From Sales to Strategy: A Lightweight Analytical Framework for E-Commerce Performance Evaluation

Abstract

Modern e-commerce generates large amounts of transactional data, although, surprisingly, often fails to leverage its strategic potential. Typically, companies are focused only on sales quantity metrics but do not ask themselves what specific products or product groups bring their company any money. In this regard, the present paper attempts to fill the existing gap by providing an analytical approach based on easily replicable methodology to convert sales data into strategic information. Using publicly available Amazon Sales Report data from Kaggle, the research applies a set of algorithms in Python via Google Colab platform to convert raw sales numbers into meaningful information that goes far beyond traditional sales reports. Since the data does not provide any information regarding costs incurred in order to make sales, the profit is estimated using a common proxy margin, making it possible to compare SKUs and categories in relative terms, without measuring absolute profits. Visual analytics in the form of monthly charts, category-level analysis, SKU ranking and contribution treemap are employed to provide insights that cannot be gained with conventional dashboarding techniques. According to the results, some of the high-sales SKUs and product categories account for surprisingly small shares of profits. With the emphasis of this revenue-profits divide, the model offers managers and analysts an effective perspective through which to focus product decisions, adjust their marketing spending, and negotiate costs based on margins. The research shows that strategy does not require sophisticated machine learning or costly business intelligence tools but rather intelligent and intelligible analytics on available data.