The purpose of this study is to develop a new method for handling social influence in the analysis of shopping path data by using the ecological system of ants. Social influence which shoppers effect on each other is believed to contained in shopping path data collected in a retail store. The existing study measured social influence by using “density” which is the number of shoppers in a sales section at a certain point. However, the obtained result differed from the hypothesis of social influence in marketing area. One reason of this result is guessed that the density is calculated with considering temporal variation of social influence at the sales section. This study reviews the result of existing study by measuring social influence by using the concept of pheromone which has a close relationship with ants’ social behaviors. In this paper, the availability of the proposed definition of social influence is examined by focusing on sales sections with large variance of criteria.