Abstract (EN)

The classic model of loyalty proposed by Dick and Basu (1994) assumes that customers can be naturally classified in four loyalty conditions. This model has been tested with cross-sectional data and measures of recalled or retrospective consumer loyalty. Therefore, these studies could not examine whether and why customers move across different loyalty conditions over time, and offer guidance to managers on how to shift customers to the more desirable loyalty conditions. In this paper, we conduct an empirical test of this model and examine the key drivers of shifts in consumers' loyalty conditions over six annual time periods. Based on data from 6,109 households and 23 stores, we find customers can be classified in three loyalty conditions only: (1) the no loyalty, (2) the latent loyalty, and (3) the true loyalty conditions. The spurious loyalty condition is not supported, probably because switching costs are negligible in the grocery retailing industry. However, we find that marketing actions, i.e., private label policy, feature advertising, end-of-aisle product display, and store pricing policy, influence customer transition across loyalty conditions.