In web analytics, a User is a distinct person–browser/app identity that interacts with your site or product over time. Practically, it’s computed by deduplicating events that share the same identifier(s) within a date range—typically a client cookie, device ID, or logged-in account. Unlike a Session (a time-bounded visit) or a Pageview (a single view), a User represents the entity behind multiple sessions and many hits.
Working definition:Users(R) = count of distinct identities observed in date range R after applying your deduplication rules.
Identity can be simple (one cookie = one user) or advanced (merge by login ID across devices; fall back to device ID). Your rules matter: change the stitching logic, and your User count changes.
Mini example (identity stitching)
Events in the same month:
- e1: cookie
cA
, not logged in - e2: cookie
cB
, loginu123
- e3: cookie
cC
, loginu123
(mobile)
After stitching by login, cB
and cC
collapse into one user (u123
). cA
stays separate.
Users = 2
Record | Cookie | Login | Device | Resolved User |
---|---|---|---|---|
e1 | cA | — | Desktop | U1 |
e2 | cB | u123 | Desktop | U2 |
e3 | cC | u123 | Mobile | U2 |
Why it matters
- Reach & cohorts: Users tell you how many distinct identities you engaged, enabling cohorting, LTV, churn, and frequency analysis.
- Non-additivity: Users over Aug 1–31 ≠ Users Aug 1–15 + Aug 16–31 (overlap!). Aggregate with care.
- Quality signals: Pair Users with Unique Pageview, Bounce Rate, and Engagement Rate to separate footprint from behavior quality.
Common pitfalls (and how to think about them)
- Cookie churn & ITP: Deletions and browser caps fragment identity; expect undercounting on strict browsers.
- Cross-device split: Same person on phone + laptop can appear as two Users without login-based stitching.
- Bot noise: Poor bot filtering inflates Users; keep filters tight.
- Scope drift: Don’t mix environments (web/app) or domains without consistent identifiers, or your Users become apples + oranges.
Related terms: Unique Visitor (often used interchangeably; check your platform’s exact rules), Session, Pageview.