Study ties ChatGPT, Gemini brand mentions to domain citation
What this answers
- Does being cited by name make a brand more likely to be mentioned by an AI engine again?
- Yes, per an analysis of AI-visibility monitoring data: prior-run mentions plus current own-domain exposure produced next-run mention rates of 80.5% (GPT) and 83.7% (Gemini), versus 1.6% and 1.9% when a brand had neither.
- How much does having your own domain cited raise your odds of being named in a ChatGPT or Gemini answer?
- Roughly 17x on GPT (2.8% to 49.0%) and 15x on Gemini (3.8% to 58.4%), rising to 91.4% and 100% when a branded citation also appears, per the same analysis.
An arXiv paper analyzing 34,960 unbranded prompt-engine observations across 75 anonymized client projects tracked by AI-visibility platform Aiso, covering 2,854 monitored prompts and repeated GPT and Gemini runs from June through September 2026, found brand mention rates of just 2.8% (GPT) and 3.8% (Gemini) when neither the brand nor its own domain appeared in the retrieval path the engine surfaced. Mentions rose to 49.0% and 58.4% with an own-domain citation but no branded fan-out, and to 91.4% and 100% when both own-domain exposure and a branded citation occurred together. Prior-run visibility also persisted independently: a previous non-mention combined with no current domain exposure produced next-run mention rates of 1.6% and 1.9%, versus 80.5% and 83.7% when a prior mention was paired with current exposure.
Why it matters: It quantifies, across 75 client accounts rather than a handful of prompts, that first getting cited compounds into future visibility — raising the stakes of the first citation an AEO campaign wins.
Via arXiv AI-search citations ↗
Posted to the wire September 28, 2026. Edited by Joe Balewski.