Page 387 -

Chapter 9 Business Intelligence Systems
386
train salespeople on the best way to up-sell to customers. The third data mart is used to analyze
customer order data for the purpose of reducing labor for item picking from the warehouse. A
company like Amazon, for example, goes to great lengths to organize its warehouses to reduce
picking expenses.
As you can imagine, it is expensive to create, staff, and operate data warehouses and data
marts. Only large organizations with deep pockets can afford to operate a system like that shown
in Figure 9-12. Smaller organizations operate subsets of this system, but they must find ways to
solve the basic problems that data warehouses solve, even if those ways are informal.

Q9-4 How Do Organizations Use Reporting

Applications?

A reporting application is a BI application that inputs data from one or more sources and
applies reporting operations to that data to produce business intelligence. We will first summarize
reporting operations and then illustrate two important reporting applications: RFM analysis and
OLAP.
Basic Reporting Operations

Reporting applications produce business intelligence using five basic operations:

• Sorting
• Filtering
• Grouping
• Calculating
• Formatting

None of these operations is particularly sophisticated; they can all be accomplished using
SQL and basic HTML or a simple report writing tool.
The team that analyzed parts in Q9-3 used Access to apply all five of these operations.
Examine, for example, Figure 9-11 (page 379). The results are sorted by Total Revenue, filtered for
particular parts, sales are grouped by PartNumber, Total Orders and Total Revenue are calculated,
and the calculations for Total Revenue are formatted correctly as dollar currency.
These simple operations can be used to produce complex and highly useful reports. Consider
RFM analysis and Online Analytical Processing as two prime examples.

RFM Analysis

RFM analysis, a technique readily implemented with basic reporting operations, is used to ana-
10
lyze and rank customers according to their purchasing patterns. RFM considers how recently
(R) a customer has ordered, how frequently (F) a customer ordered, and how much money (M) the
customer has spent.
To produce an RFM score, the RFM reporting tool first sorts customer purchase records by the
date of their most recent (R) purchase. In a common form of this analysis, the tool then divides
the customers into five groups and gives customers in each group a score of 1 to 5. The 20 percent
of the customers having the most recent orders are given an R score of 1, the 20 percent of the
customers having the next most recent orders are given an R score of 2, and so forth, down to the
last 20 percent, who are given an R score of 5.
The tool then re-sorts the customers on the basis of how frequently they order. The 20 percent
of the customers who order most frequently are given an F score of 1, the next 20 percent of most
frequently ordering customers are given a score of 2, and so forth, down to the least frequently
ordering customers, who are given an F score of 5.