TCO Starts With the End UserThe conventional view of data warehouse total cost of ownership is myopic and wrongThe recent industry preoccupation with calculating the total cost of ownership (TCO) for a data warehouse focuses on minor issues while missing the big picture. Metaphors such as, "not seeing the forest for the trees," or even, "rearranging the deck chairs on the Titanic," come to mind. Reading various white papers and listening to industry consultants would make you think that data warehouse TCO is dominated by the costs of hardware, software licenses and maintenance contracts, IT staff expenses, and services (for example, consulting). Some of the more detailed analyses break these categories down to reveal chronically unplanned costs such as hiring and firing, training, tuning, testing, and documentation. One large hardware vendor even bases its TCO advantages on reducing the number of DBAs! With all of these categories as a foundation, the various vendors "prove" that their approach is better by claiming to show head-to-head comparisons. Observers like you and me are left to wonder how the conflicting claims could be so dramatically at odds. But all of this muddling misses the big picture. As one of my former colleagues used to say, there is a hippopotamus in the room but nobody is talking about it. Bad Decisions Are CostsWhen I think about data warehouse TCO, I start by asking: Why do we have a data warehouse, and who in the organization judges the cost and the return from a data warehouse? My answer is, the data warehouse publishes the data assets of the company to most effectively facilitate decision-making. In other words, the data warehouse is an organizationwide resource, the cost and value of which you must judge by its effect on decision-making in the broadest possible sense. So, who are the decision makers in an organization? For the most part, they're not the IT employees! As Bob Lewis wrote so aptly in a recent column, "IT is an enabler of value, not a provider of value." The mainline decision makers of an organization are the end users of the data warehouse, whether they're executives, managers, knowledge workers, analysts, shop-floor stewards, customer service representatives, secretaries, or clerks. All these end users have a powerful, instinctive need to see data. With the computer revolution of the past half-century, a profound cultural shift has taken place in the way businesses are run. End users are now quite certain that if they can see their data, they can run their businesses better. So when I step back to the broader perspective of an entire organization trying to make better decisions with a data warehouse, the traditional IT-centric costs don't even make my top ten costs to be concerned about! The hippopotamus in this room is a set of problems that can destroy the value of a data warehouse. And yes, to avoid these problems entails some real costs. Here's my list of the real sources of data warehouse costs that come before the traditional hardware, software, staff, and services costs that we've been so fixated on. The list begins with the most important in my opinion, but you can adjust the ranking to your own environment:
As we'll see, most of these problems are akin to snatching defeat from the jaws of victory. When these problems raise their heads, they can threaten the entire data warehouse initiative. If the data warehouse fails, the cost analysis is really ugly. Add the traditional IT-centric costs to the cost of running your business without your data.
|
 |
Most Popular This Week
IE Weekly Newsletter
Subscribe to the newsletter
| |
| |||||||||||||||||||||||||||||||
