Session management and configuration
Updated on 20.02.26
10 minutes to read
Copy link
Overview
Proper session management is key to a smooth user experience and reducing abandonment, while thoughtful configuration helps you fine-tune your risk strategy.
Understanding these components is key to fine tuning your verification workflow, reducing abandonment rates and making data-driven decisions to fine-tune your risk strategy.
Key benefits
| Benefit | Description |
| Flexible user journeys | Pre-initialized sessions and automatic session transfer allow you to design flexible, non-linear verification flows that accommodate different user contexts and devices. |
| Reduced user friction | By saving progress and enabling seamless desktop-to-mobile handoffs, you minimize user frustration and reduce the likelihood of session abandonment. |
| Granular risk control | Configurable thresholds for liveness, face match and document checks give you precise control over the trade-off between security and user experience. |
| Data-driven optimization | Actionable operational metrics provide the insights needed to monitor performance, identify bottlenecks and make informed decisions when tuning your configuration. |
Session management
SEON provides a robust session management system to handle the entire lifecycle of a verification.
Pre-initialized sessions
You can create verification sessions in advance and share a unique, secure URL with your users. This is ideal for asynchronous workflows, such as sending verification links via email or SMS.
- 7-Day validity: Sessions remain valid for 7 days, giving users ample time to complete the verification.
- 5 retries: Users are allowed a maximum of 5 retries within a single session if they make a mistake.
- Saved progress: If a user is interrupted, their progress is automatically saved, allowing them to resume where they left off.
- URL recovery: If a user loses their unique link, it can be re-copied from the session details page in the SEON Admin Panel.
Session transfer
To ensure a seamless experience, SEON automatically handles cases where users start on an incompatible device (e.g., a desktop computer without a camera).
- A QR code is automatically generated on the desktop screen.
- The user scans the QR code with their mobile device, creating a transferred session that is valid for 1 hour.
- The mobile session resumes at the exact step where the user left off.
- Once the mobile verification is complete, the results are automatically synced back to the original session and the user can continue on their desktop.
Session outcomes
Upon completion, every verification session is assigned a clear and actionable outcome. The primary statuses are:
- PENDING: The verification process is currently in progress.
- APPROVE: All required checks were successfully passed.
- REVIEW: The system could not reach a definitive conclusion and the session requires manual review.
- DECLINE: One or more checks failed, indicating a high risk of fraud or a non-compliant user.
- ABANDONED: The user started the verification process but did not complete it within the allowed time.
Configuration and tuning
SEON’s verification engine is highly configurable, allowing you to tune its performance to match your specific risk appetite and user base. The sections below provide configuration principles, a complete reference of all tunable parameters, and the key metrics to monitor.
Configuration principles
Follow these principles when tuning your settings:
- Start with defaults: SEON’s default thresholds represent a balanced approach to risk and user friction and are a great starting point.
- Loosen cautiously: If you relax a threshold, document the reason and monitor the impact on your metrics. Avoid making multiple changes at once.
- Consider per-market tuning: Document quality can vary by region. You may need to apply slightly different thresholds for markets with older or lower-quality identity documents.
- Monitor for drift: Fraud patterns and user behavior change over time. Regularly review your metrics to ensure your configuration remains effective.
Model configurations
The tables below detail every configurable parameter, organized by module.
Document check configurations
These settings control how document verification handles image quality, fraud detection and security feature validation.
Image quality thresholds
Control how strictly the system evaluates captured document images. All settings use Match levels (Level 1 – Level 5), where higher levels mean stricter matching criteria.
| Setting | Description | Default |
| Blur match level | Controls strictness of rejecting captured images based on the amount of blur. | Level 3 |
| Glare match level | Controls strictness of rejecting captured images based on the amount of gale. | Level 3 |
| Sharpness match level | Controls strictness of rejecting captured images based on the amount of sharpness. | Level 3 |
| Lighting match level | Controls strictness of rejecting captured images based on the amount of lighting. | Level 3 |
| Hand occlusion match level | Controls strictness of rejecting captured images based on the amount of hand occlusion. | Level 3 |
| DPI match level | Controls strictness of rejecting captured images based on the DPI of the capture. | Level 3 |
| Tilt match level | Controls strictness of rejecting captured images based on the amount of tilt. | Level 3 |
| Image quality interpretation | Handles poor-quality images without automatic failure, prioritizing user acceptance. | HighAssurance |
Fraud detection thresholds
Control how aggressively the system detects document manipulation and forgery attempts.
