Perhaps it’s just coincidental, but several people have asked a similar question recently. “Should the DW or BI team gather requirements from the business?” Honestly, this question makes the hair start to stand up on the back of my neck. I’m concerned that too many organizations have overly compartmentalized their data warehouse and business intelligence teams.
Of course, some of this division is natural; especially when the resources allocated to DW/BI grows as the environment expands, creating obvious span of control issues. Also, separation of labor allows for specialization. Viewing the overall DW/BI environment as analogous to a commercial restaurant, some team members are highly-skilled in kitchen food preparation while others are extremely attentive to the needs of the restaurant patrons, ensuring their return for a subsequent visit. There are likely few waiters that should suddenly don the chef’s garb, and vice versa.
Despite the distribution of responsibilities, the kitchen and front rooms of a restaurant are tightly entwined. Neither can be successful on its own. The best chefs need a well-trained, well-oiled front room machine; the most attractive dining room requires depth and quality from the kitchen. Only the complete package can deliver consistent, pleasurable dining experiences (and sustainability as a restaurant). That’s why the chef and wait staff often huddle to educate and compare notes before a meal rush.
In the world of DW/BI, we’ve observed some teams take a more isolationist approach. Matters are further complicated by the complexities and realities of organizational culture and politics. There may be a kitchen and dining area, but there’s no swinging door between the two. It’s like there’s a transom (above eye level) where orders and plates are flung back and forth, but the two
teams of specialists aren’t communicating or cooperating. In this scenario, you end up with data models that can’t reasonably be populated. Or data models that don’t address the diners’ needs and/or leverage their tools. Or diners’ tools that are overtaxed or slow-performing because they’re repeatedly doing the work that could have been done once in the kitchen and shared throughout the organization. In the worse case, the wall becomes so impenetrable that the BI dining room substitutes a different kitchen (or creates their own) to source meals.
The data warehouse should be the foundation for effective business intelligence. Too many people have focused on one without the other. Sure, you can create a data warehouse without concern for business intelligence, and vice versa, but neither situation is sustainable for long. Isolationism is not a healthy approach for building and supporting the DW/BI environment. Even if you don’t report into the same management structure, collaboration and communication are critical.