I think both aspects are important because I've heard instances where BNPL companies try to grow so rapidly that they sometimes onboard fraudulent customers. These customers might engage in activities that essentially steal money. Quickly identifying and preventing their onboarding seems very beneficial for the business. So, I'm interested in both aspects.

Fraud is a significant challenge. It's not just about the percentage of people trying to commit fraud, but the potential financial impact, which can be substantial. When you're in credit, it's important to help the organization understand how many successful transactions are needed to cover one loss, whether it's a credit or fraud loss. Usually, the ratio is around 100 to 1 because margins in BNPL are quite low. For a fraud loss, the losses are typically 100%, whereas credit losses are a bit less.

To ensure I'm grasping this correctly, a potential fraudulent case for BNPL could involve a device farm attempting to pay in four installments for items like iPhones. They might pay the initial 25%, and after you facilitate the transaction, they receive the iPhones. Instead of making the remaining payments, they could send the iPhones abroad to be reprogrammed and sold at a profit. Were these some of the potential fraud cases you were encountering and trying to prevent?

At Klarna, we noticed mail-order mattress companies like Casper were emerging. Fraud is harder with mattresses; they're difficult to steal. We focused on these companies, directing our team towards them. GhostBed was another merchant that generated significant revenue for us. This approach allowed us to prove business viability without worrying about fraud. It also gave us time to develop more fraud prevention tools for the US market before onboarding more merchants.