Using Mathematical Programming For Marketing Plan Optimization

Abstract

While a growing number of companies are embracing analytical metrics to measure the performance of their critical business processes, many neglect to apply the same mathematical rigor to their marketing efforts. This paper examines how to apply a disciplined quantitative approach to resource allocation within the context of marketing. This project will examine a company that manufactures and sells electric power generation components. This firm’s executive management believes that developing nations in Asia have a particularly strong need for power generation equipment as they build and expand new infrastructure at a rapid pace. Management objective is to construct a marketing plan that will maximize both total revenue as well as revenue from a targeted geographic region, while working within the budgetary and other constraints. Operations Research offers sophisticated tools for translating complex business situations into mathematical expressions. These expressions may then be evaluated to identify one or more optimal solutions. Multivariate Testing is used to quantify all the relevant variables for the mathematical model. This statistical tool allows the researcher to simultaneously test the effect of various input factors on the response variable, and also test for any interactions between factors. These interactions are often significant in predicting the response outcome. Using indexed utility values produces a blended model that can simultaneously satisfy both objectives. This blended model results in a marketing plan that will increase total revenue by more than five million dollars, a 57% gain. At the same time, the plan will generate an 18% increase in targeted revenue, positioning the company for future growth. The project shows that companies can, and should, insist on the same degree of financial accountability from their marketing investments that they would expect from any other capital or operational expenditures.