprise. The stakes are higher as failure inside these systems can result in millions of dollars in lost revenue and/or significant damage to a company’s brand. In some industries, it could even mean loss of life. Failing fast in this context is not an option.
"By putting up a relatively simple activity around improvement to customer experience also cries an element of less risk to a failure-shy Executive"
Like all investments, there is a level of risk versus reward and this balance will be idiosyncratic to each organisation.
Implementing a bimodal approach may also enable an opportunity to sell the concept of managed change to a Board or Executive Committee. By putting up a relatively simple activity around improvement to customer experience also cries an element of less risk to a failure-shy Executive so this may just be the hook that captures their support.
However, the downside of such a risk adverse approach is the level of reward. Fixing a component of customer experience may not be the type of fundamental change needed to improve the company’s overall IT. The bimodal approach can have the side effect of never getting to the bottom of the underlying IT problems so very little will improve. Remember, if IT issues were easy they would have been fixed years ago – but this is complex stuff on so many levels.
On the upside, these activities may include the opportunity to experiment with new ways of working like Agile and/or Devops. These capabilities aren’t easy to operationalise within large and complex legacy environments so taking such an opportunity to assess the readiness for change on a smaller scale will be of great benefit.
But bimodal IT is not a panacea for managing organisational change. I see a number of issues with the model. First, there are the inevitable team issues (which, I might add, have been around since IT became mainstream business so this is not a new thing). IT people, by nature, want experience with the latest tools and techniques. Your staffs not given this opportunity tend to become disgruntled and leave (taking their corporate knowledge with them). Few usually want to stay just to maintain old legacy systems.
You can look at rotations but, in my experience, this doesn’t work very well, as everyone is not made to innovate and, for some, the jump to the newer technology and/or methodologies is just a step too far. Also, if you have a Board or Executive Committee over-sighting your innovation activities, you generally want your best and brightest on the project.
You need to be aware that any type of major change can likely result in some collateral damage. Staff contingency plans are a good idea anyway but when embarking on a major reform, it becomes critical.
Another issue with bimodal is fragmentation of your IT capability. This will inevitably drive costs up over time. It also causes fragmentation of processes, documentation, and standards, all of which can cause pain if not managed. By not monitoring the situation, there will be a growing gap between legacy processes and procedures and the new world order.
But my biggest concern around the bimodal model is focus. It is new and shiny and has got management attraction. This is great news for the new activities but not so much for the other, more mundane, day-to-day stuff. If this more mundane stuff is core IT business (which is highly likely) then the technology issues could get worse to the point where it could be disastrous. I’ve seen many of the organisations sitting on a ticking legacy time bomb.
In summary, all enterprise IT groups are struggling to cope with the shear amount of change and so if bimodal can help an organisation manage it then fill your boots and go for it. But my advice is, just understand the challenges, limitations, and risks of the journey and make sure you are clear around what problem you are solving by this approach and most importantly what you are not.
Founded in 1997, in Redwood Shore, USA, Oracle offers an optimized and fully integrated stack of business hardware and software systems.