Carnegie Mellon University

Busy Beaver: Increasing Transaction Size Through Data Analytics

Capstone Team: Christian Meyer, Haider Aly Reza, Nico Roussel, Paul Lim

How can a traditional retailer motivate customers to increase their basket size? The answer to this question may be hiding in Busy Beaver’s database of historical customer transactions. Included in this vast dataset is information about how customers responded to prior promotional campaigns, which primarily take the form of biweekly advertisements that promote various discounted products. The key question is how effective these promotions are at driving incremental business. Moreover, are there opportunities for improvement by adjusting product selection and/or discount levels? Finally, how should seasonality influence the design of the promotions?

The approach chosen to address these questions was to apply modern machine learning techniques to understand the relationship between the various characteristics of a promotion and their corresponding impact on customer purchase behaviors. This approach enabled the identification of key insights that will likely improve management’s ability to design their biweekly advertisements. Ultimately, these insights formed the basis for a proof-of-concept application that illustrates how advanced data analytics can be used to help improve the effectiveness of key business decisions.d for future work to continue to understand the donor base behavior, especially as more data becomes available.

Capstone Presentation