take great pride in offering comprehensive solutions to meet your business
objectives. Our approach is based on three central concepts: understanding
your corporate initiatives and architecting a custom strategy to achieve
them, identifying and integrating all necessary information into a unified
location, and providing the scientific tools and support to transform
your business information into business results. Wherever you are on your
path to success, JumpStart Point of Arrival can assist you in achieving
Business Case 1 (Response & Look-alike Modeling)
Objective The 4-Clover Entertainment group owns and operates twenty
different casinos throughout the United States. In an effort to increase
revenue, the group is preparing a promotional mailing that will offer
special incentives for spending a weekend at any of their locations in
Strategic Plan To minimize the cost associated with the mailings
and to maximize responsiveness, it was decided to selectively mail two
types of offers to three customer types. The first is the most aggressive
and will be directed towards "Best Customers" most likely to respond.
The second mailing is less aggressive and will be directed towards "2nd
Best Customers" most likely to respond. The second mailing will also be
mailed to prospective customers that resemble the current "Best Customer"
Data Integration The entertainment group already has its customer
data located in a centralized data warehouse for analysis. This data warehouse
is supplied by customer information obtained from the "Players Card" and
other transactions that occur during the stay. To increase the predictive
power of a response model, enhancement data was obtained from a 3rd party
list vendor and appended to the customer database.
Scientific Analysis (Database Marketing) A factor analysis was
performed to separate the customer database into several segments defining
the consumer behavior at the casino. A Response Model was then built for
the "Best" and "2nd Best" segments to predict those most likely to accept
the promotional offer. Finally, a Look-alike Model was built to locate
prospects from the list vendor possessing similar qualities as 4-Clover's
Business Case 2 (Business Rules & Fraud Reduction)
Objective A retailer is experiencing a high volume of fraudulent
orders from its website and would like to cut the number of processed
orders of this type in half.
Strategic Plan The retailer would like to implement new screening
procedures during processing to flag suspicious orders. They would also
like to develop some internal negative lists that will flag orders with
matching demographic information. The flag will result in a terminated
order in extreme cases, and a manual review process for lesser cases.
Data Integration The orders along with the necessary customer information
were relayed through a SAS credit screening system that produced a flag
on identified orders. This flag was then appended back to the original
order data so that a final decision could be made. Reports were also generated
compiling all historical information to build cases against fraudsters.
Scientific Analysis (Risk Management) A sample of subsequent online
orders from a 30 day window was analyzed for fraudulent behavior patterns
using various data mining techniques. Business rules were formulated as
a result of the study and implemented into the credit screening system.
The company's current fraudulent population was then analyzed and, as
a result, various negative lists were created and implemented into the
credit screening system.