The marketing mix that a manager may deploy can affect the sales of a product and can be categorized under the traditional four Ps of marketing (product, price, promotion, and placement). But the perennial question managers face concerns the combination of these different marketing mix variables that will give them maximized sales, highest share, lowest inventory, or maximized margins. Quite often, these questions are answered by historical data: for example, past sales or market share for different levels of expenditures on these marketing mix variables. In this note, we consider the design of models that allow managers to obtain robust price and advertising elasticity estimates.