This data set contains 1 year of sales data of a Polish grocery store taken from the Point of Sale system. The mom and pop owner of this store told us that revenue and profitability has been declining consecutively for the past few years. As such, the overall business objective is to make the store cometative again. We will approach this in 2 parts.
The first part is aimed at gaining business understanding through EDA and market analysis.
I will then be implementing a recommender system to generate related product recommendations for end-of-isle and in-isle promotions to increase cross-sell.
This recommender system will also be deployed as API using Docker - making it available for store's eCommerce website.
Main result of the EDA includes the identification and recommended removal of underperforming SKUs that could free up ~31% of the total working capital. This cash inflow can then be used on other investments, such as purchasing new inventory for the repositioning to a speciality retailer.
Please view the JupyterNotebooks for details.
Part 2: Combining Business Understanding Through EDA With Strategic Analysis To Recommend A Repositioning
Below is the presentation on recommendations and the strategic rationale
