Sorry, Your Order Has a Substitution : The Effects of Substitution Policy in Online Grocery Retailing

Post-purchase out-of-stock (OOS) often happens in an online store context, where products appear to be available at the time a consumer makes an order and checks out, but then become OOS when the order is to be dispatched. To mitigate negative responses from consumers, online grocery retailers often provide consumers a substitution alternative to the OOS item. This paper investigates the effects of two substitution policies where we focus on different matching strategies of the substitution with the OOS item. In policy one, we measure the effect of matching on the dominant attribute (brand vs. flavour). In policy two, we test the effect of matching with a product from the consumers’ past purchase portfolio. We investigate these two substitution policies and their interaction in two categories that differ on the level of differentiation (i.e., the degree to which distinctions are objectively measurable – vertical differentiation/VD vs. not easy to evaluate – horizontal differentiation/HD). Our dependent variable is the probability to accept the substitute. The study employs a computer-simulated purchase experiment, using two product categories: margarine (VD) and cereals (HD). 2,113 UK consumers representative of general UK shopper profile participated. Findings show that in the margarine category where brand is the dominant attribute, the same brand substitution is more likely to be accepted than the same flavour substitution. In contrast, in the cereal category where flavour is more likely to be the dominant attribute, same flavour substitution is more likely to be accepted than same brand substitution. The results also show that, in both categories, matching the substitution product with a product from consumers’ past purchase portfolio is more likely to be accepted than offering a substitute that consumers have not bought before. We also found a significant interaction between the two policy types but for cereals only. The effects of two substitution policies are mediated by perceived fairness of the substitution. The paper discusses contributions and implication for future research.