Identify patterns for returned payments for a billing paying start up. This project will take two full years of payments with important meta data and create a dashboard related to when and why a payment is returned. Some reasons may include the time to process the payment, the history of the vendor, or user error in creating the payment (sending it to the wrong state, for example).
The project will require some data wrangling to prepare the data and a simple storyboard that informs an mid-level program manager as to when and why payments are getting returned.
Success would looks like finding at least two factors that a program manager could use to intervene on returned payments. Each returned payment costs an average ~$2.25 in processing (staff, postage, bank fees) with ~50,000 returned payments in a year.