Q&A with Marianne Bunn, Product Owner and Business Analyst at Clutch
Loyalty fraud is the act of cheating in a loyalty program to unfairly earn points or benefits. We asked resident Machine Learning expert at Clutch, Marianne Bunn to explain how ML technology can be used to monitor and ultimately stop fraudsters.
As a thought leader in the customer loyalty space, what type of customer loyalty program fraud and gaming are you seeing today?
We are seeing loyalty program fraud from both customers and employees. When brands don’t require in-person loyalty scans they generally experience more customers gaming their loyalty program. On the employee side we are seeing employees adding their loyalty accounts to transactions when a customer isn’t a part of the program. On the customer side we are seeing customers add fake family members to their account or refer fake friends when referral bonuses are on the line. Our ML-driven loyalty program fraud solution is even identifying possible customer gaming that loss prevention teams haven’t seen before such as social media influencers sharing their loyalty account with their followers.
Has the prevalence of fraud/program gaming changed during COVID and how have your clients responded?
Throughout the COVID-19 pandemic we have all seen a shift towards engaging customers through digital channels. In the loyalty program world we have seen clients build more apps or add more features to their apps. With a lot of brands hosting digital accounts we are seeing brands allowing loyalty scans and behavior to live online. This makes it more difficult to catch fraudsters because they can live anonymously online and are more willing to try and game the system when no one is watching. This has made it easier to identify fraud because we have more data points, but more difficult to shut down these accounts. It is dangerous to shut down an account when the data doesn’t paint the entire picture as there is always the chance there’s a legitimate customer story behind what they’re seeing. Since we have less context on what a customer’s intentions are, bad actors use that to their advantage and really test the limits on how much they can game the system. Even when clients shut down a bad actors account, they can create a new persona and do the same things over again.
We continue to hear from brands that they realize and are ok with some forms of gaming and fraud (ex: family members sharing accounts), but they are trying to limit the most egregious forms (ones that impact the brand the most). What are your thoughts regarding how brands are approaching fraud? How do they limit fraud’s impact on the brand?
- Brands need to ensure that their focus is first on the customer’s bottom line – the customer experience. Focusing on the customer experience will naturally help their own bottom line. Allowing family members to share an account supports one aspect of the bottom line but the brand should also ensure that fraudsters can’t work to compromise a customers account – which we should remember has personal information and leads to a not ideal clean-up situation. We all remember how stressful that is with other compromised accounts.
- Brands can limit the impact of fraud by implementing features such as unique randomized single use coupon codes, login security features, ML-driven segments identifying possibly fraudulent accounts or possibly compromised accounts. One of the most impactful things they can do is think like a fraudster.
There is a lack of fraud reporting/assistance we hear from brands regarding their current customer loyalty platforms. Is this something you see? How do you help your clients with identifying fraud within your reporting processes?
All of our brands, from small business to enterprise level, can benefit from the power of a machine learning (ML) segment that identifies possibly fraudulent accounts. This segment is more than business rules and gets more intelligent with more data over time. Traditional top earner reports have a high ratio of the most loyal customers so brands end up wasting a lot of time sifting through those reports to fraudulent customers. As our segment learns the patterns of fraudulent customers over time, our brands are able to focus their time and efforts on the right customers.
There is a growing concern for employee fraud from our member community. What are you seeing regarding employee fraud, and how should brands address it?
When there is intention behind employee fraud we usually see employees adding their loyalty accounts to transactions when a customer isn’t a part of the program. This is even more rampant for brands that allow digital loyalty scans since there isn’t someone watching over their shoulder. We are even seeing employees doing the same thing but with a known charity’s card or someone in need. Employees may believe their Robin Hood act is harmless, but brands usually disagree. Brands concerned with employee fraud should first ensure that there is a clear policy communicated to employees because employees gaming a loyalty program may not know it is against their employee agreement (and sometimes it may not be). After ensuring a clear policy is in place, brands need intelligent reporting (such as our Possible Fraudulent Account segment) to identify possible fraud as it may fit within their business model. After identification, we’ve seen brands stutter when it comes to actually investigating the data and employees. It is important for the brand to develop internal processes and next steps to confirm and shut down their employee fraud.
How should brands be training their employees as to the impact of fraud?
This is another opportunity to outline the impact of fraud as it pertains to the customer’s bottom line and experience because not every employee is going to care about the impact to a company’s bottom line. Brands can describe the negative impact fraud can have on a customer’s current loyalty program experience as well as their future loyalty experience if they’re always busy chasing down fraud instead of enhancing the loyalty program. I think brands should also bring in their front line employees to help them identify suspicious loyalty program behaviors and patterns that the brand may not be aware of. When everyone feels like they have an important role to play for a common cause, they’re more likely to care.