Improve your Fraud Detection integration

Last updated: April 30, 2025

For optimal performance, the Fraud Detection solution relies on quality data input for the rules and machine-learning (ML) models. The more data your integration provides, the greater the confidence in the data quality and the effectiveness at preventing fraud.

To maximize the solution's effectiveness and efficiency at preventing fraud, we recommend the following:

Always provide as much data for your payment requests as possible.
Integrate the Risk.js package, and the Risk Android SDK or Risk iOS SDK for every customer-initiated payment flow.

The data confidence for each type of integration is as follows:

Integration	Data confidence
Flow – browser and app	High
Hosted Payment Pages	High
Payment Links	High
All card details via the API, with customer data and the Risk.js package	High
All card details via the API, with customer data and the Risk Android or iOS SDK	High
All card details via the API, with customer data	Medium
All card details via the API, without customer data	Low

We recommend providing the following data points for optimal performance:

Data point	Contribution to performance	How to provide data
Billing address country	Required	Gateway
Customer email	Required	Gateway
Customer IP address	Required	Gateway and Risk.js package
Device and browser data	Required	Risk.js package
Billing address ZIP code or post code	Optional but highly recommended	Gateway
Customer name	Optional but highly recommended	Gateway
Customer phone	Optional but highly recommended	Gateway
Shipping address country, ZIP code or post code	Optional but highly recommended	Gateway
Billing address street	Optional	Gateway
Shipping address street	Optional	Gateway

If you collect device data through a third-party provider, you can share this data with Checkout.com using the API. Device data strengthens the risk assessment and enriches your analytics in the Dashboard.

When you call the Request a payment endpoint, provide the device data in the risk.device object.

You can use the following device-signal properties in your custom rules. Provide the required data using the API as follows:

Property	Description	Required data
`:device_fingerprint:`	The unique device fingerprint. For web flows, this value is unique to each browser. Example – `rekyngyfnrte3ckzvgjlhw4jfm`	Provide the device fingerprint in the `risk.device.network.id` field.
`:device_ip:`	The device's IP address. Example – `10.256.35.48` If `:payment_ip:` is omitted in the payment request, the rule property falls back to `:device_ip:`.	Provide the device IP address in one of the following fields: `risk.device.network.ipv4` `risk.device.network.ipv6`
`:device_ip_country:`	The ISO-2 country code associated with the device's IP address. Example – `GB`
`:device_ip_city:`	The city associated with the geolocation IP address. Example – `London`
`:device_model:`	The device model. Examples: `Generic Smartphone` `iPad` `iPhone` `Laptop` `Other` `Pixel 5` `Samsung`	Provide the device user agent in the `risk.device.user_agent` field.
`:device_os:`	The device’s operating system (OS). Examples: `Android` `Chrome OS` `iOS` `Linux` `Mac OS X` `Windows`
`:device_os_version:`	The version of the device’s operating system. Example – `13.3`
`:device_browser:`	For web flows, the browser used on the device. Examples: `Chrome` `Chrome Mobile iOS` `Firefox` `Opera` `Safari`
`:device_timezone:`	The timezone of device's IP address. Examples: `America/Chicago` `Asia/Singapore` `Europe/London`
`:device_browser_is_incognito:`	For web flows, indicates whether the device browser is in incognito mode. Example – `true` or `false`

To gain insight into your integration's performance:

Sign in to the Dashboard.
Go to Payments > Fraud > Performance.

The Risk analytics screen provides an overview of metrics, transaction journeys, and rule performance. You can also generate a report for a specific time frame.

Improve your Fraud Detection integration

Data confidence by integration

Data importance

Share device data

Review integration performance