Consumer Behavior Analysis in a Physical Store Using Video Recognition AI

Point of purchase (POP) data have been studied for many years as an important advertising tool. However, the data used for evaluation are mainly from point of sale (POS) surveys and questionnaires, and there are issues with the completeness and objectivity of the information. In addition, since most of these data are obtained after a purchase decision, consumer behavior before the purchase decision remains a black box. In this research, we used video recognition AI to analyze consumer behavior, and examined the impact of digital signage, an electronic POP item, on consumer behavior before purchasing decisions. We collected and analyzed data from over 10,000 consumers in a physical store experiment. The results showed that the POP presence promotes contact with products in the corresponding category. By using video recognition AI, we were able to quantify consumer purchasing behavior such as stopping by a shelf, looking at a shelf, and reaching for a shelf. These results suggest the possibility of customer journey management in physical stores.