Identity Graph

An identity graph is a data structure that maps many identifiers — cookies, devices, logins and channels — to one entity, so a person is seen as the same across every touchpoint.

An identity graph is the engine behind an identity cluster: each identifier is a node, each piece of linking evidence an edge, and the graph decides which nodes resolve to one person. Its uses are solving the cross-device user, removing duplicate counting and multi-touch attribution. But an identity graph is as sensitive as it is powerful: one wrong edge merges two strangers into a single identity and mixes two people's data, and conversely a missing edge keeps one person as two. From a privacy standpoint, an identity graph is exactly what data-protection regulation worries about, because it builds a complete view of a person. This is why building and maintaining one is mostly the work of customer data platforms and specialized identity tools, not a simple web analytics script.