Flanking in a Price War

This study tested the notion that stock-up grocery goods would have a different pattern of price sensitivity than nonstock-up goods. We used covariance design within a Bayesian decision framework to select the optimum price treatment strategy as well as the dollar risk associated with this strategy. The Bayesian decision framework also provided an optimal stopping rule for the experiment. The test results were successfully employed in a real-life retail grocery setting when an anticipated price war occurred. A small grocery chain, rather than responding in kind to competitive price cuts, implemented a precise, profit-preserving counterattack. As a result, the chain increased market share substantially, at the cost of only 1.2 percent of its gross margin, during the price war.