As organizations become increasingly global and dynamic, CIOs are working to strengthen the ties between IT and other business units to improve business outcomes. And digitization of business is driving CIOs to organize their IT departments to address the need for both stability and innovation. This is where bimodal IT operations are effective: It helps organizations "keep the lights on," while building the agility required to face challenges and thrive in this digital age.
In this webcast presentation, Kurt Marko, analyst at MarkoInsights, dispels some of the myths surrounding Mode 1 and Mode 2 of the bimodal strategy. Read on to find out about some of the common misconceptions about the two modes, the correlation between modes 1 and 2, and the methods of deployment they use.
Editor's note: The following is a transcript of the third of four excerpts of Marko's webcast presentation on bimodal IT operations. It has been edited for clarity and length.
Kurt Marko: So, here, we'll dispel some of the myths [about bimodal IT].
Bimodal IT operations myth No. 1: Mode 1 is tired and dusty
[You'll hear people say] myth [No.] 1, I say, 'Mode 1 is boring, and it's on life support.' Nothing could be further from the truth. These applications are the lifeblood of most businesses. Some may well be at the tail end of their lifecycle, but most are probably not.
So bimodal ... doesn't imply that one side of your IT is dead and boring, and the other is really dynamic and the go-go place. And this is where a lot of IT professionals feel threatened by the concept. They think that, 'Well, if I end up stuck in this mode, if we got to a bimodal structure and I end up stuck in the Mode 1 side of the house, I'm a dinosaur and it's a career-ending move.' That's not necessarily true. And it's going to take different skill sets and different temperaments for each mode.
Kurt Markoanalyst, MarkoInsights
And, as I say, one area for Mode 1 innovation is looking at the cloud and rearchitecting applications to take some of those cloud-like attributes of maybe resilient, distributed designs, the ability to dynamically scale, or if not dynamically, more easily scale.
So, as I say, if Mode 1 applications are not, they should be rearchitected, or they probably will eventually die. And if that's a strategic choice, that you have certain systems that you know are just, you have something better in mind for the future, that's a perfectly rational and logical choice.
However, if you see a long-term need for a particular system, then the way Gartner terms it is, you want to reinvigorate those core applications. The rest of us might call that, that's just the natural evolution of IT and we see it in every generation. Big business-critical systems go through major changes, but those changes take time. And that's just fine in a Mode 1 environment.
Bimodal IT operations myth No. 2: Mode 2 is the future
The next myth is that Mode 2 is the future, so this is where it's at: Everything, all enterprise apps, they're going to move to the Mode 2. That is not true. The choice between Mode 2 and Mode 1 environment, and it is a choice, really, it hinges on the application characteristics, their maturity, and then some of the development methodologies and management methodologies that those characteristics force or encourage.
So, as I say, there's plenty of innovation left in Mode 1. It's just a different type. So, you're going to be focused on, as I say, bulletproof reliability; security. You can't underestimate that one enough, or overestimate it, I should say. Efficiency. And not necessarily, you won't, these applications, Mode 1, will not necessarily be focused on rapid development of new features or very rapid growth.
Bimodal IT operations myth No. 3: Mode 1 is always on premises, Mode 2 is always cloud
Again, just hitting this point again, but another myth you hear often is Mode 2 is the cloud; Mode 1 is on premises.
An axiom I've heard recently sums it up when it comes to the cloud is: Cloud is not a location, it's more a way of operating IT. Now, Amazon would have you think differently, that the cloud is a location. But, realistically, for most enterprises, the cloud is more a descriptor of how infrastructure is built, and operated, and procured, and consumed, and less about a place. Because cloud-like technologies really just within most enterprises and even within public clouds, most enterprises may well end up using more than one.
As I say, there is a correlation between Mode 1 and Mode 2, and the method of deployment. Mode 2 will favor the cloud for its convenience and rapid scalability. But it's not a hard and fast rule.
What is true, however, is that Mode 2 applications are cloud-native by design. They're taking their cue from the new ways of designing and building applications that focus on modularity, service-oriented architecture, APIs and distributed designs. And so they will generally start in the cloud and may live in the cloud indefinitely.
However, as Mode 2 applications grow, and we've seen this even with some startups, a few years ago, Zynga was probably the most famous one, where they started in the cloud and as they grew, it became clear that it was cheaper and more efficient to operate those internally, to build your own cloud, than it was to keep renting the capacity.
Again, hitting the Mode 1 cloud dichotomy, Mode 1 applications most certainly are not on the cloud now, but that doesn't preclude them from being deployed on a cloud at some point. It will typically be a public/private hybrid, and the general migration path there would be to a private cloud -- one that it is owned and operated by the business first, not straight to something like AWS [Amazon Web Services], or Google Cloud, or what have you.
Bimodal IT operations myth No. 4: All IT is moving to the cloud
Another myth here is that ... all IT is ... moving to the cloud. When people hear [about] bimodal, they think, 'OK, everything's moving to Mode 2. That's cloud. That means IT [will] just going to be nothing but outsourcers to Amazon.'
I would stress that every organization needs cloud expertise and capability. And for the vast majority, that should include both the ability to run in public and private clouds. And fortunately, depending on your cloud of choice, for some, notably Azure or vCloud, that may mean the same set of management tools and processes.
But it doesn't mean that everything is going to end up on a cloud. Tightly integrated, server-bound applications are going to be around for the foreseeable future. And just look at mainframes that are still around. IBM still sells them, and they show no sign of dying.
However, Mode 1 applications can be improved by adopting more cloud-like paradigms of application design and development. So, over time, they should be rearchitected. However, the goal here isn't so much to rearchitect, to kind of forklift, move them over to a cloud and eliminate your data centers -- although, some companies like GE and others are doing that.
But the real goal for most is to use the innovation that cloud services are fostering, to improve application performance, resilience and efficiency. That's where even Mode 1 IT organizations and applications can take some learning from the cloud.
Bimodal IT operations myth No. 5: Bimodal IT is an endpoint
I think the final myth here is that bimodal IT, and this gets back to the bipolar quip, is that it's ... an endpoint. It's not a terminal organizational structure, per se, although it may be that it may persist over time that you have different sides of the organization focused on different priorities. But it's really, I would say, more importantly, a journey. And that journey is to foster ties between the IT and business, and to foster innovation in digital business.
And bimodal provides both a structure for doing that within IT and a way of thinking about the problems that helps IT executives do that, to think about, 'How do I need to operate? And what are some of the behaviors I need to instill to become more business-friendly to my colleagues in other business units, and to serve their needs with new IT products better?' So, as I say, it's really a model for understanding and developing your strategies.
So, as you go through planning processes, the bimodal framework can provide a good way to partition workloads and project priorities, and then focus on the most important characteristics of each without getting the Mode 1, Mode 2 activities intermingled. And it really lets both halves succeed to their maximum.
So, as Gartner puts it, it's really about capturing opportunities that increasingly must be attacked rapidly and over some very innovative solutions that contain a lot of risk or uncertainty.