Increasingly banks are looking to leverage the full breadth of data related to customers to make better decisions, and fraud is no exception. There is a strong desire to break down silos, so fraud/risk behaviors in one area of the customer’s interaction with the bank can be utilized in other decisions impacting the customer experience.
Traditional fraud detection systems are highly specialized to compute risk for a particular access method, making an enterprise view extremely challenging. What's needed is a new approach.
Connecting behaviors holistically requires a data model that links the accounts and channels utilized by the customer. Once this data model is created, it's necessary to link various channel and account profiles to specific customers. By enabling these linkages, one can then expand on behaviors and scores occurring within channels/accounts by rolling it up at the customer level. In addition, having the customer profile allows for better utilization and personalization of decisions based on the net present value of the customer and/or regular behaviors that can help reduce false positives.
Next month, I'll be discussing the challenges of enterprise fraud management at FICO World, where I'll present the session: Analytic Innovations in Enterprise Fraud. In preparation, I welcome your thoughts and comments on this topic here on the blog. Or better yet, join us at FICO World, and share your experiences and questions with me in person.