The race for a common data model

By Shawn P McCarthy -- Logistics Management, 10/1/1998

If you plan to do business online, it's worth learning a little about data models. Data models are those boring, behind-the-scenes structures that make online supply chain tracking and electronic commerce possible. Big battles are looming over what data structure will be adopted by the future supply chain, and the outcome could affect all of us.

According to Geoffrey Moore, head of the Chasm Group and author of such business books as The Gorilla Game and Crossing the Chasm, a data model is the standardization of a business object, designed to work closely with a computer application. It can serve as a container for an online transaction or as a larger framework for a business process or data analyses.

As enterprise resource planning (ERP) became popular in the early 1990s, proprietary data structures took hold as a way of moving information between, say, a system that tracks materials on the shop floor and a manufacturer's accounting system. But what's needed today, Moore and others say, is a common inter-enterprise data model. This would allow business partners to share data and conduct transactions over the open Internet (though most likely in an encrypted format).

The market wants a common data model to follow for business and supply chain needs. Standardization could help companies avoid what we see today--costly customized programming as companies trade data and feed it into their systems.

One of two things will happen. The market leaders, such as SAP, could end up creating a default standard as more businesses adopt their systems; or open systems proponents--like Hewlett-Packard, Sun Microsystems, Oracle, and others--will win the race by making their systems a bit easier to plug and play.

To a certain extent, open solutions are expanding now with On-Line Analytical Processing (OLAP) systems. But it's far from perfect, and OLAP could stall if proprietary solutions gain momentum.

Moore says the big push right now is for ERP, supply chain software, and business process optimization vendors to get their data model adopted by as many niche markets as possible, to create a critical mass of users to assure their data model will survive.

Here's a likely scenario, painted by Moore and echoed by others at the recent Planet 98 conference in Dallas: There will be a shake-out, and two data models will survive. The ERP-centric approach may lag a bit in the development of interfaces to budding front-office applications. But there still will be a niche market for developers who don't want to take a transaction-based approach to the sharing of business information, perhaps through custom messaging technologies.

As your company moves into online business, you'll be forced to choose a data model to follow. Although it's tempting to go with a proprietary solution, work with a consultant who can keep your system as open and standard as possible. That way you'll be more likely to plug into the data model used by your trading partners.

Pointers

* SAP software can be reached at www.sap.com

* i2 Technologies can be reached at www.i2.com

* Microsoft's supply chain management solutions can be reached at http://www.microsoft.com/industry/supplychain/default3.htm

* IBM's new supply chain management site can be reached at http://www.ibm.com/e-business/supplychain/

Prediction of the Month

In the race to become the industry-standard data model, one of these will win:

* Microsoft Corp. Value Chain Initiative/Supply Chain Initiative.

* SAP Software R/3 system, now a top ERP player and expanding to the supply chain market.

* i2 Technologies is coming up fast, with its Rhythm planning software and built-in business process optimization.

* Baan, PeopleSoft, and JD Edwards also are serious players but don't seem to have the momentum of the other three.