Consumer response to retail stockouts

The conclusions and managerial implications to be drawn from this research cover five main issues: the link between a stockout experience and store image, the lack of convergence in previously published research, the non significant results obtained for demographic variables as correlates of SDL behavior, the impact of the variables Store Prices and Surprise on the Delay and Leave behaviors and, finally, the predominance of situational variables among the correlates of SDL behavior.

We showed in the early part of the results section that consumers are generally able to separate a single and recent stockout experience from other dimensions of their overall attitude toward a store. If confirmed by future replications, this result casts doubt on the general perception that stockouts significantly damage a store's overall image. With the exception of the questions about merchandise quality and product availability, both of which were rated lower by consumers who experienced a stockout, the overall perception of the store was unaffected. The managerial implication seems to be that a single stockout has a lesser impact on a store's overall image than previously thought.

With respect to the second issue, recall that Table 1 showed that the results published in four previous studies of SDL behavior did not converge. The table reported the percentage of times consumers reacted to stockouts by electing each of the SDL behaviors. The SDL percentages reported in this study also show little convergence with previous ones. This lack of convergence may lead to two alternative streams of further research. The goal of the first one is to correlate SDL behavior with independent variables. The results reported in this study fit into this research stream. Replications of this type of research may eventually lead to an understanding of SDL behavior that is valid for multiple consumers, firms and product lines. The second research stream assumes that such a general understanding is not attainable and will consequently propose that the output of research should be a methodology that individual firms may use to determine the SDL behavior of their own consumers. In this case, firms may benefit from this research by applying the same methodology to collect and analyze data. Results should be considered unique to each firm and not expected to be generalized.

Third, results relating SDL behavior to demographic variables were not significant. The implication is that management does not profit from fitting their reaction to a stockout according to a consumer's demographic characteristics. The fourth issue is that there is an important difference in the pattern of results obtained with the multinomial logit model for each of the SDL behaviors. Four variables emerged as significant correlates of the Substitution behavior. However, the coefficients obtained for each of the four variables were relatively small, ranging from 2% to 5%. This suggests a diffused pattern where none of the four variables have an outstanding impact on a consumer's decision to substitute the out-of-stock item.