Data Taxonomy
Intermediate PronunciationEN/ˈdeɪtə tækˈsɑːnəmi/
A data taxonomy is the classification structure for events, properties and their values, defining which family each item belongs to and how it is named.
A taxonomy has three layers: the subject grouping of events, one consistent naming pattern, and the list of allowed values for each property. It is not the same as a data dictionary; the dictionary holds the definition of each individual item, while the taxonomy is the order those items sit in. The common mistake is treating the taxonomy as a start-of-project document. Unless it is applied at the moment an event is defined, every team adds its own new name and the catalogue splits into parallel branches, visible as several same-meaning events with different names in reports.
FAQ
No. There should be one canonical taxonomy that every consumer reads from. Building a parallel taxonomy for a single dashboard recreates the exact problem the taxonomy was meant to solve: one concept with two names and two numbers.