Root-Causing and Responding to a Fraud Rise at a Single Merchant

Fraud at one of our merchants has been rising. How would you work out what is driving it, and what would you do about it? ```hint Scope before you hunt Before listing causes, pin down how the fraud is being measured and whether the rise is confined to this merchant or is also showing up elsewhere. ``` ### Clarifying Questions - How is the increase measured: fraud losses, chargeback rate, confirmed fraud cases or the merchant's own complaints, and is the fraud confirmed or only suspected? - What kind of fraud is it: stolen payment details, refund or claim abuse, fake accounts, or a takeover of the merchant's own account? - Was the rise sudden or gradual, and over what period? - Who bears the losses today, the merchant or us, and how important is this merchant to the business? ### What a Strong Answer Covers - Clarifying the metric and the type of fraud before forming hypotheses - A structured set of possible causes, spanning the merchant's own changes, changes on our side and outside forces, with the data that would confirm or rule out each - Immediate containment for the merchant, weighed against the cost of blocking good customers - A lasting fix and how its success is measured - Communication with the merchant, and whether other merchants are exposed ### Follow-up Questions - Suppose the rise began right after we updated our fraud-scoring model. What do you do next? - How would you decide how strict to make the fraud rules for this merchant, given that every blocked order might come from a real customer? - How would you catch a problem like this before the merchant notices it?