Event Taxonomy
Intermediate PronunciationEN/ɪˈvɛnt tækˈsɑːnəmi/
An event taxonomy is the classification and naming system for a product's events and properties, defining what gets recorded and under which name.
A taxonomy fixes three things: the list of allowed events, the naming rule, and the property set each event must carry. It differs from a tracking plan, which is the implementation document naming owners and code locations; the taxonomy is only the classification scheme itself. The usual failure is drift: one behaviour arrives under three different names and reports split into three without any error being raised.
FAQ
The taxonomy says which events exist and how they are structured; the tracking plan turns that taxonomy into work by recording the implementation location, the owner and the test status for each event. One is the model, the other is the plan to build it.
When analysts have to merge several events by hand to answer simple questions, or when one name has come to mean different things to different teams. Renaming live events breaks history, so the common path is defining a new event and deprecating the old one.