Data Governance Blog

Creativity vs. Standardization

Often the argument against standardization and data governance centers on not wanting to hobble the creativity of the application developer. The perception seems to be that in allowing the creation of customized data elements and data models, developers build a better application. In reality, the application comprises only one part of the entire information system and the customization adds tremendous overhead increasing total cost of ownership.

There are numerous examples in other industries and trades with some degree of standardization where creativity flourishes. Artists, designers, and photographers are both creative and adopters of a widespread standard. The Pantone® standard for colors allows the creativity to flourish by elimination of the problems of color matching and color management. Once the decision for a given color is made by artist, the color management standard allows efficient matching and reproduction.

At the heart of their system is a data model for color. There are three primary hierarchies Red, Blue, and Green at the first level then progressing to the next level where you’ll find blue-green, red-blue (a.k.a. purple), and so on. In reaching into their creative bag of tricks, artists and designers have come to rely upon color standards and freed their creative energy to focus on the content of their arts.

The same concept applies to application developers. In this case the standard would be the consistent use of an enterprise data model and master data management. Having one definition for product, account, and customer for example, frees the developer to draw from well defined standardized data elements. That practice allows the developer to focus their valuable time on the user experience, or the application behavior.

Standardization of data models also promotes system integration and interoperability. The engineers in manufacturing learned this when Henry Ford built the first assembly line for his cars. The genius was not as much the assembly line, rather the standardization of components and parts. A fastener of a certain specification could be relied upon to fit and function no matter where in the assembly process it was used. This approach now dominates manufacturing.

Applying the same principals in IT reduces system complexity. Moving data from Application ‘A’ to Application ‘B’ requires nothing more than a straight movement of data without transformation. Because the interface is well established and defined enterprise system interoperability and integration are greatly simplified. The absence of a transformation layer eliminates the overhead of the added processes within the system and the people to maintain them. This nirvana state requires a standardized data model and master data management.

Within information systems management, many still hold onto the idea of building transformation logic as a practical solution to move data from one application to the next. That practice equates to adding an adapter to fit two non-compliant parts together in manufacturing. Manufacturers avoid this because it adds overhead to build, store, install and maintain these adapters. Often the adapters add little value and often actually weaken the final assembly. Within information systems management, the practice stands out as an example, that something easy to do is not always the right thing to do. There are cases where it is appropriate to transform data, but it should be the exception rather than the rule.

How do you think the general public would respond to aircraft or car mechanics building custom fasteners and parts for each assembly they put together or repair? The public would not stand for it, yet they invest billions in companies whose IT systems are assembled in this manner. The investors are ultimately not going to get accurate and reliable information from the IT systems feeding financials. Right now that prospect scares me much more than getting on a plane, or getting behind the wheel of my car.

__________________________________________________________

Thanks for stopping by. My writing is intended to take a lighter look at Data Governance, and toss in some pragmatic advice along the way.  If you are interested in more information on how to implement Data Governance in your organization, please contact me via LinkedIn or the email address below.

Regards,
Tom Jesionowski

LinkedIn Profile