In today’s fast-paced global economy coupled with the availability of mobile internet and social networks, several business models have been disrupted. This disruption brings a whole list of opportunities and challenges for organizations and the domain of supply chain management. Given big data availability, data analytics is needed to convert data into meaningful information, which plays an important role in supply chain management. One of the disruptive data analytics techniques that are predicted to impact growth, employment, and inequality in the market is automation of knowledge work, better known as machine learning. In this paper, we focused on comprehensively overviewing machine learning applications in demand forecasting and underlying its potential role in improving the supply chain efficiency. A total of 1870 papers were retrieved from Scopus and Web of Science databases based on our string query related to machine learning. A reduced total of 79 papers focusing on demand forecasting were comprehensively reviewed and used for the analysis in this study. The result showed that neural networks, artificial neural networks, support vector regression, and support vector machine were among the most widely used algorithms in demand forecasting with 27%, 22%, 18%, and 10%, respectively. This accounted for 77% of the total reviewed articles. Most of the machine learning application (65%) was applied in the industry sector, and a limited number of articles (5%) discussed the agriculture sector. This paper's practical implication is in exposing the current machine learning issues in the industry to help stakeholders and decision-makers better plan transformation actions.