Retail market area analysis using a transportation network with consideration of population mobility
Krivacsy, Kevin Russell
MetadataShow full item record
Gravity models and location-allocation techniques are fundamental to analyzing current retail location’s market areas as well as to conducting retail site suitability analysis. Several current research directions are using fundamental geographic location analysis principles and incorporating some of the fundamental principles from other disciplines such as marketing, real estate and urban planning. Examples of such studies establish models of retail site selection to maximize profits or to minimize costs. An important cost factor is the distance from the customer to the retail site. While most prior studies use Euclidean distance for this type of modeling efforts, most researchers agree that Euclidean distance is not an accurate measure. One objective of this research is to implement a methodology and conduct a case study for constructing a retail analysis along a road network rather than using the more common and less precise Euclidean distance. This study looks at whether the difference between the two methods is significant. The results of this inquiry are mixed in that although there does not exist a significant difference between the two approaches in many of the cases, significant differences were observed in some of the cases. Differences between approaches are most apparent in sudy areas with a less complete road network. Another goal of this study was to develop a methodology to include aspects of population mobility, such as rates of disability, levels of car ownership and the use of alternate transportation means into the retail market area analysis. This is an exploratory investigation into one of many other factors that are typically not taken into consideration from other studies with the goal of creating retail market areas. The purpose of this is to see how the inclusion of other factors of population mobility affects the space and distance that each retail location can realistically draw its customers from. Spatial regression analysis was run in Geoda to explore the relationship between the amount of sales for an individual grocery store location and the surrounding socioeconomic variables. The dependent variable was sales and the independent variables were: white population, black population, Asian population, Hispanic population, vacancies, owner occupied housing, renter occupied housing and median household income. We typically found that there was a significant relationship between sales for the stores and median household income, black population and owner occupied housing among others results subsequently discussed.