In today's difficult economic times, maximizing return on assets is a strategic imperative. Information assets are no exception. Service-oriented architecture (SOA), with its focus on agility and reuse, plays a critical role by accelerating development of new high-business-value applications from existing information assets.

However, before new SOA applications can unleash the business value of these information assets, enterprises must first gain an understanding of and control over the assets they have. This fundamental truth is the basis of today's metadata management initiatives.

Enterprises must strike a balance between resources spent on managing metadata, and resources spent on leveraging this metadata in their SOA initiatives. In doing so, they must address the following questions:

• Where should the focus be?
• How can the enterprise deliver the most business value the fastest?
• Is there a sweet spot where metadata management and SOA align for the benefit of both IT and the business?

Managing Metadata - More Important Than Ever
The fact that metadata is important goes without saying. Furthermore, its importance is growing as it evolves from a passive knowledge store to critical content for automated development. Business intelligence (BI) and data warehousing (DW) applications were some of the first to leverage data-oriented metadata. Business process management (BPM) initiatives leverage process-oriented metadata. IT management initiatives leverage IT asset-oriented metadata. Recently SOA-based new development and legacy migrations are driving additional metadata requirements. Finally, governmental regulations, such as Sarbanes-Oxley, are turning what used to be good IT hygiene into mandates. Given this explosive metadata proliferation, enterprises are expending ever-increasing resources on understanding, cataloguing, and managing metadata.

Today's metadata management applications and techniques vary widely. This is primarily a result of three factors. The first factor is driven by the various classes of metadata identified above. For example, the metadata used to define the extract, transform, and load (ETL) scripts for a data warehouse is far different from the metadata used to help define and manage the handoffs in a complex, order-to-cash business process. Not only do these usages influence the types of data and attributes being managed, but also the frequency and impact of change.

The second factor is the compartmentalized nature of IT organizations. Different teams work on different classes of applications. These range from business-to-business (B2B), to back-office applications, to BI, and so forth. Additionally, different teams work on different layers in the application stack: database, workflow, and reporting, to name a few. As a result, each team focuses on understanding its slice of the metadata pie, with little or no concern for other users.

The third factor is the software vendor-driven approach to metadata management. As metadata has risen in importance in recent years, applications and tools vendors have added metadata extensions to their core offerings. Further, a number of metadata management specialist vendors have arisen, each attacking the metadata management problem from a unique perspective. The result is a disparate set of metadata management tools and a new set of metadata integration challenges.

Data Services - The Sweet Spot for Managing and Leveraging Metadata
With so many classes of metadata, along with the various users and tools, it's easy to get lost in the metadata management journey. However, it's important not to get sidetracked! Enterprises need to pick their battles. They need to focus on managing information assets with the highest return, while using metadata management approaches that deliver those returns the fastest. When it comes to SOA, focusing on data assets often reaps the best and most immediate returns.