Interpret Correlation Values

Correlation values indicate how strongly identity attributes—such as a person’s name—are associated with related contact information like an email address, phone number, or physical address.

Correlation values are not direct indicators of fraud or risk. Instead, they provide contextual insight that can be useful in scenarios such as:

• Supporting manual review workflows
• Evaluating contact channels before sending verification messages
• Validating recipients during peer-to-peer or account activity flows

How correlation values are used

Correlation values are returned alongside supporting explanatory indicators that describe factors contributing to the observed level of association between identity attributes.

These indicators help provide context around whether identity information appears consistent, partially consistent, or inconsistent, and should be considered as part of an overall evaluation rather than as standalone signals.

Interpreting strength of correlation

In general, correlation values represent a relative measure of confidence:

• Higher values indicate stronger alignment between identity attributes and contact information
• Lower values indicate weaker or unclear association

Correlation values may reflect:

• Strong alignment across multiple attributes
• Partial alignment (for example, name variations or incomplete data)
• Limited or no observable association based on available information

Because correlation depends on the quality and completeness of input data, values may vary across use cases and should be interpreted in context.

Best practices

• Use correlation values as supporting context, not as the sole basis for decisions
• Consider correlation alongside other risk signals and workflow logic
• Avoid treating correlation as deterministic proof of identity
• Apply additional verification steps when confidence is insufficient for the use case

Correlation values are designed to help teams better understand identity consistency, particularly in review or verification flows, while maintaining appropriate safeguards around sensitive data and decisioning logic.