Omni-channel Value of Mobile Application Usage in Off-line Retail Store: Focusing on the Use of Brand App of a Coffee Retailer

Purpose: In this study, using a marketing research company’s consumer panel data of consumer’s mobile applications (apps) usage, off-line location and payment, we analyzed the impact of consumer’s in-store mobile application (app) usage time on the store’s sales (i.e., consumer’s payment amounts). While lots of previous studies have analyzed the effects of online and offline channel operation at the retailer level rather than the consumer level and with a cross-channel perspective rather than an omni-channel perspective, we analyzed the effects of online and offline channel operation at the consumer level with an omni-channel perspective. Research design, data, and methodology: We choose one of the biggest coffee retailer who has successively introduced its own mobile app in its retail operation. In any off-line store of the coffee retailer, consumers can use the mobile app for various omni-channel shopping purposes including in-app information search, in-app order, and in-app payment. Consumers can make an in-app order while he/she is in the off-line store or just before he/she visits the off-line store, and can pickup his/her order at the off-line store’s counter including drive-through. By combining three kinds of panel data (app usage, off-line location, and payment), we traced 227 consumer panels during the two years of 2022 and 2023 to figure out how much time they used the app in the off-line store and how much they paid at the coffee retailer. Then we analyzed the impact of customer’s omni-channel shopping behavior (which was manipulated as the time spent by panel consumers using the mobile app while staying in the off-line store) on the payment amounts. To analyze the causal relationship between the in-store app usage time and the payment amounts, we applied a latent class regression analysis to reflect the consumer heterogeneity in the omni-channel shopping behavior. Results: Consumers can be divided into three latent classes (or clusters) according to the level of omni-channel shopping behavior : traditional off-line channel oriented consumers (LC1: latent class 1), on-line channel oriented consumers (LC2: latent class 2), and customers using both on-line and off-line channels together (LC3: latent class 3). While the percentage of consumers with omni-channel shopping behaviors is not high, the amount spent of these omni-channel consumers is higher than the traditional off-line channel oriented consumers. Our results also show that in-store app usage time had a statistically significant positive (+) effect on the payment amounts in LC2 and LC3. Conclusions: The implications of this study can be summarized as follows. First, this study confirmed the value of omni-channel shopping at the consumer level by empirically showing the significant effect of consumers’ in-store app usage time on the payment amounts. Second, in order to reflect the consumer heterogeneity in the omni-channel shopping behavior, we conducted a study that applied latent class regression analysis to classify consumer classes (or clusters) according to the levels of omni-channel shopping behavior and integrated causal analysis with the latent classes (or clusters). Finally, the analysis results of this study can also be meaningfully used to establish retail strategies to increase omni-channel shopping value by actively utilizing mobile apps while consumers stay in off-line retail stores.