Multiproduct Advertising Budgeting
Abstract
The problem of optimizing advertising across a broad product range, where significant cross elasticities are likely, is explored. A linear-in-logs model is proposed and allocation rules for budgeting across products and media are derived. The model is estimated for a large European variety store chain. The results suggest profits could be increased by nearly 40 percent with no change in the advertising budget, if management shifted from the conventional bottom-up approach to the more systematic method outlined. The final section describes how management implemented the model.

