AMOS - A PROBABILITY-DRIVEN, CUSTOMER-ORIENTED DECISION-SUPPORT SYSTEM FOR TARGET MARKETING OF SOLO MAILINGS

We describe a state-of-the-art, probabilistic-based system, nicknamed AMOS, to support the marketing decisions of Franklin Mint, a leading P hiladelphia-based worldwide direct response marketer of quality collec tibles and luxury and home decor products. The purpose of the system i s to target the ''right'' audiences for each promotion from among a ve ry large marketing database, based on the customers' attributes and ch aracteristics. Due to the dimension and complexity of the decision pro blem, AMOS was developed as a hybrid system involving expert system co ncepts to address semi-structural and ''fuzzy'' issues and traditional optimization and statistical models to address well-defined structura l issues and decision problems. The system has been implemented in 199 2, and currently constitutes the primary promotion selection for the F ranklin Mint, applied for all solo mailings in the US which meet minim um size requirements and statistical validity.