Skip to main content
Making decisions based on current market behavior

Test-and-learn cycles have taken on new importance since the financial crisis. In many markets today, there's been so much change that historical data is no longer sufficient for predicting customer behavior. The speed with which creditors can understand and adapt to how consumers are currently behaving in the market is becoming critical to success.

A FICO banking client is one of the first to raise its learning capacity to a higher level by implementing an analytic learning loop. This is a way of accelerating and providing shared access to feedback about the market performance of models and decision strategies. The initial project aims at driving growth by targeting the right card products to consumers most likely to respond and use the card in a manner that generates profits.

The analytic learning loop provides the means to drive rapid test-and-learn cycles that compare what was expected with what actually happened. Because the loop can be "threaded" through existing systems, this quick market feedback becomes a shared resource across customer lifecycle decision areas. In this case, marketing is gaining early access to originations decisions and outcomes. It's soon clear whether campaigns are attracting the customers that the bank needs to be profitable, and if these customers are showing early signs of using the products in the way the bank intended. 

New-Normal-Analytic-Loop

Variances between forecasts (based on simulations during development) and actual results are opportunities for learning. Marketing can laser in on underperforming population segments and change its strategy for that segment. While a campaign is still underway, the company can adjust the offer, the language and packaging of mailed and online promotions, and even call center and branch staff training to improve results.

related posts