Digital Lexicon Hub

The process of linking identifiers such as emails, device IDs and account numbers that belong to the same person or company into one profile.

Without a resolved identity, one buyer appears as several anonymous visitors, leads and contacts, which inflates audience counts, breaks attribution and causes sales and marketing to contact the same person with conflicting messages.

A SaaS company configures identity rules in its CDP so that an anonymous web ID merges into a profile when the visitor submits a form with a work email, and the same email links the CRM contact ID and product user ID.

Merge rules that are too loose join different people, for example on shared devices or generic inboxes, and wrong merges are hard to undo. Probabilistic matching adds reach but lowers accuracy, and linking identifiers may require consent under GDPR.

Deterministic matching joins records that share an exact identifier, such as a hashed email or login ID, while probabilistic or ML matching scores similarity across fields like name, address and IP. An identity graph stores the links, and rules set which identifiers must be unique and which sources may trigger merges. Inputs are normalized first, for example lowercased and trimmed. It underpins Customer Data Platform (CDP) profiles and Cross-Device Tracking.