Customer data glossary

Identity resolution

By Art Kukharenko · Published · Updated

Browse learn glossary resources

Identity resolution is the process of matching records from different sources that refer to the same person or household, so that one profile replaces many fragmented ones.

Part of the Learn hub and the customer data glossary.

What it means in practice

Customer records arrive with different identifiers: an email address in the marketing tool, an account number in core systems, a device ID from the app. Identity resolution decides which belong together. The result is usually stored as links in an identity graph and surfaced as a unified profile inside a customer data platform (CDP).

Two families of method are used. Deterministic matching links records on exact, trusted identifiers. Probabilistic matching scores partial agreement between attributes such as name, address and date of birth. Most production systems combine them: deterministic rules first, probabilistic scoring for what is left, and human review for the uncertain middle.

The central trade-off is between a false merge (two people combined into one) and a missed match (one person left split). Which error is worse depends on the use. Sending a duplicate newsletter is a nuisance; showing one person's information to another is an incident. Good implementations record why each link exists, allow a merge to be reversed, and carry the most restrictive consent state forward when profiles combine.

Example (healthcare): a health system has a patient-portal account, a call-centre record and a pharmacy-loyalty entry that may belong to one person. Matching on a verified portal login is safe enough to personalise an appointment reminder. Matching on name and postcode alone is not safe for anything that reveals health context, because a wrong match could expose one patient's information to someone else. The threshold and the permitted uses should be set per use case.

Where it matters

Frequently asked questions

What is the difference between identity resolution and an identity graph?

Identity resolution is the process of deciding which identifiers belong to the same person. The identity graph is the stored result: the identifiers, the links between them and the evidence for each link.

How accurate is identity resolution?

There is no universal figure. Accuracy depends on the identifiers available, the matching rules and the threshold. Measure precision (how many merges were right) and recall (how many true matches were found) on a labelled sample of your own data.

Is identity resolution personal data processing under GDPR?

Yes. Linking identifiers to a person is processing of personal data, so it needs a lawful basis and a defined purpose, and the links themselves must be covered by access and deletion requests.

Back to the customer data glossary or the Learn hub.

Further reading

Published: · Updated:

Publication date recorded for the approved release. Editorial updates do not establish legal approval or a customer result.

Need help?

Ask the PrivyCDP Assistant