In the last decade, while the whole retail market has been sluggish, e-commerce (EC) has steadily expanded its market share. However, EC has a higher logistics cost than other retail formats, in particular, inventory management is a problem. In EC, the demand for goods tends to fluctuate greatly depending on its price because the price comparison in EC is much easier than that of other retail formats, and it makes the inventory management more difficult. Since the demand for goods fluctuates greatly due to price setting, EC companies suffer losses from the disposal of goods due to excessive inventory or from the loss of sales opportunities due to under stocking their inventory. Conventional methods, such as periodic ordering and quantitative ordering, are not appropriate to cope with the fluctuation of demand in EC. Thus, an inventory management method which can correspond to the fluctuation of demand due to price change is required. This paper proposes a demand forecast inventory management method which predicts demand by using planned pricing data and decides order timing and order quantity based on the prediction. This paper shows that the demand forecast inventory management method we propose can reduce inventory costs by 10% compared to periodic ordering by applying the method to the actual sales data and inventory data of an EC company.