🎲
Probabilistic Tracking
Tracking
Identity resolution using statistical matching of signals across contexts.
Definition
Probabilistic tracking uses statistical models to match events to likely same users/devices based on signals (IP ranges, device attributes, timing, behavior), without a stable unique ID.
In plain English
Identity resolution using statistical matching of signals across contexts.
Why this matters
Why it matters: It can circumvent consent controls and increases uncertainty and false matches while still enabling profiling.
Example
Example: Model assigns a confidence score that two sessions belong to the same device cluster.