| Setting | Description | Default |
| Screen match level | Controls strictness of detecting screen-based document fraud. | Level 3 |
| Photocopy match level | Controls sensitivity to detecting photocopied documents. | Level 3 |
| Photo forgery match level | Controls detection of photo manipulation. | Level 3 |
| Barcode anomaly match level | Sets sensitivity for detecting barcode inconsistencies. | Level 3 |
| Static security fratures match level | Verifies document security features like holograms and watermarks. | Level 3 |
| Data match level | Checks data format consistency across the document. | Level 3 |
| Treat expiration as fraud | Flags expired documents as fraudulent if enabled. | True |
Document check behavior
| Setting | Description | Default |
| Strict last name match | Enforces strict matching for last names during verification. | Off |
| Extracted fields | Displays all Microblink-extracted data during document check in SEON. | Off |
Selfie and liveness check configurations
These settings control biometric verification, including liveness detection and face matching.
| Setting | Description | Default |
| Liveness check quality threshold | Quality score below which the liveness check is considered failing. | 0.5 |
| Liveness check probability thresholds | Failure, review and passing thresholds to consider when the quality threshold is met. | 0.4, 0.5 |
| Face matching similarity threshold | Failure, review and passing thresholds for the face matching component of liveness checks. | 0.4, 0.5 |
| Return extracted capture frame URLs | Returns signed S3 URLs of capture frames used in liveness and face match checks. | Off |
| Display face images | Shows extracted face images in session summary and liveness results in Admin. | Off |
Proof of address configurations
These settings control how proof of address documents are validated and matched.
| Setting | Description | Default |
| Name fuzzy matching | Allows slight name variations during verification. | Enabled |
| Address fuzzy matching | Allows minor address discrepancies during verification. | Enabled |
| Address validation | Validates address against official records, allowing minor variations but blocking major discrepancies. | Enabled |
| Document type validation | Accepts various document types without restriction. | Disabled |
| Document country validation | Accepts documents from any country without country-specific checks. | Disabled |
| Name matching similarity threshold | Similarity score above which compared names are considered matching. | 0.8 |
Duplicate detection configurations
These settings control how the system identifies potential duplicate submissions across sessions.
| Setting | Description | Default |
| Face similarity threshold | Similarity level above which two faces are considered as duplicates. | 0.3 |
| Face similarity limit | Maximum number of similar faces linked to a single session. | 25 |
Session configurations
These settings control session behavior, including timeouts and retry limits.
| Setting | Description | Default |
| Session lifetime | Defines how long a verification session stays active before it expires. | 1 hour |
| Session attempt limit | Limits how many times a session can be restarted before requiring a new one. | 5 |
| Check retry limit | Limits the number of retry attempts for actions like document verification. | 5 |
| Unique session validity period | Limit validity period of generated unique session links. | 1 week |
| Override declined sessions to review | Converts all declined sessions to review status after evaluation. | Off |
Webhook configurations
These settings control how SEON communicates verification results to your systems.
| Setting | Description | Default | Configurable |
| Webhook call URL | Webhook calls are sent to this URL. | Not set | Yes, in the SEON UI |
| Maximum webhook call retry count | Sets max retry attempts for calling the webhook before discarding. | 5 | Contact Customer Success |
| Webhook call timeout | Sets max time for webhook response before triggering a retry. | 10 seconds | Contact Customer Success |
| Webhook retry count | Defines time gaps between webhook retries, increasing gradually to reduce load. | [10s, 20s, 30s, 40s, 50s] | Contact Customer Success |
SDK configurations
These settings control the behavior and appearance of the SEON SDK integration.
| Setting | Description | Default | Configurable |
| SDK design customizations | Allows styling the SDK to match your brand across Web, iOS and Android. | SEON branding | Yes (SDK) |
| SDK base URL | Specifies regional SEON IDV service URL; must match account’s assigned region. | — | Yes (SDK) |
| SDK language ISO code | Specifies preferred language ISO code for the SDK; defaults to app locale or English. | Locale language | Yes (SDK) |
Operational metrics to track
Focus on these actionable metrics to guide your tuning decisions, rather than abstract academic measures like FRR/FAR.
| Metric | What it measures | Why it matters |
| Approval rate | The percentage of total sessions that result in a verified identity. | This is your primary success metric for frictionless, successful onboarding. |
| Rejection rate | The percentage of total sessions that are automatically declined. | A key indicator of fraud pressure, but a sudden spike could also signal that a threshold is too strict. |
| Abandonment rate | The percentage of users who start a session but do not complete it. | A high rate can signal user experience issues or problems with the capture process. |
| Forced entry rate | The percentage of users who are asked to re-capture an image. | A direct measure of capture quality and a signal that your users may be in a challenging environment (e.g., poor lighting). |
| Check-level failure distribution | A breakdown of which specific checks are failing most often. | This is your most important tuning signal, telling you exactly which thresholds may need adjustment. |
Need help with configuration?
Our team is here to assist with any tuning your organization requires. To request configuration changes or discuss optimization strategies, reach out to your Customer Success representative or contact us.