Paper 116-29
8 Characteristics of a Successful Data Warehouse
Marty Brown, Lucid Analytics Corp
Abstract
Most organizations are well aware that a solid data warehouse serves as the foundation from which to build
meaningful business and analytical intelligence. Unfortunately the gulf that exists between being aware of
the need and actually building a data warehouse seems too large to cross. Instead of building a data
warehouse, many organizations build their analytical solutions from spreadsheets and query tools.
There is a better way. This paper servers as a roadmap from which to bridge the gap between concept and
reality. Included are brief discussions in the following areas:
• The executive sponsor
• The project team
• The business
• The data model
• The metadata
• Extract, transform, and load (ETL)Q
• Marketing and change management
• The users
Once complete, the data warehouse allows organizations to unleash the power of their data in the form of
analytical intelligence. It is at this point that they can begin to achieve a true competitive advantage.
Introduction
As organizations seek to move up the intelligence value chain, it becomes clearer everyday that their
previous strategies for delivering information are becoming inadequate. What was once “good enough” for
reporting and analytics has now been legislated into a more robust and controlled environment. Businesses
simply cannot be run on spreadsheets anymore. Official reporting data that touches many hands is quickly
becoming a thing of the past.
Enter (or re-enter as the case may be) the data warehouse. What was once primarily considered an IT
centered project, now has the eye of the business. With the passing of legislation like the Sarbanes-Oxley
Act, corporate executives must feel comfortable that the financial (and other public) reports that their teams
prepare and that they sign are accurate. Many of the failed data warehouse projects of the past lacked true
commitment on the part of the business. The fact is it is nearly impossible for a data warehouse to be
successful without active participation from the data owners, i.e. the business, on a daily basis.
The focus of this paper is getting from here to there. It includes the roles the business plays as well as
some suggestions on how the warehouse, once implemented, remains the firm foundation from which
organizations build out the rest of the intelligence value chain.
Characteristic 1 - The Executive Sponsor
At the heart of the data warehouse is the executive sponsor. This is the person who believes in the value of
the warehouse and is willing to use his/her political clout to fund the project. The executive sponsor is the
one individual that must clearly understand the ROI of the project and be able to articulate to his/her peers.
So what level of an organization does the executive sponsor represent? Generally it’s the “C” level (CFO,
CMO, CRO, etc) of an organization. By definition a data warehouse crosses over functional boundaries
within an organization so it’s extremely important that the executive sponsor carry enough clout and have a
high level of buy in from his/her peers. It is important that there be an individual as the executive sponsor
rather than a group of executives that collectively make up the “sponsorship”. If the scope of the data
warehouse is going to cross all functional areas, then many times the CFO makes the best sponsor. This is
SUGI 29 Data Warehousing, Management and Quality
