Challenge

The client was looking to cater to customer demand better by improving their Demand Forecasting capability. Their existing demand forecasting system is a black box with limited accuracy of 10-20%, and this resulted in lost sales and high inventory cost. They were looking to get adequate visibility into the Demand Planning process and obtain higher accuracy. NPI - In the Apparel and Footwear Industry, 70% of the styles contribute from NPI (New Product Introduction). These styles don't have a history or there is a shortage in history to generate appropriate forecasts. Styles having similar attributes to the NPI styles are picked up in generating forecast. In the current setup, the process was carried over through a legacy system

Approach

Rigorous sessions with the Demand Planning Team to understand current setup of NPI. Choosing of product attributes in picking up appropriate LikeStyles based on various filter conditions which the client expected.

Outcome

We were able to generate forecast for the New Products based on the similarity score of the historical products.