As is true with any type of house, a data warehouse must be structurally sound, have a trained staff of builders, and work with a reasonable budget. These, however, are simply a few of the numerous risk factors and pit falls that can present themselves during data warehouse implementation. Chuck Kelley addresses the major problems he sees occurring time and again in an article on information-management.com:
Risk factors include, but not limited to a)not having a corporate sponsor high enough up the corporate ladder, b)project management that has never built a data warehouse and insists that it is done like a transaction system, c) in-fighting within your team, d)designing the database to be transaction-oriented vs. aggregation-oriented, and e)not having your user community involved in the requirements and development process.
A strong sponsor is key, or else you risk people becoming unmotivated. Furthermore, even if people are motivated and a strong sponsor is present, there must be a strongly defined need when constructing a data warehouse. Without need, strategic objectives cannot be defined and overall organization will suffer. Worse yet, time is wasted with a lack of defined need.
Proper funding is also a requirement for success. The data warehouse must be constructed with good, sturdy materials (meaning quality data people actually wish to reference), or else it simply will not be used. Think of this as stairs in a home. If someone steps on the first step and their foot goes through it, they will not want to attempt any of the other stairs up to the top floor. Finally, poor project management will mean the death of data warehouse implementation, as it has meant the death of numerous project ideas in the past.