What is LLM Citations?
LLM citations are sources, links, or attributions an AI answer shows when it grounds a claim in retrieval or browsing. In visibility audits, they also mean how often a model names your brand with enough context to count as a substantiated mention on buyer prompts.
Teams track citations alongside plain mentions because a sourced or comparison-style answer usually reflects stronger association than a one-line name drop. That is what most buyers mean by “citations in an LLM” when they review ChatGPT or Perplexity output.
| Signal | Looks like | Visibility takeaway |
|---|---|---|
| Explicit citation | Linked or numbered sources in the UI | Strong provenance; good for trust-heavy categories |
| Named recommendation | “Use [Brand] for …” without a link | Counts toward recall even when no URL is shown |
| Generic mention | Category advice with no brand | Gap to fix with positioning and third-party coverage |
What are citations in an LLM?
They are references the model attaches to a claim: URLs, document titles, or numbered source lists when retrieval is on. When retrieval is off, you may only see confident wording without links — audits still score whether your brand is named with substantiating detail.
What does LLM citations mean for my brand?
It means how often models cite or clearly attribute recommendations to you versus competitors on realistic buyer questions. Improving it usually requires third-party comparisons and review surfaces models already trust, not only on-site copy.
Why it matters for marketing agencies
Agencies reporting on AI visibility need a shared vocabulary with clients. Citations and named recommendations are the proof points clients recognize in ChatGPT-style answers. Framing work around citation and recommendation quality aligns deliverables with what executives see in AI tools, not only SEO dashboards.