This thesis studies effects of shelf layout in a department store on buyer behavior and then extends the research to cover the similar issue in layout of e-bazaars. Shelf layout –in terms of parameters which influence buyer behavior- is formulated by two variables; pattern of prices and pattern of popularity. In measuring both of these variables, a weighting system should be used to fine tune effect of distance between customer and different items in the layout which are also in the eyesight of the customer. In this study, weights are manipulated in such a way to optimize prediction power of the model. As a result, these weights determine an average span of vision for buyers. This measure can be used then in analysis to improve shelf layouts. Based on the analysis, we came to conclusion that in our case study –Shahrvand department stores – the first factor is the key determinant in customer buying decision. A rather similar result was obtained in e-bazaars case but with much greater lack of fit measure. At the end, a set of recommendations for shelf layout is developed based on the results of the research.