Feedback API Overview

Feedback is critical to closing the loop between Signifyd’s real-time decisions and the outcomes you observe in production. By sharing whether a user ultimately completed verification, was denied, or proceeded through your flow, you help refine Signifyd’s models, improve rules, and tailor risk assessments to your unique customer base.

Fraud tactics evolve quickly, and legitimate user behavior changes over time. Feedback provides essential ground truth, allowing Signifyd to:

• Improve model precision – Reduce false positives and false negatives through better training data.
• Adapt to your business – Tailor risk scoring to your specific flows and user behaviors.
• Support reporting and insights – Ensure dashboards reflect true business outcomes.
• Tune rules and thresholds – Identify areas where policies may need adjustment.

Feedback should reflect the final action taken by your system and the observed user result—not speculative fraud risk or downstream outcomes. The call can be made asynchronously but should be tied to the original decision as closely as possible.

Typical feedback scenarios include:

• The user successfully completed a STEP_UP challenge.
• A DENY recommendation was overridden and the user was allowed to proceed.
• An ALERT was shown to the user and no further action was needed.

Each Account Integrity flow (Account Opening, Login, and Modification) has its own Feedback endpoint to capture these results. Specific details for calling each Feedback endpoint appear in the corresponding API reference sections.