The full-time Kelley MBA Program has consistently been ranked among the top 25 MBA programs in the U.S. Indiana’s business course work emphasizes teamwork, skills and knowledge integration, leadership and service learning — all necessary tools to succeed in the new global economy.The Marketing Concentration at Indiana University focuses on the study of all activities required to determine which products and services to are most desired by consumers as well as the how to design and implement programs to efficiently communicate and distribute products and services.

PolyAnalyst PRO for Windows NT is being used as the main data mining tool for the course Marketing Intelligence Management (M549) taught by Prof. Raymond R. Burke to the second year MBA students of Kelley School of Business at Indiana University.

Course background: Marketing Intelligence Management. Marketing research is in a state of transition. New technologies including point-of-sale UPC scanners, frequent shopper programs, credit/debit cards, in-store tracking, caller-id systems, and electronic shopping interfaces have created an explosion of data on customers and their purchase behavior. The Internet and on-line retrieval systems have opened the door to volumes of secondary research on market trends and the competitive environment. Syndicated databases provide detailed information on the buying habits of individual households. The task of analyzing marketing data is no longer relegated to research specialists. Managers are now using networked, desktop computers to process this information. However, more computing power and data will not necessarily lead to better decisions. Managers must understand what research is available, what assumptions underlie the research methods, how to analyze and interpret the data, and how to apply the results in a business context. The goals of this course are to introduce students to the emerging field of marketing intelligence management. The course will discuss three major types of customer and competitor databases: geodemographic information systems, point-of-sale scanner databases, and catalog and frequent shopper databases. We will review existing data analysis tools used to manage the data explosion, track sales and market share, and diagnose changes in product performance. We will apply data-mining software to profile customer segments, analyze promotion effectiveness, and identify cross-selling opportunities. We will discuss recent innovations in tracking customer behavior and measuring customer value. The course concludes with a discussion of how to create a “knowledge-based organization” which incorporates marketing research into a firm’s ongoing operations to build a closer relationship with customers.

Full text of the review: I am very enthusiastic about the PolyAnalyst package for both education and commercial research. The software provides a unique and powerful set of tools for data mining applications, including promotion response analysis, customer segmentation and profiling, and cross-selling analysis.

PolyAnalyst provides several different methods for data analysis. The “Find Dependencies” tool sifts through large quantities of data in order to remove outliers and create a new, cleaned dataset. The “Stepwise Linear Regression” module uses a stepwise procedure to select those variables from a large set of alternatives which have significant linear relationships with the dependent variable. Unlike most other statistical packages, PolyAnalyst automatically dummy-codes categorical variables.

The “Find Laws” engine is perhaps the most powerful procedure in the PolyAnalyst package. It can identify complex, nonlinear relationships that are missed or incorrectly specified by existing, statistical methods. Unlike neural network programs, PolyAnalyst displays a symbolic representation of the relationship between the independent and dependent variables. This is a critical advantage for business applications, because managers are reluctant to use a model if they don’t understand how it works.

To help users visualize this relationship, PolyAnalyst can plot the estimated function with the “Rule Graph” procedure. Users can then run “what if” scenarios by manipulating the values of the independent variables with sliders and observing the effect on the dependent variable. PolyAnalyst has a simple and intuitive user interface. Datasets, rules, charts, and reports are represented as icons on the desktop. The user invokes analysis procedures by selecting options from the Windows toolbar and menus. Students find the object-oriented interface to be easy to learn and use. This is especially important in business classes where the focus is on data analysis and interpretation rather than on programming. Students are using PolyAnalyst PRO installed in Kelley School of Business on 75 PC’s for in-class workshops and individual data analysis. The main projects carried out during the course include the analysis of:

Marsh Superstudy program performed by Marsh Supermarkets, Inc. The effectiveness of the Anheuser-Busch beer promotion program over a retailing chain in a particular city. Possible ways to improve the Rodale Press’ cross-selling initiative.Design and effectiveness of the Simon DeBartolo Group’s MALLPeRKS program. MALLPeRKS is a proprietary customer incentive program that targets high-dollar, high-volume shoppers to encourage customer loyalty, stimulate sales, and create a valuable database of customer information.

Raymond R. Burke
E.W. Kelley Chair of Business Administration
Kelley School of Business
Indiana University