Track.js: User Analytics Without the Surveillance Tax

In 2014, user tracking meant Google Analytics, Mixpanel, or Kissmetrics. All three approached user identity the same way: identify users by their email address or a persistent cookie, build a profile around that identity, store everything forever.

This created a data liability every company was accumulating — reams of personal information that was useful for analytics but dangerous to hold. The breaches that would make this liability obvious in 2016-2018 hadn't happened yet, but the structural problem was already there.

Track.js approached user identity differently.

Anonymous-First Design

Track.js generated a random identifier for each user — not tied to any PII — and associated behavior with that identifier. Session data, page views, events, commerce interactions all attached to the anonymous identifier.

Identity resolution happened explicitly: if a user logged in, or provided an email, the anonymous identifier was linked to the identified record in a separate, access-controlled store. But the event log — the high-volume behavioral data — never contained PII.

The architecture had immediate benefits:

• The event log was safe to retain indefinitely (no PII exposure in a breach)
• Cross-device tracking worked without requiring login (probabilistic device fingerprint + deterministic identity resolution on login)
• Deletion requests were simple: delete the identity record, the anonymous event data is already de-identified

The Behavioral Model

Track.js was designed to answer the question that mattered for commerce: "What sequence of behaviors leads to conversion?" — not "Who is this person?"

The commerce intelligence we needed came from behavioral sequences: visited product page → added to cart → abandoned at shipping → returned three days later → converted. That sequence is analytically rich. The name and email address attached to it add legal liability without analytical value.

What Happened in 2018

When GDPR came into force in May 2018, companies scrambled to audit their data practices and build consent flows. Companies running Track.js had a structural advantage: their most sensitive behavioral data was already de-identified. The compliance work was a fraction of what surveillance-first analytics required.

This was not foresight — we built what was technically correct in 2014. It turned out that "technically correct" and "legally defensible" were the same architecture.

---

Track.js is at github.com/hanzo-js/track.js. The core principles of anonymous-first behavioral tracking are foundational to Hanzo Analytics.