Intelligent Enterprise

Better Insight for Business Decisions

RSS
Webcasts
Whitepapers
Subscribe
Home

May 31, 2003

The Quest for the Universal Data Model

Like the Holy Grail, the universal data model may be unattainable

by Joshua Greenbaum

Continued from Page 1

But the melody lingers on, no doubt led by the realization that you can take a reasonable solution to this problem right to the bank. And the desire to cash in has led to a number of interesting claims regarding universal data models. And a fair amount of confusion about what has been promised and what is possible.

The Road to Utopia

Take standards efforts like RosettaNet. The original Rosetta stone was famous for allowing the translation of Egyptian hieroglyphics by matching them to a slightly less ancient form of Greek. Its utility in deciphering the communications of one completely vanished culture by using another whose heyday had also long since passed is a fitting goal for RosettaNet: a means to understand communications between applications based on moribund data models and no-longer-strategic information. The really important communications, the value-added ones that differentiate one company's way of doing business from all the others, won't be expressed in RosettaNet. The standards effort may wind up interesting to data archeologists, but it won't solve anyone's universal data model problem.

The other universal data model effort worth mentioning is SAP's master data management (MDM) initiative. Unlike RosettaNet, important parts of MDM make basic sense. That basic sense stems from SAP's ability to use its market position and influence to dictate some form of a universal data model that derives from SAP's applications (and user base) and can become a de facto standard within the larger SAP ecosystem. SAP believes that if it can define a universal data model and make it part of its NetWeaver composite applications platform and enterprise services architecture, everyone will flock to the opportunity.

SAP's market clout does allow it to dictate these terms. But its clout is big only in terms of its relative size: SAP, like everyone else in enterprise software, lacks anything resembling a market monopoly. If NetWeaver takes off, SAP believes that its clout will increase as the data models that SAP hopes to embed inside NetWeaver become more prevalent. Of course, just having the data models will make NetWeaver appealing, even if its takes time for SAP's data models to approach the universal. Meanwhile, SAP's first uses of MDM will help SAP's customers reconcile the disparate data models that exist within their different versions of SAP software, a good first start and a highly laudable goal.

Rate This Article

Comments:

Optional e-mail address:

But total universality reminds me too much of Esperanto, which by synthesizing the Tower of Babel into a single language rendered itself largely useless. Languages, like data models, are the expression of unique cultural attributes that, by definition, defy universality. SAP will probably be able to take a leadership position within its customer base, which by definition may someday force the Rick Bergquists of the world to support SAP's version of the universal data model. But there's still an "obvious obscurity" to the overall concept: Truly universal data models won't emerge any time soon. Any more than the Holy Grail, whatever that is.

Joshua Greenbaum [josh@eaconsult.com] is a principal at Enterprise Applications Consulting. He researches enterprise applications and e-business.