This presentation provides a short conceptual introduction to the CF (Climate and Forecast) Conventions, presenting CF not simply as a set of file-level rules for netCDF, but as a semantic and structural metadata model for representing quantities over a domain. The talk is aimed at climate data scientists and technical practitioners who work with climate data products and need a clearer mental model of how CF expresses meaning, comparability, and interpretability. The presentation covers these main ideas: CF as a data model rather than a checklist of attributes or a collection of format-specific tricks; how meaning in CF emerges from the combination of descriptive and structural metadata, beyond `standard_name` alone why richer metadata matters in practice, using examples based on aggregations, thresholds, and climate indices; The material was adapted from earlier work prepared in 2017 at ECMWF on recommendations for encoding products in netCDF using CF, but has been substantially refactored and updated here with a stronger emphasis on the CF data model, semantic equivalence, and practical interpretation of metadata constructs such as coordinates, `cell_methods`, `bounds`, and threshold coordinates.