E-commerce is using the ABC analysis to define stock levels, as a quick and easy evaluation of all goods in the warehouse. For each of the categories, the company is using different inventory policies.

Not considering the different patterns of demand, at the end of the year company has a high level of stocks.

The company would like to redefine inventory policies and get rid of the articles with unpredictable demand that don’t bring high revenue.

First, we separated articles into three categories based on ABC analysis to represent the current situation, based on which inventory policies have been made.

We added fluctuation of demand to the inventory classification process. With ABC XYZ analysis we got nine categories of articles. All articles from the CZ category will be put on a flash sale and removed from the assortment. This is the practice that the company will do on a yearly basis.

Next, we simulated multiple different inventory policies for each article in the AX category, as the most important one. The same is done for the rest of the categories and an optimal policy per each article is established. With that, predicted total stocks at the end of the next year will be lowered by 24%.