Analyzing and anticipating competitor moves has been central to modern competitive strategy. In contexts involving intense interfirm interaction, the value of a particular strategy depends in large part on how competitors will react to it. Despite many developments, anecdotal evidence indicates that the effective use of techniques to gauge decisions based on competitive considerations has been scant in practice. Our paper intends to fill this void. Using data from the auto insurance industry in Brazil, we compare strategies that do and do not anticipate competitor reactions. Basically we show that it does pay to anticipate those reactions. An optimal strategy will explore both demand elasticities and competitors’ patterns of reaction. We show that such “strategic” policy is expected to outperform a “myopic” approach which ignores competitor reactions. We also develop an methodology to compute demand elasticities, reaction functions and numerically compute optimal reaction strategies.