We took a four-phased approach to this challenge:

                              • Data from disparate?sources were consolidated into a single data warehouse, improving the usability of data
                              • The analytical data layer was then prepared after carrying out data treatment procedures and applying business rules that were appropriate for the client’s business
                              • Elementary data analysis helped classify stores based on potential products they could house
                              • A regression model was then applied to identify the impact of changing assortment quantities on the top-line sales and the correlation model helped identify the af?nity between different products
                              • The combined insights from the models helped us arrive at the optimal assortment strategy for the client.

                              KEY BENEFITS

                              • Our easy-to-use solution enabled the client to identify closely related products and plan assortments accordingly
                              • It included recommendations on the mix of products that a store should carry and store-level revenue prediction based on the assortment mix


                              Our collective efforts paid off when the client’s Stock Keeping Unit (SKU) level prediction enhanced the weekly sales by 6%.