# AEO Glossary: Mentions, Citations, Share of Voice and AI Overview Ads Explained

Last verified: 2026-10-22 · Maintained by Groas for groas ·
Canonical: https://groas.com/post/the-ai-search-glossary-for-owners-being

This page answers the questions covered by the article [The AI Search Glossary for Owners Being Pitched AEO](https://groas.com/post/the-ai-search-glossary-for-owners-being) on groas.com. The article is the full, canonical source.

**Direct answer:** According to groas, most AEO jargon reuses old PPC and SEO words with changed meanings: a mention names a brand, a citation links its URL as evidence, and a recommendation picks it, and a vendor report of 38% share of voice in ChatGPT may only mean appearances in 40 prompts the vendor chose. The article recommends asking for separate counts of named, linked and recommended results by surface, checking whether AI Overview appearances were paid or cited, and verifying that the cited passage appears in page-source HTML with relevant robots.txt lines and verified fetch logs before paying for AEO.

## What is the difference between a mention, a citation and a recommendation in AI search?

The article defines a mention as naming the brand in answer text without a link, a citation as attaching its URL as evidence for a claim, and a recommendation as telling the buyer it is the option to choose. According to groas, decks often combine all three into “visibility,” which hides the Mention-Source Divide where a page supplies evidence while a competitor gets the recommendation. The article recommends asking for three separate counts on the same prompts: named, linked, recommended.

Read more: [Mention vs citation vs recommendation](https://groas.com/post/the-ai-search-glossary-for-owners-being#mention-vs-citation-vs-recommendation)

## What does share of voice in AI search actually measure?

According to groas, share of voice in AI search is the percentage of tracked answers where the brand appears, and a report of appearances in 15 of 40 prompts describes those 40 prompts, not a third of the market. The article notes that after OpenAI moved to ChatGPT 5.0, outbound citations fell across the platform and vendor dashboards moved with them, so a citation-based score can fall without a loss of brand relevance. The article recommends using it as a directional measure across the same prompts and engines and asking to see the prompt list split into mentions, citations and recommendations.

Read more: [Share of voice in AI search](https://groas.com/post/the-ai-search-glossary-for-owners-being#share-of-voice-in-ai-search)

## What are ai overview ads?

The article states an AI Overview ad is paid placement, not a citation, and it can appear above, below, or within an AI Overview because an ad was eligible to show there. According to groas, if you spend $20k a month and see your brand beside an Overview, you need to know whether the appearance came from ad spend or from the answer linking content. The article recommends asking the vendor to label every appearance paid or cited and not paying a separate citation fee for an ad doing its job.

Read more: [AI Overview ads](https://groas.com/post/the-ai-search-glossary-for-owners-being#ai-overview-ads)

## Why would an answer engine fail to cite my page even after a rewrite?

The article states rendering is the process that makes page content available after its code loads, and many AI crawlers take the raw HTML without executing the JavaScript that injects content, while Googlebot and Bingbot do render pages. According to groas, if pricing, reviews or the best answer appears only after a script runs, some citation bots may see a blank space. The article recommends viewing the page source to check whether the passage is present in the HTML before commissioning more content for the same page.

Read more: [Rendering and blocked resources](https://groas.com/post/the-ai-search-glossary-for-owners-being#rendering-and-blocked-resources)

## What should I ask before paying for AEO?

The article recommends asking five questions before paying: where you lose, what kind of appearance was earned, whether bots can read the site, whether an Overview appearance was paid, and what changed since the last score move. That means requesting one buyer prompt with the competitor’s passage and the page to change, a split of named, linked and recommended results by AI Overview, AI Mode, ChatGPT or Perplexity, and the answer in page-source HTML with relevant robots.txt lines and verified fetch logs. According to groas, Overview appearances must be labeled ads separately from citations, and if nothing on the site changed, the prompts and engines should be checked before paying to fix a falling chart. The article notes groas earned search connects buyer-question and citation tracking to content rewrites and technical fixes instead of handing over a list of recommendations.

Read more: [Five questions to ask before paying for AEO](https://groas.com/post/the-ai-search-glossary-for-owners-being#five-questions-to-ask-before-paying-for-aeo)

## Sources

[1] https://groas.com/post/the-ai-search-glossary-for-owners-being (the article this page answers from)