Databases Reference
In-Depth Information
that case, a single Market Basket BI Table can be leveraged for two or more
exception reports.
Market Basket KPI Exception Reporting
The Market Basket KPI Exception Reporting application uses a SUM and
GROUP BY query, which was defined as a K PI during Market Basket
Analysis. During analysis, an analyst discovered an aἀ nity that confirms
the current state of the enterprise or predicts a future state of the enter-
prise. That aἀ nity was defined as a query, for example:
• What is the aἀ nity between motor oil and oil ilters?
• What is the aἀ nity between hamburger meat and hamburger buns?
• What is the aἀ nity between steak and dessert?
Each of these queries will deliver an aἀ nity. If that aἀ nity falls below a
threshold value, that is an indication that the enterprise needs attention in
that area. For example:
• What is the aἀ nity between motor oil and oil ilters?
• Good—between 50% and 75%
• Bad—between 40% and 50%
• Red Alert—below 40%
• What is the aἀ nity between hamburger meat and hamburger buns?
• Good—between 70% and 95%
• Bad—between 50% and 70%
• Red Alert—below 50%
• What is the aἀ nity between steak and dessert?
• Good—between 80% and 95%
• Bad—between 70% and 80%
• Red Alert—below 70%
The exception reporting method leveraged by the Market Basket KPI
Exception Reporting application can be the standard method used by
the data warehouse. There is no need to invent a new wheel for excep-
tion reporting if one already exists. If, however, the data warehouse does
not already have an exception reporting method, then the creation of an
exception reporting method should be a prerequisite for the implementa-
tion of a Market Basket KPI Exception Reporting application.
