---
Source: https://docs.microblink.com/platform/capabilities/trust-score
Title: Trust score
Description: Calculate composite trust scores from multiple verification checks
---

# Trust score

:::tip[Availability]
Available in SDK v1.2+
:::

The Trust Score capability calculates a single composite score for the results of previous background checks and biometrics.

The following capabilities are included in the final trust score:

- [Personal data checks](personal-data-checks.md)
- [PEP, AML & Adverse Media checks](pep-aml-am-checks.md)
- [Selfie, Liveness & Face Match](selfie-face-match.md)

Document-related steps (Scan ID and Scan & Verify ID) don't contribute to the trust score.

## How it works

Each relevant check produces a score after it runs.
You assign a **weight** (0 to 5) to each of those checks:

- **0**: the check has no impact on the trust score
- **1 to 5**: the check contributes to the trust score, with higher values giving the check proportionally more influence

The weighted scores are then combined into a single trust score.
By default, all checks carry the same weight.

You then define a threshold: if the combined trust score falls below that threshold, the step can be configured to result in a **Rejected** outcome or send the transaction to **Manual review**.

## Placement in the workflow

Trust Score can be placed at different points in the workflow depending on your needs:

- **Mid-workflow**: place it after background checks but before additional verification steps.
  This lets you route high-risk users through a stricter path (for example, requiring a selfie or additional document checks) while letting low-risk users proceed directly.

- **End of workflow**: place it at the end to produce an overall risk signal for the transaction record, without branching.

For workflow branching based on the trust score outcome, combine this capability with an [If/else branch](if-else-branch.md).


Last updated on Sep 25, 2026