Identity Graph

A database that maps and connects all known identifiers for a single person - such as email addresses, device IDs, cookie IDs, and phone numbers - into a unified profile that represents one real human.

Also known as: identity resolution graph, ID graph

Why It Matters

A single person might interact with your brand through a dozen different identifiers: a mobile device ID, a desktop cookie, a work email, a personal email, a social login, an in-store loyalty number, and more. Without an identity graph, that one person looks like a dozen different users in your analytics, inflating your user counts and fragmenting your data.

Identity graphs solve this by linking identifiers together based on deterministic matches (same login across devices) and probabilistic signals (same IP plus similar browsing patterns). The result is an accurate count of real people and complete cross-device, cross-channel profiles for each one.

For analytics accuracy, identity resolution is foundational. Conversion rates, retention rates, lifetime value calculations, and cohort analyses are all wrong if you are counting one person as multiple users. Getting identity right is the prerequisite for getting every downstream metric right.

Industry Applications

E-commerce

A multi-channel retailer builds an identity graph that links online customer IDs with in-store loyalty card numbers. They discover that their "online-only" customers actually make 30% of their purchases in stores, fundamentally changing the ROI calculation for their digital marketing spend.

SaaS

A B2B platform uses an identity graph to connect individual user activity to company accounts. When three people from the same company are all using the free tier independently, the identity graph connects them, triggering an enterprise sales outreach that converts them to a team plan.

How to Track in KISSmetrics

KISSmetrics maintains an identity graph that links anonymous IDs, email addresses, and custom identifiers into unified person profiles. The alias method lets you explicitly connect identifiers when you know they belong to the same person. KISSmetrics automatically resolves identities across sessions and devices when a user authenticates.

Common Mistakes

  • -Merging identifiers with low-confidence matches, which incorrectly combines profiles of different people
  • -Not handling identity splits when shared devices cause two different people to be merged into one profile
  • -Ignoring privacy regulations that restrict the combination of certain identifier types without consent
  • -Assuming identity resolution is a solved problem - it requires ongoing maintenance as new identifiers and channels emerge

Pro Tips

  • +Use deterministic matching (same authenticated login) as the backbone of your identity graph and add probabilistic matching cautiously
  • +Implement quality checks that flag suspicious merges - like two profiles with different billing addresses or plan types being combined
  • +Design your identity graph to handle splits (un-merging accidentally combined profiles) not just joins
  • +Track your identity resolution rate (percentage of events tied to identified users) as a data quality metric
  • +Consider the order of precedence when identifiers conflict - which source of truth wins?

Related Terms

See Identity Graph in action

KISSmetrics tracks every user across sessions and devices so you can measure what matters. Start free - no credit card required.