Here’s a good find I wanted to share with Dashboard Spy readers. The IBM document (link provided below) contains 4 excellent segments:
1) Eight strategies for delivering business intelligence on the web
2) Use analytics for strategic business information implementation
3) Making the operational case for data warehousing
4) 10 ways to begin a data warehouse project
The last topic was of particular important during a conversation that I had with a data modeling consultant that called me for help. Let me share some of the article:
First, here’s the link to the IBM Data Warehousing and Business Intelligence PDF:
10 ways to begin a data warehouse project
IT departments typically launch data warehouse projects without input from business partners, building the business case themselves. But what steps should you take once the project gets greenlighted? These pointers will help you get off on the right foot.
High functioning value‐added IT departments operate in a consultative mode, using the enterprise business model and strategic plan to work effectively with business partners to identify technology‐based solutions in response to requirements as articulated by the business. Projects are launched based on collaboratively deciding what’s needed.
A data warehouse, however, is one of the few examples of a project that’s typically initiated independently by IT without input from the business. IT has to build the initial business case for the data warehouse, since few people outside the technology discipline understand what a data warehouse really is or what kind of value it can provide. What’s the best way to get started? Here are some suggestions.
(see pdf for the full discussion of the follow 10 points.)
#1: Determine your organization’s appetite for change
#2: Identify the most likely business unit to benefit from a data warehouse and approach it proactively
#3: Determine data access controls
#4: Assemble the team
#5: Decide on the implementation approach
#6: Identify the project scope
#7: Establish the success criteria
#8: Conduct the “25 question” analysis
#9: Assess current data quality and pre‐cleansing efforts
#10: Don’t forget the metadata