Introduction:

This was a project for my college course: BNAD 277--Analytical Methods For Business. My team and I were tasked with choosing a company and then conducting complex regression analysis to predict profit per order based on a number of independent variables. There were many key components to this project, as outlined in the table of contents of the report. The main point of this project was to test multiple regression models to determine not only the highest adjusted-R^2, but also the best combination of independent variables to accurately predict the profit per order for Amazon.

​

Description:

The report on Amazon's order data focused on predicting profit per order to inform strategic decisions. It involved analyzing 9,900 rows of customer data using Excel's data analysis toolpack. The study identified key variables affecting profit, such as product categories, order destinations, and seasonal effects. The team tested over 15 regression models to find the best fit, ultimately creating a season-based dummy variable to enhance prediction accuracy. Findings indicated higher profits were associated with technology products and home office segments, while categories like furniture and orders during winter were less profitable. Recommendations included focusing on high-revenue segments with fewer discounts to increase profitability.

​

Results:

The end result of this project was extremely rewarding. Not only did I get an A on the project and was recognized as one of the best projects in the class, but I learned a lot of new things past the material taught in class. Being able to do a hands-on statistical regression project for a real company helped me solidify these rather complex topics. Furthermore, working with a team of 5 including myself was important, as we all had different perspectives and opinions which ultimately lead to the high grade and recognition on this project. After completing this project, I feel much more confident working with Excel, regression, and statistical models in a work environment.