Continuing with the house analogy, we will now create the blueprint for your house based on the building code from Chapter 3 and the fraud scenarios identified from Chapter 2. In this chapter, we will design each room of the house based on the fraud scenarios identified in your fraud risk assessment. Hopefully, we will avoid change orders, although I believe the nature of fraud data analytics is an evolving process.
In this chapter, we will discuss the methodology for building a fraud data analytics plan. There are eight stages to building the plan. Initially, you may feel the process is bureaucratic or redundant. In some ways the reader is right. However, it is critical to ask as many questions as possible before creating the data interrogation routine. Otherwise, the plan may result in either excessive false positives or, worse yet, false negatives (a missed fraudulent transaction). In time, the process of developing a fraud data analytics plan will become intuitive.
Asking the right questions, in the right order, is critical from a logic perspective. However, for my free thinkers, I offer the list not to constrain you but to give you a checklist of questions you need to consider. Within each step there will be many considerations. The decisions should be understood and documented as part of the workpaper process. So, what are the steps or questions to building a fraud data analytic plan?