Booking.com processes payments in more than 220 countries. Andreas Zodhiates, their Head of Payment Fraud, described the challenge their team faced: rule-based systems were good at catching known patterns, but they had a device-level blindspot. Fraudsters who cycled through cards and fresh accounts looked clean to every system that evaluated identity at the transaction level.

They integrated Fingerprint to close that gap. The device signal they added was persistent across browser sessions, card resets, and account creation events. For the first time, a fraud ring that created 40 accounts over six months was traceable to a single cluster of devices rather than 40 unrelated identities.

The second half of that quote matters as much as the first. When the device is known and trusted, Booking.com reduced friction at checkout. Conversion improved for legitimate customers because the system finally had a durable signal to distinguish them from bad actors.

The same dynamic applies at Stripe's scale. Device identity lets you concentrate friction where fraud risk is real, not distributed across all unfamiliar cards and IPs. Stripe already handles the transaction layer. Fingerprint adds the device layer that completes the picture.

The full case study covers how the integration was architected and what they measured. Worth a read if you are thinking about the same problem.

Fingerprint, Inc.  ·  Unsubscribe