Consumer response to retail stockouts

Zinszer and Lesser pioneered research into the product characteristics and shopping situations as correlates of stockouts." Their model looked at how stockouts affect consumers of different demographic characteristics, whether the item was on sale and how the stockout affects store image and intended future patronage. Farris, Olver, and Kluyver developed a simulation model to show a positive, curvilinear relationship between distribution and market share." Brands with a larger share benefit more than brands with a smaller share when a small share brand is out-of-stock. This is because it was assumed in the simulation that retailers prefer to restock shelves with the best selling brand. This restocking policy leads brands with higher market share to attain better distribution which, in turn, contributes to further market share gains. This spiraling process accounts for the curvilinear relationship between distribution and market share. Private brands were not included in the simulation.

Over a period of four days, Emmelhainz, Emmelhainz, and Stock removed five items from the shelf of a discount grocery store. 6 These were selected from the best selling item of the leading brand in the following product categories: ground coffee, orange juice, toothpaste, peanut butter, and tomato sauce. Consumers were interviewed at the checkout lane about intended SDL and other behaviors following the stockout. Results were quite different for the five test items when compared to other items found out-of-stock in the store. The results are also included in Table 1.

Straughn was the first to use scanner data in a stockout study." She attempted to estimate the effects of stockouts on brand share for candy bars. The short-term effect was negligible. The long-term effect, defined as more than five weeks following the stockout condition, was substantial. Decline in brand share averaged 10 percent. Because the methodology used in this study was unique, it is necessary to review the key measures and their operational definitions. First, since the data were aggregated weekly, a stockout could be corrected within the same week and consequently not show in the weekly numbers. To solve the problem, a stockout was defined as a reduction in the weekly sales average for an item that fell outside a 95% confidence interval of a Poisson distribution. Second, the longterm effect of stockouts on brand share was defined as the difference between actual market share and the market share that would be observed in the absence of stockouts. The latter number was estimated from observed sales in weeks during which no stockouts occurred.

Using a two-firm game theory model, Balachander and Farquhar examined the conditions under which it would be profitable for firms to reduce availability and accept stockouts.18 They concluded that firms could actually benefit from stockouts (and thus have a vested interest in them) because the reduced product availability will restrict supply and, consequently, increase prices. This theory implies a deliberate conspiracy among competitors. Competitor A is able to sell product at a higher price when competitor B is out-of-stock. Competitor B expects a quid-pro-quo at a future date. This conspiracy also benefits the company that allowed the shortage to occur because incoming replenishment inventory can be sold at a higher price. These results counter the general body of literature on stockouts.