Experimentation and the problem of “shadow” IT

IT departments often live in a world of contradiction. On the one hand, they must “keep the lights on,” by keeping servers and networks up, by delivering reports on time, and by ensuring that systems and data meet regulatory obligations such as Sarbanes-Oxley and other forms of compliance. These
requirements are nothing if not rigorous—and essential.

On the other hand, they and their business partners desire innovation: new programs and new  applications to support both new and evolving business opportunities, to better serve their customers,
and so forth. Yet the costs of IT operations—sometimes 70–80 percent of the overall budget—reduces the ability of IT to spend on new programs and innovation.

In many cases (in fact, in every enterprise we know), there are occasionally applications created and
deployed outside of the IT department in response to critical business needs. These unofficial applications are often referred to as “shadow” IT. Instead of going through the usual budget,
requirements analysis, design and deployment phases typical in the creation of a new IT application, a  marketing department publicizing a new campaign might simply create a new website on their own.

Because it eliminates the capital-expense investment component (i.e., servers, storage, and network)
of application development, the cloud makes this sort of rapid innovation much, much easier. In effect, all that is needed are a few coders to write the application—and a credit card.5

IT executives should realize that this sort of innovation and experimentation is inevitable, and in many cases actually desirable. As the business climate rapidly evolves, it is critical for both businesses and IT organizations to foster rapid experimentation and innovation. It will be important to educate businesses on the importance and consequences of regulatory issues and noncompliance, of course. IT departments can actually help them by providing controlled, managed access to critical data, such as customer information, rather than letting them gather and manipulate the data on their own.

As soon as a company starts this process of envisioning and creates the culture of experimentation, it
learns a disruptive truth: in the cloud era, you must experiment, fail fast, and learn fast. It is as important to experiment in order to learn quickly both from successes and from failures. Learning
from how you succeed and what makes you fail provides the basis for delivering the disruptive
innovation and value from the cloud.

As you can by now expect, these phases shape the cloud migration principles used for the rest of the
process, these principles are go fast, push the boundaries, make data-driven decisions, simplify, and,
finally, communicate to succeed.

 Go fast exemplifies the spirit of the experimentation phase. For some, it might represent a new
way of thinking for IT because, with the cloud, you can “spin up” new projects quickly with a few
clicks rather than having to plan, allotting datacenter space, procuring equipment, and so on. We
call this the try many, use best approach because the cloud uniquely facilitates the ability of IT
departments to choose the best of many solutions.

 Push the boundaries suggests that wherever possible, IT should not simply adapt to the new paradigm of the cloud, but embrace it and adopt new architectures and processes quickly to best
exploit the new opportunities.

 Make data-driven decisions proposes that you carefully track and measure the numbers, including the cost effectiveness of the cloud for financial reasons, system telemetry for technical efficiency
reasons, and so on. Following the data carefully will make it possible for you to make informed
decisions about which applications are generating the most return, about which you should prioritize, about which are performing well in the cloud, and where potential problem areas exist.

 Simplify focuses on retiring, right-sizing, and consolidating as many services and applications as
possible. Applications that are infrequently or rarely used often generate significant costs for an IT
organization, with little return. Retiring them and consolidating them with applications that
perform similar functions can, conversely, generate savings in a number of areas such as hardware, system software licenses, and maintenance. Consider generating metrics around “hot” and “cold” applications based on CPU, network, and database utilization; for example, an application that averages two percent of CPU and has few authenticated users might be just such a “cold” application.

 Communicate to succeed is the single most important mechanism that guarantees continued success and not just the migration of a single application or a service. Establish a clear and continuous communication channel for stakeholders to visualize success and impact as well as to understand the failure and the lessons learned from them. Key stakeholders remain engaged and continue to invest when they feel their participation in the joint effort required to make this a continuous journey and not just a single trip.

Source of Information : Microsoft Enterprise Cloud Strategy