Glossary · Business Analytics
Identity Resolution
also: identity stitching · identity graph · probabilistic identity · deterministic identity
Definition
Identity resolution is the process of linking pseudonymous user signals (cookies, device IDs, IP/UA fingerprints) into a coherent person-level view. Logged-in deterministic matches are typically 8% to 18% of traffic; probabilistic matches fill the remainder with calibrated confidence scores rather than absolute IDs.
Identity resolution sits at the center of the post-cookie measurement stack. The spectrum runs from deterministic (email hash, logged-in ID, with effective 100% confidence when present) through probabilistic (IP plus user-agent plus behavior, with confidence scores typically in the 0.55 to 0.85 range) to aggregated cohort approaches (Topics API, FLEDGE/Protected Audience). The 'match rate' problem is the dominant operational constraint: most consumer sites have 8% to 18% logged-in traffic, so probabilistic methods carry the bulk of stitching work. Data clean rooms (LiveRamp, Snowflake Clean Room, AWS Clean Rooms) let two parties match on hashed identifiers without exposing the underlying PII.
Essays on this concept
- Business Analytics
Identity Resolution in a Cookieless World: A Probabilistic Reality
The cookie was always probabilistic. Cookieless makes the probability legible. Operators who treat new identifiers as deterministic will misattribute spend and contaminate downstream measurement.
- Business Analytics
Server-Side Tagging Beyond Compliance: The Operational Case
Privacy compliance is the entry point for server-side tagging. The operational case is broader: latency, ad-blocker resilience, data quality, and the cost model of running an event router at production scale.
- Business Analytics
Event Taxonomy Design as Data Engineering
Event taxonomies are schema problems, not marketing problems. Teams that treat tracking plans as living documents (with versioning, validation, and PII boundaries) avoid the drift that quietly costs everyone else.
- Business Analytics
Privacy-Preserving Analytics: Differential Privacy in Practice
Differential privacy is a formal guarantee about what an analyst can learn from a dataset. The operational question is when the guarantee is worth its accuracy cost, and when a weaker model is the honest answer.
- Business Analytics
Cohort Analysis at the Action-Set Level (Not User-Level)
Sign-up-month cohorts confuse arrival with behavior. Action-set cohorts predict retention earlier and more honestly, at the cost of an event taxonomy, materialized views, and resolved identity discipline.
- Business Analytics
The GA4 Transition Forensics: What Universal Analytics Did Better
An honest post-mortem of the UA to GA4 migration. What broke, what is genuinely better, what remains unchanged, and the opportunity cost question that nobody at Google wants to discuss in public.
- Marketing Engineering
Unified Measurement Architecture: Connecting MMM, MTA, and Experimentation Into a Single Source of Truth
MMM says Facebook works. MTA says Google works. The incrementality test says neither works as well as you thought. Three measurement systems, three different answers, here's how to reconcile them into one coherent picture.
Related concepts
Authoritative references