A Method for Data Warehouse Design - Data Warehouse Systems: Design and Implementation - page 386

Identify Users
In this example, three groups of users were identified:
1. Executive: the members of the board of directors of the Northwind
company who define the overall company goals.
2. Management: managers at departmental levels, for example, marketing,
regional sales, and human resources.
3. Professional: professional personnel who implement the indications of the
management. Examples are marketing executive ocers.
Determine Analysis Needs
This step starts with the specification of the goals. We will just address the
general goal: increase the overall company sales by 10% percent yearly .This
goal can be decomposed into subgoals :
1. Increase sales in underperforming regions.
2. For customers buying below their potential, increase their orders (in
number of orders and individual order amount).
3. Increase sales of products selling below the company expectations.
4. Take action on employees performing below their expected quota.
In the next step, further sessions with the users are carried out to understand
their demands in more detail and operationalize the goals and subgoals. As we
explained above, the queries can be expressed in free natural language. Then,
the terms must be aligned with a data dictionary or common vocabulary
during a process of cleansing, disambiguation, and prioritization. Below, we
give some examples of the queries that operationalize the goals above. We
show the queries already expressed in a common vocabulary that we assume
has been previously defined in a data dictionary. We omit here the process
of prioritizing queries and users. We then identify potential dimensions,
hierarchy levels, and measures. To facilitate reading, we use different fonts
for dimensions , hierarchy levels ,and measures .
1. Increase sales in underperforming regions:
(a) Five best- and worst-selling (measured as total sales amount) pairs of
customer and supplier countries .
(b) Countries , states ,and cities whose customers have the highest total sales
amount.
(c) Five best- and worst-selling (measured as total sales amount) products
by customer country , state ,and city .