Fraud is a Human Behavior
 

Paul McGinley, LPC
Regional Loss Prevention Manager
Dollar Financial Group

In the D&D Daily dated June 24th, Stuart Levine (CEO of the Zellman Group), stated that “Fraud is not a person – it is a dynamic grouping of statistics that deviate from the norm.” While the lesson illustrated by this statement is a good one, I must respectfully disagree.

In our ever-continuing goal to be data-driven, we must not lose sight of where that data comes from. Data is nothing but numerically coded results of human interactions. These interactions might be human to human, human to system, or initiated by a human to be system to system. In any of these circumstances, data is not real. Think of data like a photograph: it might be a perfect representation of a moment, but the photo isn’t the moment. Data is the same. It is a wonderful momentary representation, but that is all it is.

Dynamic statistics are more like a movie than a photograph. But just as with the photo, the movie is not the event. The movie tells a story, sometimes the story is so full and robust that you truly have a comprehensive vision of events, not just in the story, but in the universe around it. However, again, that movie is not the event. The movie is simply evidence of the event. No matter how robust or dynamic they are, statistics are the same. They simply tell stories. The stories that they tell are stories of human interactions. Take a P&L for example. EVERY line on your P&L is simply coalesced data on the human interactions between your customers/employees and other customers/employees/systems. Just as your P&L is not your business, but rather a representation of it, your fraud statistics are not your fraud, they are a representation of it.

Attempting to combat fraud statistics will result in two very unintended consequences: 1) though you will very likely see excellent results in the metrics you are tracking, fraud and loss will migrate to other areas 2) you will create an unwelcoming environment for both your customers and your employees. By addressing the statistic you are simultaneously creating a whack-a-mole and fishing trawler approach; you will only knock down an issue when it pops-up, and you will catch lots of innocent, yet uniquely transacting, customers/employees in your net. This is to not say you should not address your fraud statistics, as stated previously, they are telling you a very important story. Just as in an illness where the symptoms tell the story, the symptoms are only to be managed while the illness itself is treated.

None of this should diminish the importance of dynamic fraud statistics, rather it should emphasize how important they are to help you to identify the root of the issue. What it should also illustrate is the root will always be people related. There is no such thing as a fraud scenario that occurred because of statistics, each fraud event occurs due to an action initiated by a person takes. When approached it from a people perspective you will be able to craft tailored fraud solutions that do not have the two unintended consequences above. It has been our organization’s experience, time and time again, that appropriate fraud solutions based on a people perspective result in three very intended consequences: 1) the fraud statistic that you are addressing will show positive results 2) “adjacent” statistics will also improve (other types of fraud, top line revenue, positive middle of the P&L impacts) and 3) the customer and employee experience is greatly improved. The people perspective provides an additional benefit in that it creates an alliance between customers, frontline employees, and investigative personnel.

Ultimately, fraud is not a person, but it is also not a group of statistics. I would offer this alternate thesis statement: Fraud is a human behavior that is typically identifiable by utilizing dynamic groupings of statistics that deviate from the norm.