Over the past year I’ve read a couple of articles that described the task of meta data management as “meta data misery”, “a necessary evil” and “meta data morass”. In these articles the authors discuss meta data management as if it is some horrible ill that companies are forced at gun point to endure. I am quite surprised by this viewpoint as implementing a managed meta data environment (MME) isn’t any more difficult than building any very large, highly complex enterprise spanning application. For example, the success rates for MMEs are significantly higher than those of customer relationship management (CRM) applications which fail around 90% of the time. Moreover, while enterprise MMEs are expensive projects, their costs pale in comparison to typical enterprise level CRM, ERP and/or data warehousing initiatives. It is these same individuals who state that corporate executives don’t care about meta data management. There was a recent survey that asked CIOs about the goals of innovation and the technologies that they believe will deliver innovation. Therefore, I would like to examine the results of this survey and ask the question: “should your CIO care about meta data management?”What Do CIOs Care About?
Before we can decide if the CIO should care about meta data management we need to first understand what CIOs do care about. Table 1 shows the results of a survey of CIOs that marked each of the technologies that they believe would provide the greatest levels of innovation.
Table 1: What Are The Key Technologies For Innovation
We are going to walk through these key innovative technologies and see where and if, meta data management plays any role.
Redesigning or Rationalizing IT Architecture
The need to redesign/rationalize information technology (IT) architecture was by far the survey leader with 73% of CIOs citing this as an innovative technology. It may be surprising to some to see rationalizing IT architecture as being the top innovative technology choice for CIOs; however, this is a distinct trend that has emerged from the glut of redundant applications and needlessly interdependent processes that has handcuffed most large organizations IT departments.
Redesign IT architecture is not a specific technology so it is important to understand what technology best enables an organization to redesign your IT architecture? Before we can address this question we need to have a clear understanding of the key components of your IT architecture. Key architectural components include:
- Data Attributes
- Manual Processes
Before it is possible to redesign these components one must understand (the following is a partial list):
- what components exist?
- where do they exist?
- what are they used for?
- who uses them?
- data stewardship
- What does the data mean?
- how does each component interact with each other component?
For those of you who are meta data practitioners you realize that all of this information is really meta data. In fact every successful enterprise architecture team that I have ever worked or spoken with, have come to the realization that they are a meta data management team. The reason being is that a managed meta data environment which provides the key functionality for redesigning an organizations IT architecture. In fact, to attempt this without meta data management would be impossible for a Global 2000 company or any substantial government organization. Just imagine if the average company attempted to manually track the 500,000 – 3 million individual data attributes in your IT architecture. Now think of the effort it would take to manage the 30 million or more data lineage relationships between these data attributes! Regardless of how intelligent a person or group of persons may be, it is impossible to manage this meta data without the assistance of an application. The application that gathers, retains and disseminates this information (meta data) is the managed meta data environment.
55% of CIOs believe that data warehousing and access to data technologies can provide critical innovation to their organization. Many companies 1st generation data warehouses focused on integrating data from their source systems and presenting that data to their end users. Typically these “data warehouses” had limited data volumes and more resembled a data mart rather than a “true” data warehouse. As companies become more sophisticated their data warehouses entered into their 2nd generation. This iteration tended to evolve their data warehouses into large, multi terabyte enterprise applications. Of course the demands of the business have continued to push the data warehousing forward into the “active” world. We see that companies no longer want to supply chain information that is updated as of last night. Now they want the data to be as current as is possible.
It is common for end users to complain that they have lots of data “thrown” at them but they don’t understand the data and they don’t have “actionable” information. By “actionable” I mean information that answers whatever question that the end user needs answered so that they can do their job better, faster and more effective.
Anyone that has ever worked substantially in the data warehousing field understands that meta data management is critical for evolving your data warehouse through it different iterations. In addition, business meta data that is presented to the end users is an absolute necessity in transforming data into actionable information.
Next issue I will continue to walk through the key innovations that CIOs want and the vital role that meta data management plays.
 CIO Magazine, April 1, 2005 (respondents checked all that applied)
 For a more detailed discussion on this topic see Chapter 1 – Facing Corporate Challenges with an MME, “Universal Meta Data Models”, David Marco, Michael Jennings, John Wiley & Sons 2004