Introduction to the IBM optimization subroutine library
IBM Systems Journal
Future Generation Computer Systems - Special double issue on data mining
| Hi-index | 0.00 |
We present an algorithm for performing multi-channel marketing event optimization. Previous related work, which made use of a complex mixed-integer linear program to generate a marketing plan, was capable of providing plans for only groups of individuals, referred to as micro-segments, and considered only the direct mail channel. Today, most firms use multiple channels, such as e-mail, call centers, and direct mail, to contact customers with marketing events. Our method successfully overcomes many past restrictions and reduces the run time for the marketing scenarios so that computation can take place on a daily basis. The algorithm used is an advanced form of a greedy heuristic which--given a set of marketing events by channel, a set of individuals, some constraints, and the concepts of saturation and cannibalization--determines the optimal set of marketing events to present to an individual customer. The algorithm is embedded in a solution that is designed to operate as an interactive what-if scenario planner, or as a batch-oriented job that can continually maintain each customer's future contact plans in an optimal fashion.