Retail sales forecasting of agricultural products marts in a mid-sized city using the Huff model

Author: Pan-Jin Kim ¹ and Myoung-Kil Youn ²

Received 19 January 2010, accepted 26 March 2010.

Abstract

This study aimed to predict retail sales of agricultural products marts in mid-sized cities of South Korea in conjunction with the Huff model. The Huff model is widely used in retail sales forecasting because of its easiness of use and the accuracy of its predictions. To forecast sales volume using the model, the parameter λ should be estimated. Although several studies have reported the estimation of λ, its value is often assigned arbitrarily. Besides, no previous studies have conducted a prediction of sales volume with agricultural products marts. For this, first studies included the sales forecasting agricultural products marts in a mid-sized city, designated to be Daejeon and explored the idea that the parameter may be different as the size of the city differs. Consumers in a mid-sized city were less sensitive to travel time (or more sensitive to store size) than those in metropolitan areas. Our estimate of λ for the mid-sized city is 0.38, which indicates that store size has a greater effect than travel time, while the existing studies in the metropolitan areas report λ greater than 1, indicating the reverse condition. The implication from the results is to help locations decision makers estimate a potential sales volume in a mid-sized city especially in the Korean context.

Key words: The Huff model, the new probability model, agriculture product, agricultural product marts, the gravity model, top-of-mind brand recognition.