Define: Project template elements from here.
Most importantly: The goal, the reports required, the data points and their sources and business names, the benchmark report or query (the current status).
Make a brainstorming session for requirements and include all stakeholders! document requirements and get all approved by all, otherwise it will make conflicts and reduce trust and adoption of BI.
– Sketches of the Dashboards: Validate them with the stakeholders and end-users.
– Architecture Diagram: of the status-quo and the to-be status.
Analyze: Roots of variances between the definition of the project and what was found in measurements/profiling. a meeting is needed to analyze and find the roots and decisions to treat such variances.
1. *Query containing all fields required, normalized (fact) / denormalized (dimension).
Add calculated fields (fact:total amount, dimension: full name from first & last names). Default row/value if no
2. *DW structure (may be done within the 3rd step: ETL package): Include Add constraints (PKs, FKs, indexes), lineage columns (surrogate keys, Ts_Id, last_processed),
3.* ETL package.
4. ETL job.
5. *Dimensional model (SSAS, PowerPivot, etc.)
6. Dimensional Calculated measures.
7. Reports needed or demonstrating the data points developed.
Steps marked with asterisk (*) are to develop for each DW entity, so development time in DW project may vary a lot based on their number.
Demo: describing the data model and show casing reports that consume it.
Control: Long term plan:
Spread the word.
Increase the engagement.
Usage evaluation – Get continuous feedback from the BI solution. Handling new features and change requests policy.
Ensure smooth and beneficial usage of your BI solution.
Maintain security and access model.