DM Review Published in DM Review in October 2000.
Printed from

Meta Data and Data Administration: Meta Data ROI: A Competitive Advantage

by David Marco

This column, an adaptation from my book, Building and Managing the Meta Data Repository: A Full Life-Cycle Guide, marks the second in a series on specifically defining the value that meta data provides to a corporation. This month's column will cover meta data driven business user interface and data quality tracking.

The reason we exist as IT (information technology) professionals is to meet the informational needs of our business users. Unfortunately, our current systems are falling well short of meeting the needs of the business. One of the reasons is that instead of designing systems that speak to our business users in the business terms they are familiar with, we have built systems that communicate to them in IT terms. Meta data holds the key to resolving this challenge. Meta data addresses this situation as it provides a semantic layer between our IT systems and our business users. In simple terms, meta data looks to translate the systems' technical terminology into familiar business terms. Figure 1 illustrates a Web-enabled decision support system. This Web front end is designed with the business user in mind. One thing that we need to understand is that the business users of our systems do not care whether the information they are looking at comes from a data warehouse, data marts, operational data store or meta data repository. They just want to be able to find the information they want quickly and in a manner that they understand. Let's suppose that the business user wants to see the numbers on monthly product sales. The user would go to the decision support Web site (see Figure 1) where they would have the ability to search for this flavor of report.

Figure 1: Meta Data Driven Business User Interface ? Decision Support System Web Site

Once the business user gets to the search page of the decision support Web site (see Figure 2), the user could search for any decision support reports that have "monthly product sales."

This is where meta data comes into play. In the meta data repository, there will be meta data that has business definitions for each of the decision support reports. Therefore, the business user can search through the meta data business report definitions for the reports that have the words "monthly product sales" in their meta data definitions. The results of this meta data search appear in Figure 2.

Figure 2: Meta Data Driven Business User Interface ? Search Results

The user can select from the reports that are returned or enter a new query. For our example, it turns out that our business user wants to see global, summarized product sales, by category, on a monthly basis by region. The second report returned looks exactly like what our user needs. The business user may want to know exactly how U.S. sales dollars are being calculated. The user could merely "right" click or hit a "hot" key on the "U.S. Sales $" field and see the business meta data definition for it. This type of business definition for U.S. sales dollars is stored in the meta data repository. By integrating this business meta data into the decision support report, the business user will understand exactly what goes into U.S. sales dollars. This type of information makes the data in the decision support system much more valuable and improves the accuracy of decision making.

As we can see, utilizing a meta data driven access has vastly improved this decision support system's value to the business user. In addition, the value of the actual information in the decision support reports is vastly upgraded through the use of business meta data. The value that business meta data provides includes:

  • The business users have a much easier time using the system; therefore, IT-related problems are reduced.
  • The system has a great deal more meaning to the business users, which allows them to do their jobs much better.
  • Business users of the system are able to locate the information that they need and understand the data presented.

Data Quality Tracking

Data quality is a significant issue impacting many, if not all, corporations competing in today's marketplace. Companies realize that IT systems are strategic weapons that can provide a significant advantage over their competition. However, if the data in these systems is redundant, inaccurate, missing or incomplete, the corporation is placed at a severe and distinct disadvantage. In addition, many companies have mission-critical initiatives such as e-business, customer relationship management and decision support. All of these initiatives will typically require data from the company's existing legacy systems. If the quality of the data in these systems is poor, it will directly impact the reliability, accuracy and effectiveness of any of these initiatives. The old IT saying of "garbage in, garbage out" illustrates that data quality or the lack thereof is critical to any enterprise.

Business/Technical ValueROI Measures
Improved Business Decision MakingData quality is improved, which provides the business users more acurate systems and reports.
Reduction of IT- Related Problems Data quality is improved, which reduces many system- related problems and IT costs.
Increase System Value to the BusinessBusiness users of decision supprt systems can make better decisions if they are aware of possible errors skewing reporting numbers.
Improved System PerformanceAs data quality improves, system errors are reduced, which improves system performance.
Figure 3: Meta Data ROI ? Data Quality Tracking Benefits

Meta data is a critical component of any data quality initiative. Meta data provides the mechanism for monitoring and improving the quality of the data coming from the operational systems into the decision support system. Meta data tracks the number of errors that occurred during each data warehouse/data mart load run and can report to the IT staff when certain error thresholds are exceeded. For example, if we are loading transactional sales records into a decision support system for the marketing department to view, we may decide that if more than two percent (our threshold) of the dollar amount of all of the sales transactions is in error, we need to stop the decision support system load processes and investigate the problem. It is important to note that on data records that have dollar amount fields, it is typically more applicable to set the error thresholds on the dollar values of the records in error than on the number of records in error. For example, let's suppose that there are typically 100,000 records, totaling $20,000,000 in transactional sales that are loaded into the decision support system on a monthly basis. If 2,000 (two percent of records) of these sales records totaling $20,000 (0.1 percent of sales dollars in error) erred out without loading into the decision support system, the business users my not feel that this is a large enough error to skew their decision-making process. However, if 10 (.01 percent of records) records erred out totaling $2,000,000 (10.0 percent of sales dollars in error) in sales, it is highly probable that the business users could not make accurate decisions. Keep in mind that it is important that the business defines what the error threshold should be since this is a business decision.

In addition, all of these decision support system's data quality metrics should be stored in the meta data repository and kept over the history of the decision support system. This allows corporations to monitor data quality improvement over time.

In decision support systems, it is common to compare field values from different time periods. A decision support report showing global corporate sales on a monthly basis could be used to compare U.S. sales from October 1999 to November 1999 for the holiday buying season. If the sales amount for November seem to be a little low, the business user could check the data quality statistics and see that 8.4 percent of the records in the November decision support load run erred out and were not loaded. The business user would know the margin of error when making decisions based on this report.

Unfortunately, quite often companies do not want to spend the money or the time to uncover, evaluate and resolve their data quality issues.

David Marco is an internationally recognized expert in the fields of enterprise architecture, data warehousing and business intelligence and is the world's foremost authority on meta data. He is the author of Universal Meta Data Models (Wiley, 2004) and Building and Managing the Meta Data Repository: A Full Life-Cycle Guide (Wiley, 2000). Marco has taught at the University of Chicago and DePaul University, and in 2004 he was selected to the prestigious Crain's Chicago Business "Top 40 Under 40." He is the founder and president of Enterprise Warehousing Solutions, Inc., a GSA schedule and Chicago-headquartered strategic partner and systems integrator dedicated to providing companies and large government agencies with best-in-class business intelligence solutions using data warehousing and meta data repository technologies. He may be reached at (866) EWS-1100 or via e-mail at

Copyright 2004, Thomson Media and DM Review.