Making meanings out of the huge amounts of data that are generated almost every second across the globe is becoming an important data science concern. Discovering a unique feature(s) that can aid in the classification, prediction, and general analysis of a particular system under consideration could be considered a major task of data science. Data science tools are desperately needed to draw insights from the vast amounts of data that are generated for critical decision-making and planning for the government, businesses, military, politics, and academia, amongst several other critical organizations. In this paper, our motivation is to investigate whether Benford’s law can serve as a data science tool. For this, experiments were performed on Point of Sale (POS) datasets. POS key features such as TranTime (Transaction time in seconds), BreakTime (Break time including idle time in seconds), ArtNum (Number of items, i.e., basket size), and Amount (Transaction value) served as inputs into Benford’s law. Results obtained showed that the Amount feature of the POS system perfectly conforms to Benford’s law based on its plots and chi-square divergence. The results showed that normal Amount transactions on POS systems followed Benford’s law, whereas fraudulent/tampered POS Amount transactions deviated from this law. We found that Benford’s law can actually serve as a data science tool by giving us insights into POS operations.