In my last post, I talked about technology and, in particular, components in a fraud-detection architecture. But when it comes to building a fraud strategy, it’s not just about technology. Perhaps equally, if not more, important are the people who analyze and review cases to catch fraudsters. This post is part one of a two part series discussing hiring fraud analysts and building a fraud / risk team. First, let’s take on hiring, key traits to look for and how to assess them.
Hiring managers often think about the following questions: Is it better to hire from inside or outside? What qualities should great fraud analysts have? What skill sets and backgrounds are needed and most crucial?
Not unlike the build it or buy question, hiring internally or externally requires a solid evaluation of your current situation. I’ve hired both internally and externally with similar success rates. While an internal hire may bring greater product and business understanding, an external hire may bring prior fraud detection knowledge and a different perspective on a situation. There is also no straight path and to hiring fraud analysts and building a team can be done with people from diverse educational and technical backgrounds. Individuals with training in engineering or statistics may take on more technical work such as building features in machine learning models, but some of the analysts with the best investigative capabilities with whom I’ve worked have come from non-technical backgrounds such as art, music, and writing.
So what should you look for and how should you evaluate? Having worked with hundreds of fraud analysts, I’ve found that evaluating on the following three dimensions produces the strongest candidates during the hiring phase: 1) Analytical Ability, 2) Curiosity, and 3) Enthusiasm. Your best analysts should display strong abilities and traits in all three areas.
1) Analytical Ability
Your fraud analyst must be able to visualize different scenarios from multiple view points and extrapolate different “stories” from the data. This is critical because they make decisions on whether to approve or reject transactions and customer accounts. Having the ability to recognize a valid purchase when it looks fraudulent on the surface, or vice versa, makes the difference between losing a good customer or incurring a loss for the company. In addition, they should have the capacity to extract relevant info and spot patterns from a large data set while also be able to clearly communicate the rationale for their decision.
The fraud analyst is a problem solver, breaking things down into their component parts. He or she uses deductive reasoning and draws conclusions from givens, applying judgment to reach decisions from a combination of evidence and assumptions.
How Do You Assess Analytical Ability
One of the ways that I evaluate an analyst’s analytical ability is by having them walk through sample fraud cases. For example, I’ll provide a list of user accounts, both good and bad, with various types of background info on each user (e.g. social profiles, IP addresses they’ve used, purchase history, site pages visited, connections to other users, login activities) as well as a list of valid and fraudulent transactions and event logs. With this information, I’ll ask the analyst to correlate and match the user accounts to a corresponding transaction or event most likely conducted by that user, and also map out what led them to that decision. In some instances, I may exclude certain pieces of information (e.g. provide incomplete data) in order to gauge how the analyst arrives at a verdict through both proof and inference.
Through this exercise, I’m ultimately looking to see that the analyst can not only digest a lot of information, but also quickly and efficiently both organize and prioritize the data that’s important to tell a story of fraud or no fraud. More importantly, what I can find out from this activity is whether the analyst can identify repetition and patterns within the data that he or she can translate into a review process across other accounts and transactions.
One of the most important traits of a good fraud analyst is a sense of curiosity. This trait is an indicator of a range of other qualities. Creativity, innovation, and the ability to learn quickly all spring from curiosity. Fraud analysts often investigate and do research on transactions and activity on user accounts. Whether it’s to gain a deeper understanding of the situation or to understand the fraudster’s behavior, the drive to know more pushes the fraud analyst to be better at reaching the right conclusion. For example, the analyst who is driven to dig a bit further into the data might uncover the mismatching information that finds the fraudulent transaction initially deemed to be good, or the fake sign-up originally approved as legitimate. One of the best analysts that I’ve known was constantly researching and learning about trends in another industry. Through his findings, he would always alert us to potential fraud attacks on our site based on what he saw in this other industry, which helped us see these attacks sooner and mitigate any damages.
How Do You Assess Curiosity
One way to determine whether the analyst is a curious person is to ask him or her the following questions during the interview:
- Describe something interesting that you taught yourself recently.
- How did you teach yourself this new skill/subject and what was the outcome?
- What do you spend time reading, researching, and becoming an expert on?
The responses to these questions offer insight into whether or not they are innately curious and dedicated to continual learning. If the analyst exhibits a positive energy level when talking to any of these questions, it’s usually a good sign of curiosity. If they actively seek out new information weekly or daily, that’s a good indicator of their desire to learn. Getting to know the steps they took to satisfy their interest helps you evaluate their ability to perform research.
Another way to determine whether the candidate is curious is to give him or her an assignment that requires research and note how in-depth and detailed his or her findings are. One assignment I like is to provide candidates a list of IP addresses and have them find as much information as they can on each IP, including the type, and explain how they confirmed that the data is accurate. I’ve been impressed and amazed with the level of creativity candidates have achieved to get results.
Last but not least, keep an eye out for people who ask follow-up questions throughout the interview. Almost every candidate will ask questions at the end of the interview, but an intellectually curious individual will come up with new questions on the spot based on what you’re discussing.
A strong sense of positivity and composure under pressure are two great signs for a fraud analyst. They also take what they’ve learned and look to actively improve the operational processes for detecting fraud. They enjoy collaborating with other teams to reduce risk and safeguard new product features. Fighting fraudsters every day can be taxing. Analysts are exposed daily to the darker sides of the business and even when they do a good job of keeping fraud down, there may not be a celebration to recognize the achievement. Fraud is not something that’s often talked about externally and as soon you celebrate, something will go wrong (as has been my experience).
Sometimes fraud analysts need to be the bearer of bad news to customers when they’ve found an occurrence of fraud. It’s not easy conveying negative details to customers. Mistakes will be made and fraud will be missed, or good transactions will be incorrectly marked as bad. Both of these scenarios lead to negative outcomes and losses for the company. Being able to stay focused and move on to the next case is a very crucial quality for the analyst to possess.
How Do You Assess Enthusiasm
Assessing for enthusiasm is partly like evaluating for organizational fit. Ask situational or behavioral questions. I’ve found that getting people to describe a challenging situation they’ve faced recently at work and how they tackled it works well to clarify the candidate’s working style under tension. Pay attention to whether or not they show great interest, emotion, and enthusiasm when describing situations and people they work with in the workplace. Ask them to describe an organizational challenge and see if they have a constructive recommendation and steps to solve it.
What’s important here is listening to how the candidate talks about the challenging situation. Are they negative and complaining or are they upbeat and making the most of the setting despite not being able to influence a change? Do they share positive examples of their experiences that show their willingness to go beyond the call of duty? Above all else, you want to assess that the person is motivated to keep on going in the face of hurdles because there are plenty when it comes to battling fraud.
Although there are numerous other areas that can be assessed, by focusing on analytical ability, curiosity, and enthusiasm as the dimensions for hiring, you’re creating a solid foundation for a successful fraud analyst team. Hiring is a continuous process and it’s only one part of building your team. To complement hiring, you’ll want to build upon your team and provide them the best environment to succeed, which I’ll tackle in my next blog on optimizing your team’s value.
If you’re interested in diving further into anyone of the areas discussed in this blog, please email me at firstname.lastname@example.org. I’m always up to learn more about your challenges, your company, and how I and/or DataVisor can help you. In the meantime, look out for part two of the hiring and building a fraud analyst team series coming soon.