People also ask
In practice
The two events have different causes and different fixes, which is why they are tracked separately. Mention without citation: the brand is in the model’s or retrieval’s consideration set — likely from broad mention footprint — but the answer draws its sourcing from other pages. The lever is citability: pages on the brand’s domain that answer the category’s questions in quotable, current, well-structured form, plus presence on the pages that do get cited.
Citation without mention: the brand’s page is being used as evidence — good — but the answer names competitors instead, often because the cited comparison lists them more prominently or describes them more completely. The lever is the placement layer: presence on and shape of the comparison pages the engines cite, so the brand’s entry is prominent and accurately described. Both: the strong quadrant — named and sourced — defended by keeping the citation layer fresh and the description consistent.
Neither: the gap matrix work — publishing where competitors are cited and you are absent, entity clarity so the engines can describe you, and the prompt-set baseline so progress is measurable. The diagnostic runs on the same monitoring infrastructure as everything in GEO: fixed prompts, repeated runs, per-answer recording of who was named and what was cited. The honesty constraint that shapes client conversations: the ratios are observable, the levers are actionable, and the causation between acting and appearing is still not publicly provable — so the programme reports movement, not guarantees.
See also: AI citations, AI brand mentions, Mention-to-citation workflows.
Related service: Brand mentions.
