What is identity resolution?

Unifying a customer's scattered identifiers — devices, emails, logins — into one coherent profile.

What's the difference between deterministic and probabilistic matching?

Deterministic links via known shared IDs (high confidence); probabilistic models likely matches from behavioral signals (broader, less certain).

Why does it matter more now?

As third-party cookies and cross-app IDs vanished, first-party, consent-based identity resolution became the durable spine for personalization and measurement.

Growth Marketing Glossary

Identity Resolution

i·den·ti·ty res·o·lu·tion/aɪˈdɛntəˌti ˌɹɛzəˈluʃən/noun

One person, twelve devices, three emails, and a dozen sessions — identity resolution is the work of realizing they're all the same human.

Schematic — fragments resolved into one profile

Term: Identity Resolution
Goal: One unified profile per real person
Methods: Deterministic (logins, email) vs. probabilistic (modeled)
Home: CDPs, clean rooms, first-party graphs

Forms & parts of speech

identity graph · phrase

The map of who-is-who.

"The identity graph finally connected the app user and the email subscriber — same person, double-counted for years."

Definition in plain terms

Identity resolution is the process of recognizing that a customer's many scattered identifiers — different devices, browsers, email addresses, app logins, loyalty IDs, session cookies — all belong to ONE real person, and stitching them into a single unified profile. Without it, the same customer appears as a dozen strangers across systems, fragmenting personalization, inflating audience counts, and breaking measurement. With it, marketing can recognize and serve the whole person rather than disconnected fragments.

The mechanics

Two methods anchor it: DETERMINISTIC matching (linking via known shared identifiers — a logged-in email, a loyalty ID — high confidence, the gold standard) and PROBABILISTIC matching (modeling that two devices are likely the same person from behavioral and contextual signals — broader reach, lower certainty). The output is an IDENTITY GRAPH connecting the fragments. In the privacy era this shifted decisively toward FIRST-PARTY, consent-based, deterministic resolution: as third-party cookies and cross-app IDs vanished, the durable identity spine became the customer's relationship with YOU (logins, emails, accounts), often unified in a customer data platform (CDP) and matched with partners inside privacy-preserving data clean rooms. The governing constraint is consent — identity resolution done without it is the surveillance the privacy era exists to curb.

When it matters

Identity resolution underpins everything that requires knowing WHO a customer is: personalization, suppression (don't retarget someone who just bought), cross-channel frequency capping, accurate CLV and cohort analysis, and measurement that isn't double-counting one person as many. It became strategically central exactly as cross-site/cross-app tracking collapsed — the brands that can resolve identity from their own first-party relationships hold an asset their cookie-dependent competitors lost. It's the connective layer between scattered data and any coherent view of the customer.

Worked example. A retailer treats its email subscribers, app users, and website visitors as three separate audiences — and unknowingly markets to the same customers three times, double-counts them in every report, and retargets people who already bought in-app. An identity-resolution project unifies them: a CDP stitches logged-in sessions, email, and loyalty IDs deterministically into one profile per person, with consent tracked throughout. Suddenly suppression works (no more ads to recent buyers), frequency capping spans channels, CLV is calculated on whole customers instead of fragments, and personalization recognizes the person across surfaces. The data was always there — identity resolution finally made it about people instead of pseudonyms.

Failure modes to watch. Probabilistic matching treated as certain; resolving identity without consent (the privacy-era trap); leaning on third-party graphs that degraded with cookies; and skipping it, so the same customer is double-counted and cross-targeted as strangers.

Synonyms & antonyms

Synonyms

identity resolutionidentity graphcustomer identity stitching

Antonyms

fragmented identityanonymous-only profiles

Origin & history

*No single coiner is on record; the account below is reconstructed from how the industry actually used the term. Identity resolution grew from database-marketing and customer-data-integration practice (the 'single customer view' of 1990s-2000s CRM); the term sharpened in the 2010s martech era with CDPs and identity graphs, and became strategically central as the privacy era forced the shift to first-party, consent-based resolution.

Etymology: source.

Usage trends

Search interest for this term over the last five years:

View interest-over-time on Google Trends →

Common questions

What is identity resolution?: Unifying a customer's scattered identifiers — devices, emails, logins — into one coherent profile.
What's the difference between deterministic and probabilistic matching?: Deterministic links via known shared IDs (high confidence); probabilistic models likely matches from behavioral signals (broader, less certain).
Why does it matter more now?: As third-party cookies and cross-app IDs vanished, first-party, consent-based identity resolution became the durable spine for personalization and measurement.

Related tools & calculators

toolCAC calculator
toolLTV-to-CAC ratio

Resources & people to follow

referenceCDP Institute — identity resolution frameworks
referenceData clean room / privacy-preserving matching documentation
referenceRGM analysis — consent-based first-party resolution is the durable asset

Curated, non-competitor resources verified per term.

Related training

moduleMarketing analytics

Disciplines

Areas of marketing where identity resolution is a core concern:

MeasurementGrowth strategy

Sources

trendsGoogle Trends — "identity resolution"