Data-Driven Enterprises for Economic Recovery after Pandemic

A Study Case in Consumer-Packaged Goods Industry

Abstract

Businesses across Latin America are working towards economic recovery. Among the different strategies in place, organizations are transforming to data-driven enterprises to use analytics to achieve their goals in a variety of business processes. This Consumer-Packaged Goods (CPG) industry study case shows how business intelligence can help to reach their goals, applying business analytics and machine learning to forecast the demand. The study focuses in one product in one presentation, which has more than thirty flavors available sold by 275 stores across country. The study case discusses how enterprise current use of traditional forecasting methods such as autoregression and moving average are being used, the results they are getting and more important, what limitations and challenges they are facing.  The current situation is compared against the proposed machine learning based approach leveraging sales data from two whole years, data visualization was transformed from static and manual procedures into an automated continuous delivery approach with insights available near real time. Also, it is discussed why data granularity played an important factor to create the proposed forecasting models and how machine learning capabilities provide a better and more accurate results. Lastly, it is discussed the implementation of data-driven philosophy explaining what components were required, how they interact among each other, data gathering alternatives, what types of transformation were required, and what technologies were selected. This comprehensive study case can help other enterprises looking to start their quest to become data-driven organization relaying on business analytics to make better decisions during their efforts for economic recovery.