How does a data dictionary differ from an event taxonomy?

A taxonomy is the scheme: it decides how events are grouped and what pattern names follow. A dictionary records the instances: every name that actually exists, what it means, which properties it carries and who owns it. A taxonomy without a dictionary is never executed, and a dictionary without a taxonomy quickly loses order.

Data Dictionary

Intermediate

Updated Jul 11, 2026 Shahin editorial 2 min read

PronunciationEN/ˈdeɪtə ˈdɪkʃəˌnɛri/

A data dictionary is the shared, authoritative list that records the exact name, meaning, value type and owner of every event and property.

Each dictionary entry describes one event or one property: the agreed name, a sentence that locks down its exact meaning, the value type, the allowed values, whether it is required, whether it is active or deprecated, and the team that owns it. Its main job is to make two teams read one name the same way; without it, sign-up means a submitted form to marketing and a verified email to product, and the two reports never reconcile. A data dictionary enforces nothing on its own: until a validation layer rejects a non-conforming event, the document describes rather than controls. A dictionary written once and never revised is worse than none, because a stale document still commands trust.

FAQ

Sources:Mixpanel — Lexicon