Don't Just Serve — Enable: A New Model for IT Organizations 08-29

Don't Just Serve — Enable: A New Model for IT Organizations

by Bill Franks  |  

The classic model of an IT organization as a central point of control over all things data was never ideal, and today it simply doesn't work. Business users need increasingly fast, broad, and flexible access to data. It is no longer possible for IT to address every single business request. So, what's an organization to do? A small shift in thinking and process—moving IT from a serving mode to an enabling mode—can make a huge difference.

The traditional IT model is much like a classic restaurant. First, IT acquires and stores all corporate data in much the same way that restaurants acquire and store all the ingredients for the dishes they serve. Next, IT provides a pre-determined menu of specific data you can have and in what combinations. 

Just as with restaurant menus, you can ask for substitutions or to leave out something, but you may not get what you want or, if you do, the change will take extra time. Any modification adds complexity and cost, so restaurants try to avoid them. Similarly, many business people are forced to consume what IT has chosen to serve them rather than what they really want and, in many cases, need.

IT can no longer serve precisely measured and pre-determined views of data to users. To be successful in today's world, IT must to shift focus to enabling access to the range of raw data elements that go into those traditional views. Then, users must be able to mix and match data as required for their specific problems. One method of doing this is to create "discovery" environments where users can freely explore data.

 These environments include "sandboxes" or "data labs" which are slices of a production environment's resources that are allocated to users. Within a discovery environment, users can query a broad range of data, create output data (which is not typically allowed in such systems), and even load new data. Of course, there are limits on how much data users can load or create. Those limits simply must be high enough to effectively navigate the discovery process.

I have seen many large companies accelerate the development of new analytics through the use of a discovery environment. Not only are users able to experiment more broadly and freely, but since they are already working within the scalable systems that will be used to deploy their findings, it is much easier and faster to move from prototype to final product. 

This in turn allows more iterations through the discovery process which allows the benefits to compound. With more flexibility and less bureaucratic overhead, users are able to be more innovative, more efficient, and are typically much more satisfied as well. After all, wouldn't you be happier if you weren't constrained arbitrarily in how you are able analyze data?

It is important to note that this isn't about ripping out and replacing an organization's data storage and analysis systems. It really doesn't require much in the way of capital investment. Rather, it is simply about changing how data storage and analysis are configured and how users interface with them.

 The systems in place at large companies today can easily handle the updated model as many Fortune 500 companies have proven. This shift requires some cultural and mindset changes, of course. At the same time, for those users who still prefer a predefined menu, there is nothing stopping IT from offering that too. The key is to enable users to choose what they prefer.

A model of enabling broad and flexible analysis of data as opposed to serving predetermined views of data isn't all that radical if you really think about it. Why wouldn't you want people empowered to freely search for the next great business-changing analysis? While it is a model that many organizations have yet to embrace, make sure that yours makes the shift.

View at the original source