During the design of a Business Intelligence infrastructure, some essential steps should be followed:
- Prioritizing the business processes
- Developing a roadmap for a phased implementation
- Analysis of business requirements with the SME's
- Data source & organizational readiness assessment
- Translation of business requirements into data modeling
- Definition of analytical applications to be implemented
- Decisions on the sourcing model
It is clearly identified by many BI thought leaders that BI professionals need to have both business & IT know-how and skills.
The success of a advanced business intelligence endeavor often depends on the ability to 'see the whole picture'.
In case of a Customer Intelligence project, combined skills of the project leader, on the identification of the appropriate marketing strategy and the implementation of the latter based on a successful design & implementation, allow him to identify issues early-on and steer the project successfully.
On the other hand, if we have to decide on the most critical success factor, business requirements analysis comes first. A leader with a clear vision of the business goals and the strategy to achieve them, can probably search and find the rest of the skills.
Some may argue though, that it's not possible to identify the full set of business requirements in advance, in a dynamically changing environment. Therefore the design of a flexible & extensible data model would allow future enhancements.