A influência do big data analytics na previsão de demanda: um estudo de caso na Amazon

Abstract

Contemporary supply chain management faces the challenge of operating in volatile markets, where traditional demand forecasting methods, based on historical time series, prove insufficient to avoid inefficiencies such as excess inventory and the Bullwhip Effect. In this context, Industry 4.0 and Big Data emerge as transformation vectors, promising to replace speculation with precise information. This work aims to analyze the influence of Big Data Analytics on demand forecasting and operational efficiency, using Amazon.com, Inc. as a case study. The research, of applied nature and qualitative approach, was conducted through a documentary case study, based on the analysis of technical literature and recent scientific articles. The results demonstrate that the application of predictive algorithms and the integration of unstructured data (Volume, Velocity, and Variety) allow overcoming the reactivity of classic models. The Amazon case, through the practice of Anticipatory Shipping, proves that real-time visibility enables the transition from a "push" production model to a predictive and "pull" model, resulting in the breaking of the historic trade-off between cost and service level. It is concluded that Big Data acts as an enabler of "demand sensing", although its implementation imposes significant barriers regarding technological infrastructure and data privacy.