My company is implementing a DW, later this year an EAI project will be initiated. Are there any design considerations, primarily for the DW that we have to account for? I believe will we do much of the same work twice by putting the EAI project on hold instead of getting the EAI platform in place and then implement the DW. Am I right? What other considerations might there be?


Les Barbusinski’s Answer: You are correct in assuming that much of the work will have to be done twice if you delay the installation of the EAI product. Installing the EAI tool first will give you the option to a) structure some of your data acquisition processes to capture source data in near real time using EAI message broker technology, b) using the message broker technology to feed multiple data marts in parallel, c) using EAI adapter/integration manager technology to perform some of your simpler data transformations, d) using EAI portal technology to publish data warehouse information to multiple Web applications (e.g., CRM, ERP, e-business, etc.) and so on. One area that you will sorely miss the EAI tool is in the extraction of data from one or more of the commercially available ERP packages (e.g., SAP, Baan, Peoplesoft, JD Edwards, etc.). Internally, these ERP applications consist of thousands of tables replete with tens of thousands of hidden relationships and business rules, making direct data extraction a highly problematic venture. Most of these packages have external APIs for data access, but these APIs can number in the hundreds and can be difficult to master. In addition, they do not easily provide the querying system with “net change” data. Having the EAI technology in place can be a major help because many of the EAI tools provide “adapters” for the most popular ERP packages, and the J2EE- compliant EAI tools can even utilize third-party ERP adapters. This can significantly cut data acquisition costs and development time.

Doug Hackney’s Answer: Contact Neil Raden at Archer Decision Sciences (nraden@archer- decision.com). He has done some work in implementation of both of these technologies and has some good thinking on their relative strengths and weaknesses.

Joe Oate’s Answer: Well, everything is a tradeoff. Waiting until all of the enterprise applications are integrated could take several years. In the meantime, the benefits of having the analytical capability of a data warehouse would be lost. I think that this would probably be detrimental to your company’s ability to compete. As an example, one of our global clients decided to go ahead with a data warehouse, the first iteration of which was for purchasing. This company does in excess of a billion dollars in purchasing each year. In the first six months, the data warehouse identified nearly $100 million in possible savings. Obviously, not all of this will ever be achieved, but they should be able to realize 20-50 percent of these savings. The affect to the company’s bottom line will be the same as if the company were to increase sales by a billion dollars. If this company had waited until all or a significant portion of the operational systems had been integrated, significant shareholder value would have been lost and the company would not be nearly as competitive.

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