PRE-DECISION SUPPORT
by Robert S. Seiner

The industry of data warehousing has single-handedly raised the awareness of the need for corporate data management. Beyond data management, data warehousing has also raised the awareness of the need for meta data management and more precisely meta data that supports the data warehouse.

Decision support databases are often considered similar to data warehouses. Companies put themselves through the rigorous tasks of creating data warehouses to, simply put, provide the ability to make better decisions from corporate data. Less attention is paid to the need for a second (earlier) level of decision support that is very necessary. This level of decision support can be called pre-decision support.

Pre-decision support is a name that I use to label "support for decision makers before they make decisions" from the data in the data warehouse. Much of this support comes through the use of meta data about the data warehouse. Let me explain further...

In most cases, warehouse end-users ask specific questions about warehouse data that, if left unanswered, jeopardize the success of an accepted decision support environment. Most individuals want to know simple information about the warehouse data, such as, what types of data exist, how the data is named, defined, and referenced, and how the data can be selected. There are also end-users that want to know more complex information about warehouse data, such as, what business rules were used to select the data, how the data was mapped, transformed, cleansed, and moved to the data warehouse, and how the data is audited and balanced. Perhaps, THIS information may be more important to many of the warehouse end-users than the actual data itself. All of this information is contained in meta data.

Without some level of pre-decision support through the use of this meta data, warehouse end-users are left guessing on what data exists and what data should and can be used to help them to make better decisions. Without this support, these same users spend more time performing 'data-hunting' activities, or at a minimum, they become less inclined to take full advantage of the data in the warehouse. This dampens the results of the work and expense that goes into building the data warehouse. In these situations, return on investment in the warehouse suffers and the decision support environment falls short of addressing the business and warehouse developer's expectations and requirements.

Before decision makers and data analysts make decisions from data, they are required to make decisions on selecting data for their analysis. Before decision makers can select the data for their analysis, they must understand and trust the data. One way to improve trust is through improved documentation.

Enterprise data asset catalogs (meta data repositories or databases) are a logical place to store documentation about warehouse data. Companies often shy away from producing data asset catalogs (and thus pre-decision support) because of the expense in time, money, and people. Many of these same companies also end up asking questions about why the data warehouse is not being used to its greatest potential. Most warehouse consultants will tell you that the two are clearly related.

Warehouse meta data to populate an enterprise data asset catalog can come from a variety of sources. Tools that contain meta data include data modeling tools, data extraction tools, data cleansing tools, data reporting tools, data base management tools, data security tools, and so on. Almost every tool in the warehouse development process (and application development process) makes use of meta data and stores it is a tool-based repository.

Tool based repositories can be proprietary in nature (meta data format is not made available to the purchaser) or they can be provide an open architecture (meta data structure is known to the purchaser). It is recommended that this determination be known prior to selecting the tools for your organization.

These repositories provide adequate support for the activities of the specific tools, but often do not address the broad scope of meta data required to supply the level of pre-decision support defined earlier. In most companies, more than one tool is used in the development of a data warehouse. Progressive warehouse and meta data managers understand the meta data of each tool and toolset and make the required meta data available to warehouse end-users.

Uses of meta data are often rooted in the data warehouse development process. However, uses of meta data reach beyond the warehouse into areas of data quality, data accountability, data security, application development, object management, year 2000 initiatives, and most likely, data modeling and data resource management (or data administration).

The statement is often made that "I wish I had better data from which to make my decision". Do not be surprised if those around you begin saying "I wish I had better meta data to know which data I need to use to make better decisions". Pre-decision support through the use of meta data will enable most organizations to ease the decision maker's pain and provide the company with better decision making power.