All banks have pricing models, but few know what will actually happen if you change rates.
A leading US regional bank
The bank set rates using pricing models but felt there was opportunity to improve pricing for some customers. The bank had a history of changing rates and hoped to learn from customer response to those changes to inform future strategy. However, rate changes happened frequently and affected different sets of customers. Customers also responded differently based on their key criteria like tenure, income levels, or age. Additionally, the macroeconomic environment and competitive landscape were always changing and also affected performance. There was no way for the bank to come up with a clear pricing strategy in this chaos.
The bank used APT's Test & Learn™ solution to automatically find a set of customers whose rate had changed and a set of customers whose rate stayed the same. These groups of customers were different across key characteristics such as balance, tenure, and income, but the software was able to identify a subset of customers that were comparable across these dimensions.
Comparing these similar groups of customers clearly showed the impact of the rate change. The software then scanned thousands of factors to identify which types of customers responded best to the rate change. A predictive model incorporated these factors to identify the optimal rate strategy for each customer.
The bank determined the optimal rate for each customer and improved annual EBIT by $20M. More importantly, the bank was able to use APT’s Test & Learn™ Management System to constantly refine pricing, respond to changes in the economic environment, and continue to maximize profitability.