Implemention Technology
Data Transformation
  Data Junction
Data Extraction
  DJEngine
Warehouse Engine
  Microsoft SQL Server

Taubman, a veteran of the shopping center industry for almost 50 years, needed an automated solution to combine 1.6 million records stored in multiple, isolated legacy systems and to convert the data into useful information. We wanted a cost-effective, easily maintained system that could negotiate several different data formats and perform unattended transformations on a nightly basis. Taubman turned to Data Junction, a versatile and cost-effective tool, to centralize access to the information contained in its various legacy systems.

Founded in 1950, Taubman is recognized as one of the preeminent retail/developers/owners/managers in the United States, and its 17 properties are among the most productive in the nation. Taubman Centers has one of the most prestigious portfolios of retail properties in North America, including such landmark destinations as The Mall at Short Hills (Short Hills, N.J.), Beverly Center (Los Angeles, CA) and Cherry Creek (Denver, CO). Sales productivity of Taubman's malls is over $425 per square foot, a productivity level that eclipses the national average for regional retail centers by a wide margin. In November 1992, Taubman Centers, Inc. (TCO), a real estate investment trust, debuted on the New York Stock Exchange.

Like many businesses of its size, Taubman stores huge volumes of business data in isolated legacy systems. Buried within that data is invaluable information about everything from accounts payable to the company's latest marketing strategies. The problem, however, is getting at that information. To help us combine the raw data in operational data stores and convert it into useful information, we deployed a powerful business intelligence solution based on software from Data Junction Corporation. The sophisticated solution is empowering Taubman to deploy data marts in as little as a day ­ saving the company an estimated 300 hours and thousands of dollars a month on administration as well.

Of Malls and Marts

Our search for a business intelligence solution followed a corporate decision to create specialized data marts for use by business analysts. To bring this plan to life, we first had to build an automated solution to extract, transform and integrate the data in our operational databases and deliver it to the data marts. We needed a system that was cost-effective, easy to maintain, could negotiate several different data formats and perform unattended transformations on a nightly basis. And since Taubman continues to aggressively develop and acquire retail properties, the solution had to be both scalable and flexible.

After carefully researching our options, we deployed a data transformation and exchange solution that combined Data Junction's toolset with Microsoft SQL Server, running under Microsoft Windows NT Server. Data Junction's graphical environment enables us to rapidly build and implement robust, automated data transformation processes from source data received in any format.

We evaluated several solution providers before implementing our data marts with Data Junction. All of those tools cost $50,000 or more and require long training periods. Data Junction provides easy administration and cost a lot less. We need data marts that are easy to administer because we don't have a huge IS staff. That's the primary strength of our data mart approach.

Data Marts in a Day

Thanks to Data Junction's ease of use, I became proficient with the system in just a few weeks. Today, the company can implement a complete data mart in as little as a single day.

Data Junction is very easy to use. Even more complicated transformation routines can be designed in a matter of minutes.

Data Junction is a general-purpose data conversion tool that acts as a universal and neutral "junction" connecting structured data sources. Data Junction software permits users to translate data to or from almost any structured file format, including databases, flat files, spreadsheets, ASCII, binary, EBCDIC, SQL, ODBC, COBOL, ISAMs, vertical applications, mail merge, stats/math and many others. Filter, edit and mapping capabilities exist to modify characters, fields and records into the exact format required. Specifications can be selected interactively through a powerful graphical user interface or on the fly via the pure execution engine.

I use Data Junction at my workstation to visually design transformation, validation and cleansing rules, and we use DJEngine to automatically execute all of the transformation routines for each data mart. DJEngine is a programmable, embeddable engine that includes an application program interface (API) for integration with other programs as well as a command line interface.

DJEngine runs on a dedicated server powered by Microsoft Windows NT Server. Starting at 11:45 p.m. each night, DJEngine loads more than 1.6 million records into eight data marts, all running under Microsoft SQL Server. A custom ad hoc querying tool enables business analysts to study that data flexibly and intuitively.

Bottom Line Transformation

Data Junction and DJEngine have proven to be a potent combination of technologies for us. That they're so intuitive has meant shorter learning curves, faster implementation and rapid return on investment ­ but no sacrifice in power or versatility. More importantly, the solution is cost-effectively empowering Taubman to get its managers the business intelligence they need to make sound strategic decisions. Data Junction costs thousands less than similar solutions we evaluated, and we estimate that those systems are saving us 300 hours and thousands of dollars a month on administration. Thanks to Data Junction, there will be no more shopping for data at Taubman Centers.

Practical Advice

Anticipate the requests as best you can and provide enough data to meet those requests without including unnecessary data. The warehouse should be as lean as possible but still able to answer most queries. This includes filtering unnecessary tables or rows which will provide speed and ease of use.

Where practical, mirror legacy system data structures. Implementing a known and familiar model in the data warehouse, one that mirrors a well-designed legacy system, eases comprehension for all people involved.

Utilize a query tool that hides as much of the physical data structure and storage from the user as possible. The user's primary concern should be analysis of the results, not proper join syntax or proper table relations.

In some circumstances, denormalization of some of the information in the data warehouse may provide better performance than that achieved by following strict normalization rules. Remember that the data warehouse is an analytical tool, not a transaction processor.

Carefully consider the data freshness requirement to determine a proper data warehouse loading strategy. Will day-old information be good enough, allowing a nightly batch load, or must individual transactions be posted to the warehouse?

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