Copilot is changing how people find information — but does that mean metadata, content types, and taxonomy no longer matter?

This session puts that question to the test. We'll compare document discovery results from a single library using structured metadata (managed metadata, content types, site columns) against the same content in a document library with no metadata. We will reason over both sets of data evaluating both for accuracy, relevance, and trustworthiness of results.

You'll walk away with a practical framework for deciding where to invest your time: when strong information architecture pays off, when agents can compensate for messier content, and how to set governance expectations for site owners and content managers as agents become part of everyday search.

This session is aimed at architects, knowledge managers, and site owners responsible for content structure and Copilot readiness — no development experience required