Identity Verification Confidence Score Calculator
Calculate a composite identity verification confidence score (0–100) based on document quality, biometric match strength, data consistency, and risk signal penalties. Used in KYC, onboarding, and fraud prevention workflows.
Formula
Step 1 — Component Scores:
Doc Score = Document Quality (0–100) × Document Type Weight
Bio Score = Biometric Match (0–100) × Liveness Weight
Data Score = Data Consistency (0–100)
DB Score = Database Check Result × 100
Step 2 — Weighted Base Score:
Base Score = (0.25 × Doc Score) + (0.35 × Bio Score) + (0.25 × Data Score) + (0.15 × DB Score)
Step 3 — Risk Penalty:
Total Penalty = Number of Risk Signals × Penalty per Signal
Step 4 — Final Confidence Score:
Score = clamp(Base Score − Total Penalty, 0, 100)
Thresholds:
≥ 80 → High Confidence (Approve) | 60–79 → Medium (Review) | 40–59 → Low (EDD) | < 40 → Reject
Assumptions & References
- Component weights (Doc 25%, Bio 35%, Data 25%, DB 15%) reflect industry-standard KYC prioritisation where biometric liveness is the strongest signal.
- Document type weights are based on FATF and eIDAS trust-level classifications for identity documents.
- Liveness check multipliers follow NIST SP 800-63B guidance on presentation attack detection (PAD) levels.
- Risk signal penalties are configurable; typical signals include VPN/proxy use, device fingerprint anomalies, unusual geolocation, velocity checks, and behavioural anomalies.
- Watchlist/database check values align with OFAC, EU Sanctions, and PEP screening outcomes.
- Thresholds (80/60/40) are commonly used in financial services onboarding; adjust per your organisation's risk appetite and regulatory requirements.
- References: FATF Guidance on Digital Identity (2020), NIST SP 800-63-3, eIDAS Regulation (EU) 910/2014, ISO/IEC 30107-3 (Biometric PAD).