AEO and GEO glossary: the terms for AI search advertising

As of July 2026, AEO is the practice of getting a brand cited inside AI answers from engines like ChatGPT, Perplexity, and Google AI Overviews. Our ChatGPT capture covers 2,655 advertisers running 89,133 placements across 108 niches. AEO and GEO describe the same job: optimizing for citation inside a generated answer.

Why a glossary for AI search advertising

Search has moved from ten blue links to one AI answer. Rank became citation. Click became recommendation. The traditional SEO keyword list gave way to a new vocabulary for the new job: getting chosen by the model. The terms below are the ones advertisers and SEO teams now use to plan, measure, and buy against that job. For deeper competitive breakdowns by advertiser and niche, see our intelligence hub.

AEO (Answer Engine Optimization)

AEO is the practice of optimizing a brand, page, or dataset so an answer engine cites it as the response to a user's question. Where classic SEO chases a blue link at position one, AEO chases the citation embedded inside the model's answer. The work looks familiar: structured data, clear sources, an entity the model can resolve, and content that maps tightly to the questions users actually ask. The output is not a ranking, it is a named brand inside a generated paragraph.

GEO (Generative Engine Optimization)

GEO is a sibling acronym coined at the same time. In practice, AEO and GEO refer to the same activity, and the trade press uses them interchangeably. A minor distinction some teams keep: AEO emphasizes being the direct answer, while GEO emphasizes being woven into a longer generated response, the paragraph that follows the citation. Treat both terms as the same job, with subtle stylistic preferences.

Share of Voice and Answer SOV

Share of Voice, often shortened to SoV or Answer SOV, is the share of an answer engine's recommendations that name a given brand. It replaces the old rank or impression share metric from search ads. Compute it by counting how many answers to a target prompt list your brand, divided by total answers counted for that prompt. As of July 2026, Mastercard leads our ChatGPT placements at 3.92% share of voice, with Monday.com at 3.30% and Salesforce at 2.01% rounding out the top tier.

Trigger prompt

A trigger prompt is the user query that causes an answer engine to surface an ad or recommendation in the first place. The same query can trigger different brands on different engines, which is why advertisers now run prompt-level tests the way they once ran keyword A/B tests. In our capture, prompts like "best platform for hiring across Europe with local compliance support" are logged as active trigger prompts, and the same query often surfaces several advertisers at once. That is why prompt-level tracking has replaced keyword-level tracking for many teams.

Placement (and placement vs. impression)

A placement is one instance of an ad appearing inside one AI answer. As of July 2026, our ChatGPT capture logs 89,133 placements across 8,803 distinct ads. A placement is not the same as a search impression: one placement counts a single ad inside a single answer, while an impression repeats for every user view of that answer. The shift from impression-based to placement-based measurement is one reason legacy SEO reporting has had to be rebuilt for AI search.

Advertiser and niche

An advertiser is the brand whose product is being promoted inside the AI answer. A niche is the vertical the question belongs to, such as hiring, security, real estate, or fintech. As of July 2026, we track 2,655 advertisers spread across 108 niches, with top advertisers like Mastercard logging thousands of placements each. Niche matters because trigger prompts cluster by topic, and a brand's share of voice only makes sense within the niche its prompts belong to.

More AEO and GEO terms

A few shorthand terms come up often in AEO and GEO work. AI Overview is Google's name for the AI-generated answer block that appears above traditional results. Brand mention is any time the model names a brand, cited or not, and AEO teams track mentions as a leading indicator of citation. Citation rate is the share of answers that name a brand out of answers that name any brand in the same category. Position zero refers to being cited above other cited brands in a generated response. Query fan-out is the model's habit of breaking one user prompt into many sub-queries, which is why a single user prompt can produce placements for several advertisers at once. Zero-click describes an answer that fully resolves the user's question without any click or site visit.

Frequently asked