Deterministic vs probabilistic identity matching
Deterministic matching joins identifiers on exact shared data, like the same login or email. Probabilistic matching infers a link from evidence like device and behavior. They trade precision for reach.
Deterministic matching
Deterministic matching links identifiers when they share exact data, most often a login. If the same account signs in on a laptop and a phone, the two are certainly the same person. It is precise and easy to trust, but it only works once a customer identifies themselves.
Probabilistic matching
Probabilistic matching infers that identifiers belong together from patterns: the same device, network, location, and behavior. It can connect activity before a customer logs in, which deterministic matching cannot. The cost is that it is a guess, and guesses are sometimes wrong.
Which to use
Most teams use deterministic matching where they can and probabilistic matching to fill the gaps. The right balance depends on how much certainty a decision needs. Charging a customer wants deterministic certainty; adjusting a layout can tolerate a probabilistic guess.
See it on your own traffic
Add the SDK, define one attribute, and see an attribute updating against your own traffic in the first session. The free tier runs on the real engine: no card, no sales call.