A Look at Fraud Prevention and the ROI Curve
0 comment(s)
What's the of point fraud prevention? Perhaps the simplest answer is that it pays by recovering losses (costs) to the organization. Think about employee fraud for a second. The news reminds us of how prevalent employee fraud is, and when it strikes it may feel like the blade of the guillotine because the losses are usually substantial and it's committed by a trusted employee. So proactive employee fraud detection mitigates a very real risk, and as many of you know this can be difficult to quantify. But not all fraud is like the guillotine blade waiting to fall. An awful lot of fraud is more like death by papers cuts. Check fraud is a good example, where a steady flow of small cuts can lead to a rather damaged organism and costly losses. For fraud of this flavor, the payoff can be understood well enough to determine the solution.
Okay, so what is this payoff? Well, there are a few different ways to describe it, and I'm going to take the tack that the mathematician in me likes best. With that said, I suppose I better fess up now that I am a mathematician. My job is to build statistical learning models that detect fraud. So, with this disclaimer in mind, let's examine the curve in this graph:
Perhaps math isn't your favorite subject or maybe it's been a while – so let's go over the graph: This is a return on investment curve or ROI curve. This is created using real deposit fraud data and it is for one of our deposit fraud solution queues. This shows a view of the payoff that I like best, as it allows me quantify the utilization of fraud management resources with respect to optimizing fraud loss savings.
This ROI curve is a graph of the annual return on investment in millions of dollars as a function of the average number of alerts worked per day. Said another way this is the loss avoided (or recovered) minus the cost the organization incurs to find that fraud. By understanding the optimal point of resource utilization as it relates to loss mitigation it empowers you to make better decisions about how you allocate your limited fraud prevention resources. An example to think about is the grocery store. There are a few items you need, but you know they cost more at the grocery store and you could easily save $5.00 by getting these items at Walmart. So the question becomes is the incremental savings worth the cost (time and energy) of going someplace else?
But I digress, let's get back to the ROI curve. One of the reasons it has this steeper shape is because at Memento we ranks alerts more accurately based on both the risk profile and the potential fraud loss amount. Look at the ROI curve again, notice how it rapidly peaks then slopes downward as more alerts are worked. As the ROI curve peaks it shows the power of Memento's advanced analytics and accurate risk ranking of alerts. It means your fraud prevention team reviews and investigates alerts with the highest probability and greatest potential for loss first. As the curve slopes downward – it shows diminishing marginal returns because the incremental cost of preventing fraud by working more alerts may not be worth the savings on the losses avoided. Only you can decide what is best for your company based on your data.
So this is what the ROI curve looks like, but the more important question is: Why is the ROI curve the right curve to look at?
I'll discuss that tomorrow so stay tuned!