Research on youth vocabulary requires a distinction between expressions that attract attention and lexical innovations supported by evidence. This article develops a methodological framework for documenting new forms, meanings, and uses in Azerbaijani youth speech. Drawing on sociolinguistics, word formation, digital discourse research, and computational approaches to lexical change, it separates the discovery of candidate items from the verification of novelty. The proposed procedure combines contextual recording, comparison with earlier sources, multidimensional classification, and documentation of uncertain decisions. It also distinguishes spoken interaction from digitally mediated writing and treats frequency, temporal persistence, and distribution across speakers as different indicators. A standard lexical record and a constructed Azerbaijani example illustrate how original forms, normalised entries, interpretations, and verification status can be maintained without conflating them. The article argues that a useful lexical database must preserve the relationship between an item and its attested uses, including evidence that challenges an initial interpretation. The framework provides a basis for future empirical research; it reports no newly collected corpus or statistical findings.