Browsing by Subject "Vähittäiskauppa"

Retail location analysis has been a widely researched topic during the last decades, and models able to estimate consumer and purchasing power flows are valuable to retailers and investors. As the retail market is getting increasingly competitive and consumer habits are changing, the demand for sophisticated modeling techniques remains high. Spatial interaction models (SIMs) have been proven effective for simulating consumer behavior and estimating store revenues in a high degree of accuracy. This master’s thesis examines the suitability of spatial interaction models for simulating the grocery market in Helsinki metropolitan area (HMA). Three iteratively calibrated SIMs are developed, and generated revenue estimates are compared to each other and actual revenue figures of HMA grocery stores. Based on the modeling results, factors preventing more accurate results are identified and suggestions are given for future model development. The study focuses on year 2019, and main datasets used are Nielsen grocery store register and Statistics Finland’s Grid Database. Travel times between supply and demand zones of the model are based on Helsinki Travel Time Matrix 2018 by Digital Geography Lab. The developed models can forecast revenues of the 323 studied stores quite accurately, and in the best case over half of the revenues are forecasted within a 25% error margin. A high coefficient of determination is achieved even with a simple model, and the disaggregated versions further improve the results. The models estimate the revenues of large hypermarkets the best, while there is more variance in the estimates for smaller stores. The results indicate that a spatial interaction model suits well for modeling the grocery market in the study area. The lack of empirical consumer flow data did not prevent the calibration of the model, and passable results were achieved with the iterative calibration approach. However, the developed models remain theoretical due to the lack of empirical datas. This, in addition to other results of the study, underline the importance of empirical calibration data when developing robust modeling solutions. If suitable empirical data was available, combining it with highly granular demographic data such as Grid Database, might enable very accurate modeling of consumer flows. Models able to consider current changes in the grocery sector, such as e-commerce and the refurbishments, could be valuable to both scholars and commercial operators in the retail sector.