I'll be the first to say that building an extended intelligent enterprise (XIE) IT infrastructure with a zero-latency, real-time performance operational data store at it's heart is probably the best long-term use of today's tight IT dollars. However, just as the hardware and business intelligence (BI) applications must be suited to the task, so must data management and exchange capabilities. How well organizations manage their internal data ­ specifically their meta data ­ and the technology that underpins data exchanges with trading partners will play a large part in determining the success or failure of any XIE project.

I cannot stress enough how important it is to define and maintain meta data. It is the foundation of organizational data because it describes what that data is. However, meta data management is still one of the thorniest issues in most organizations. As Stamford-based Gartner indicates, "Since the dawn of information technology, the problem of understanding meta data ­ what data and processes exist within the organization ­ and sharing this information across people and tools has been a problem. There are many facets to this issue, some technical and others cultural."1 They're right.

For one thing, many IT folks just don't realize how critical meta data management and standardization is for any IT project ­let alone a project that aims to integrate all essential internal and trading-partner systems to serve a zero-latency decision-making environment. The second problem with meta data management is that there are still, even after a decade of wrangling, no coherent standards.

In theory, the lack of education concerning meta data management should be easy to solve. However, too often, meta data planning is sacrificed in the project plan along with other "niceties" such as project management time and logical data modeling. The prevailing opinion among many gung-ho, cash-crunched IT development staffs is "We'll get to it later" or "The meta data will take care of itself as the systems are built." Either of those meta data pseudo- philosophies is dangerous in any IT project, but they are fatal when it comes to XIE projects that must cope with data from other organizations, not just other departments!

What can organizations do to make sure that they define and maintain good meta data that will serve their IT projects well and help ensure their success? The first thing is to ensure that capturing, defining and maintaining meta data is a top priority. Because meta data is data about data, the genesis of any data quality initiatives within the organization will be at the meta data level. Without good meta data, the entire IT project is built on shaky ground.

Also, make sure the IT systems- development team knows what types of meta data exist within the organization, how that meta data is used in different situations and how best to capture meta data when the need arises.

Third, get a top-notch meta data repository and meta data management tool suite. Larry English has established criteria for a first-class toolset. The best tools will perform one or more of the following tasks:

  • Assure conformance to data naming standards.
  • Validate data name abbreviations.
  • Assure all required components of data definition are provided.
  • Maintain meta data for control of data reengineering and cleansing processes
  • Evaluate data models for normalization.
  • Evaluate database design for integrity, such as primary key to foreign key integrity, and performance optimization.2

Remember, however, that there's no substitute for organizational knowledge. Tools can only assess the technical aspects of meta data; they can't tell whether or not the meta data organizations have captured is truly the meta data needed in order for the business users to do their jobs.
Finally, as much as possible, standardize meta data throughout the organization and throughout the entire value chain. The recent merger of Object Management Group (OMG) with the venerable Meta Data Coalition (MDC) has resulted in a new meta data standards model that will consolidate its own Common Warehouse Model and the MDC/Microsoft-owned Object Information Model.

The new model will tackle issues such as a common metamodel, common syntax and rules for data import/export, and knowledge-worker processes such as OLAP and data mining.3 It provides a promising start to addressing the lack of meta data standards.

There's no reason for any organization not to make meta data a top priority and increase efforts to capture and maintain meta data properly. Remember, meta data is data about data. If the meta data is not high quality, the IT systems built from it won't be either.


1. Gartner Group. M. Blechar. "OMG's Common Warehouse Metamodel Specification," 28 July 2000.

2. English, Larry. Improving Data Warehouse and Business Information Quality: Methods for Reducing Costs and Increasing Profits. John Wiley & Sons, Inc. 1999, p. 323.

3. Taken in part from Richard Adhikari's article, "Metadata Conundrum Carries On." In Application Development Trends, October 2001, p. 51.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access