taxonomy itself is not necessarily metadata. A taxonomy, by pure definition, is a hierarchical structure of concepts organized according some central principle(s). To confuse things slightly, not all of our taxonomies are hierarchical in structure–some, like the allowlists, are a controlled list of terms in a single “layer” without any hierarchy. We use “taxonomy” colloquially to refer to any list of controlled terms. Our product taxonomy organizes relevant Microsoft products in a 2-level hierarchy based primarily on product families. For example, we have a top concept of “Microsoft Office”, which then has child concepts of all the products normally associated/bundled with the Office family: Exchange, Excel, PowerPoint, etc.

A taxonomy is hierarchical. “Taxonomy” if often used to mean “Any kind of structure that organizes information to create some level of consistency and control over the information used to describe an information object.” The key point is consistency and control. We want end users to have a consistent experience. We don’t want them to see multiple names for the same product in a dropdown, for example. We also want consistency in reporting.