A conversational query is one turn in a multi-turn AI search conversation that the engine treats as its own independent query, rather than as part of the opening question. It's the individual unit of a conversational search session — where that term names the dialogue-style mode, a conversational query is a single measurable turn inside it.
The concept became concrete in mid-2026 when Google confirmed how AI Mode records conversations. Per Google's documentation, when a user asks a follow-up question within AI Mode, they are essentially performing a new query, and its impression, position, and click data are attributed to that new query. So a five-turn conversation produces five conversational queries, each logged separately.
Because follow-ups are replies, not fresh searches, conversational queries surface as short, dependent fragments — "yes, pricing", "go on", "the cheaper one" — that only make sense in context. In Google Search Console they land in the main performance report mixed with classic queries and carry no AI Mode label, so they can be spotted by shape but not cleanly segmented. For measurement, the implication is that AI visibility is now earned turn by turn: a brand can be cited at the opening query and dropped three turns later, which single-query tracking would miss entirely.