This work describes the efforts of the leading Mexican convenience store company to look for opportunities to reduce a high level of customer service variability at the store echelon. The company was having high customer service variability originating that 42% of the time, on average, stores were out of stock. After an analysis of the structure of the distribution network and the actual forecasting and inventory management schemes, the authors found that the main cause behind the deficient performance was poor demand forecasting accuracy and incomplete inventory management scheme structures. Therefore, the improvement initiatives designed were developed to select correct or adequate forecasting procedures for the different demand patterns and the modification of the structure of the inventory schemes. Initial results are obtained from a pilot study carried out in all SKU´s of all the stores of the Puerto Vallarta plaza of the company. The resulting forecasting mean squared error (MSE) was decreased significantly in the order of 43%. The firm estimates a reduction of about 21% of the level of inventories and of 22 percentual points of customer service variability at the retailing stores due to the improvement in forecasting accuracy.