Abstract: With the liberalization of the policy, pharmaceutical e-tailing keeps surging forward challenging the status of traditional pharmacies. Unlike the daily use products sold by traditional e-tailing, pharmaceutical products sold by pharmaceutical e-tailing have extremely high brand concentration which means the difference between similar products is brand only. In this article, the decision making problem is studied from the perspective of pharmaceutical e-tailing in order to expand its market share.Customers usually are unwilling to go through all the website space of pharmaceutical e-tailing, but only browse the home page and make the purchasing decision. Therefore, how to optimize the website space especially the home page in order to expand the market is an interesting question. Comparing with the shelf space optimization in offline channel, the website space optimization in online channel is quite different. The traditional shelf space optimization regard all the positions in single shelf is homogeneous, while the website space optimization consider the perception of viewer to define the space of website especially the home page is heterogeneous. First the market share and substitutions between products are estimated through the demand exogenous model with the input of historical sales volume. Then we the market share and substitutions between products are regarded as the input of the website space optimization in order to generate the display list and order in home page of pharmaceutical e-tailing.In traditional research of category management, Multinomial Logit model which has vulnerability that it assumes the stochastic components of utility follow independent identical distribution is adopted. In reality, there are many factors which can simultaneously affect the utility and the market share of different choices. Considering the extremely high brand concentration of pharmaceutical products, Multinomial Logit model is invalid here. Therefore, the exogenous demand model which has more degree of freedom is employed to incorporate market variables into the consumers' buying behavior.Based on the shelf management model with slotting fee and the maximum potential market share obtained by exogenous demand model, ignoring the number of exhibition,an website space optimization model of integer programming based on an brand-new concept named the biggest market share is construeted to maximize the utility of the pharmaceutical e-tailing by deciding the sequence of posted products.Meanwhile, using the sales data of 111.com during August 1^st to October 30^th in 2015, an empirical study is conducted to investigate the potential performance of the proposed website space optimization model. The data is first sifted in order to get rid of the abnormal ones and input the normal data into the estimating model. Results demonstrate that the website space optimization model we proposed can significantly improve the sales of pharmaceutical e-tailing by 8.06% in average, and up to 27.21%. Although we lack of chance to evaluate the effect of changing the display number of product in home page, we theoretically measure that fewer display may also reach the profit that pharmaceutical e-tailing expects.

Key words: exogenous demand model, pharmaceutical e-tailing, category management, website space optimization