Capital One Virtual Card Numbers: North Star, Funnel and Fraud Metrics

You are a product manager at Capital One. Cardholders can generate virtual card numbers: stand-in numbers that bill their real credit card account while keeping the actual card number away from online merchants. How would you measure whether this feature is a success? ```hint Value and timing Separate what the cardholder gains from what the bank gains. The outcome the bank cares about most, fraud avoided, is rare and shows up weeks after a purchase, so think about what can move sooner. ``` ### Clarifying Questions - Is the feature newly launched or already in wide use? That decides whether adoption or protected usage should be the headline metric. - Is the bank's main aim lower fraud losses, more online spending on the card, or customer trust and retention? - Where are virtual numbers created: in the mobile app, on the website, or automatically at checkout through a browser extension? - Can a cardholder tie a number to one merchant, pause it or delete it? ### What a Strong Answer Covers - The feature's goals for cardholders and for the bank, and a north-star metric chosen with the feature's maturity in mind - An adoption-to-habit funnel with a metric at each step - Security outcome metrics (fraud, disputes, reissued cards) that account for how rare and delayed fraud is - Guardrails for declines, checkout friction, refunds and support contacts - A way to separate the feature's effect from the self-selection of the cardholders who adopt it ### Follow-up Questions - Adoption is high, but most virtual numbers are used once and never again. Is that a problem, and how would you find out? - How would you estimate the fraud losses the feature prevents in dollars? - Declined transactions on virtual numbers rise after a release. How would you diagnose it?