All use cases

Transaction risk

Score a payment or checkout in real time using device and session signals, so you can approve trusted transactions, step up the uncertain ones, and challenge the risky ones.

Fingerprint fit Supporting signal

Overview

Transaction risk is the practice of evaluating an individual payment or checkout as it happens and deciding how to handle it: approve low-risk activity with no friction, step up verification when something looks off, or challenge and hold the clearly risky. It exists because transaction fraud, using stolen or fake payment details to make unauthorized purchases, follows a familiar arc from data theft through monetization, and the same payment can look clean on the surface while carrying signals that suggest otherwise. The aim is a graduated response rather than a blunt approve-or-decline, which catches more fraud while sparing genuine customers unnecessary friction and reducing chargebacks. This matters most to e-commerce and to fintech and other financial services, where each transaction carries direct financial exposure.

Real-time transaction risk scoring draws on payment and network data, behavioral and velocity patterns, identity and authentication context like 3DS and step-up, and device intelligence, combining them into a score that drives a decision in milliseconds. Device and behavioral signals are valuable inputs that add context a transaction record alone does not carry, such as whether the request is automated, comes from an environment built to evade detection, or shows a device changing key attributes at an unusual rate.

Fingerprint feeds the device and session layer into that scoring. Bot Detection flags automated checkout attempts, including the scripted activity behind card testing and bot-driven payment fraud. Velocity Signals surface a device cycling through data points like IP, country, or payment attempts at an unusually high rate, a pattern consistent with stolen-card abuse. Proxy Detection adds context when a payer routes their connection to obscure origin, and Suspect Score combines many device attributes into a single risk indicator a business can threshold on.

How to implement it

  • Call Fingerprint at checkout or payment authorization before the transaction completes. Check Suspect Score and Bot Detection, and store the Visitor ID with the transaction in your database for downstream review.

  • Use Suspect Score as a real-time risk indicator on each transaction, thresholding it to approve low-risk payments, step up the uncertain ones, and hold the clearly risky.

  • Use Bot Detection to flag automated checkout attempts, including the scripted card testing behind bot-driven payment fraud.

  • Use Velocity Signals to surface a device cycling through IPs, countries, or payment attempts at an unusual rate, a pattern consistent with stolen-card abuse.

  • Use Proxy Detection to add context when a payer routes their connection to obscure their origin during a transaction.

  • On mobile, feed Active Call Detection into the score as session context: a call in progress while a payment is authorized is a common sign the payer is being coached, and it is most telling on high-value transactions or a first payment to a new recipient.

Also known as

transaction risk · transaction risk scoring · transaction scoring · payment risk evaluation · transaction anomaly detection · payment fraud scoring · real-time transaction monitoring

Ready to identify?

Start building with a free trial or speak to an expert to explore our Enterprise plan.