A Method of Detecting Competitors of a Store and Evaluating their Attractiveness in Micro Scale Retail Environment

Abstract

The availability of data set itself and the method of handling the data set are essential factors to achieve effective decision making in retail marketing, especially for small privately managed retail stores. In the field of retail and marketing geography, attractiveness of a store is analyzed with questionnaire survey and logit model in general. However, the problem is that huge cost is required for the process, which becomes more difficult under a micro scale retail environment. Considering the fact that small retail stores often have their own purchase history data of customers, the cost is reduced if an effective method exists that is applicable using purchase history data only. Hence, the purpose of this paper is focused on evaluating and detecting competitors of a small retail store, with limited data availability, to understand the status of its market area from the geographical aspect by clarifying where its own strong competitors are located. The method uses a probabilistic store choice model, which is usually estimated from the data of consumers' store choice behavior collected by a questionnaire survey. Instead of such detailed but costly data, this paper uses actual sales data of the target store to estimate the model. The method is applied to simulated choice behavior, then, to actual data based on the findings understood from the simulated choice behavior analysis. Results show that the method effectively detects strong competitors of a store. Furthermore, the result may provide a guideline for the store to establish an effective marketing strategy.