A method using various data sets and algorithms to identify (with high probability) the same user across different devices and applications.
As each device has its own unique identifier (e.g. device ID and cookies), it’s hard to know whether it’s the same user on all those devices.
Probabilistic matching is used when deterministic matching can’t be used, such as when there’s no common identifier connecting one user to multiple devices (like an email or social media account). It may also be used in combination with deterministic matching to improve match rates and enrich user profiles.