For historical reasons, optimization has traditionally been slightly outside the mainstream of system dynamics. However, computer technology has made both quantitative data more abundant and optimization more feasible. At the same time, modelers are encountering real situations and clients with high-stakes questions that are nearly impossible to answer without optimization—the systems involved are not only dynamic, and not only highly interconnected, but also combinatorially daunting. In particular, corporate marketers currently make allocation decisions impacting billions of dollars of shareholder value on the basis of intuitive “anchor and adjust” strategies, which can be far from optimal. This article presents an anonymized case study of one such situation. The company is in a high-tech industry undergoing rapid change. The company needed to fashion a go-to-market strategy balancing traditional and unfamiliar markets, an important component of which was allocation of marketing resources. Optimization revealed a potential valuation increase of roughly 30 per cent relative to executives' intuitive allocations. Scenario analysis revealed the basic policy direction (more advertising) to be robust, and in the process resolved several traditional conundrums in dealing with adverse events in the marketplace. Copyright © 2003 John Wiley & Sons, Ltd.