• Ensures a single version of the truth across the enterprise
• Supports accelerated & simplified access for data analysts
• Consolidates data from across disparate systems
• Ensures data accuracy and integrity
• Conforms complex data with standard business rules
• Accelerates reporting project development
• Reduces costs by eliminating redundant data silos
   

Consider This We have identified four common mistakes among failed data warehousing initiatives. Here’s how to avoid them:

Don’t rush your data model.
While it’s human to try and get things done quickly, far too often the data model is the corner that gets cut in data warehouse development. Nothing is more important to your long-term success than establishing the right data warehouse/data mart data model. You simply cannot “come back later” and fix this stage of development. Rushing the data model inevitably produces a less-than-perfect system requiring endless “work-around” solutions which over time can generate unnecessary complexity and enormous expense.

Don’t throw in “everything including the kitchen sink.”
While the goal of a data warehouse is to include data from throughout the enterprise, the temptation to incorporate too much too soon can be a recipe for disaster. Even the simplest data sources bring with them a long list of things that have to go right and that could go wrong.

Don’t deploy “everything to everyone.”
Here’s the scenario: The vendor showed you how incredibly easy the interface is, you have a few hundred users clamoring for access, and the first set of reports are ready to roll. You’re feeling pressure to deliver and are considering just “opening up the system.” WAIT! Nothing is worse for the long-term success of a data warehouse than a botched deployment and the horrible PR it generates across the company. Instead, consider the following: "

• Pilot Groups:
  Initially, there will be issues with both the technology and the data. Accept this and limit your exposure to help find and fix these before releasing to the masses.
   
• Segment Users & Functionality:
  Today’s reporting tools are powerful. In untrained hands this can spell disaster. Identify who needs what and segment your users groups appropriately. Keep in mind, your power-users will come to you and identify themselves.
   
• Garden Paths for Analysis:
  If you know the types of business questions your users will be asking, develop helpful “Garden Paths” of analysis for them. This can often be accomplished through carefully structuring access and defining specific methodologies for building reports.

Don’t keep your users in the dark.
Communicate with your users. If , for example, the nightly data load fails, let your users know. Users are more understanding and trusting of your ability to handle issues in the future when you clearly communicate the issue, the reason for failure and the expected resolution.