Analyzing market dynamics and optimizing sales performance with machine learning

In today’s rapidly evolving business landscape, understanding the market is crucial for maximizing sales. However, the overwhelming amount of market factors to consider makes it extremely difficult for companies to pinpoint the key drivers of sales performance. Focusing on the most influential market factors is crucial, as it enables businesses to concentrate on what truly drives sales. This precision can make or break a business, and failure to understand sales trends can leave companies struggling to compete in a dynamic market. In this research market study, we used machine learning regression models of sales data from Corporación Favorita, the largest Ecuadorian grocery franchise, to analyze which factors influence sales in grocery retail. We hypothesized that macroeconomic factors have a larger impact on sales compared to geographic and seasonal features. In this project, we used the sales data for training lasso and ridge regression models, which were subsequently examined through Shapley analysis. We found that macroeconomic features, particularly the size of Ecuador’s labor force, exert the strongest effect on sales. However, we also found that other select features, such as city altitude and holiday proximity, also have a high impact on sales performance and should be incorporated along with macroeconomic conditions when developing business strategies. This research applies interpretable machine learning models for market analysis to improve profits and provide a competitive advantage to businesses in grocery retail.