We analyzed the sales situation at a farmer's shop using POS data for a certain month provided by a JA in Hyogo Prefecture. The provided data included information such as the date of the sale, the customer's age, gender, the purchase amount and quantity of each item, and the product category. Using these data, we performed time series analysis, attribute analysis, parallel sales analysis, and the measurement of the effectiveness of promotional events. We will introduce the results. Prior to these analyses, we used the table function of JMP to aggregate data by purchaser attributes. We will also introduce how the excellent GUI function, which is not available in commonly used tools such as Python, made the preprocessing of the analysis more efficient. We also performed univariate analysis, etc., and were able to visually grasp the characteristics of the data. Then, we used Python and other tools to perform time series analysis to grasp the trends in the number of visitors and purchase amounts by day of the week, and analyzed the changes in sales by product category. We also linked the results to external data to explore the factors behind the fluctuations in sales. Next, in the attribute analysis, we clarified the relationship between the customer's age and gender and the products purchased. In the parallel sales analysis, we used graphs to visualize the associations between products that are often purchased together. Finally, we measured the effectiveness of promotional events and evaluated the impact of specific promotional events on sales. Through these analyses, we plan to understand the characteristics of each store and provide information that will be useful for future store operations.
