Glossary
The AI visibility glossary
AI visibility is how often and how favorably AI assistants mention you when people ask. This glossary defines the category terms in plain English: GEO, AEO, AI brand monitoring, share of voice, AI sentiment, grounding, citations, and the difference between model-only and compare. Saidly measures all of them across Claude, ChatGPT, Gemini, and Grok.
What do the AI visibility terms mean?
Each definition below is short enough to quote and specific enough to act on. The first three (GEO, AEO, AI visibility) are the umbrella terms most people search for. The rest are the working parts you will see inside any AI monitoring tool, including Saidly.
- GEO Generative Engine Optimization
- GEO is the practice of shaping how AI assistants describe and cite you when people ask them questions. Where classic SEO competes for the ten blue links on a results page, GEO competes for a place inside the single generated answer the model writes. The work includes publishing clear, citable content, earning mentions on sources the models trust, and watching whether those changes move the answer. You cannot improve what you cannot see, so GEO starts with measurement. See the AI visibility monitoring guide for the full playbook.
- AEO Answer Engine Optimization
- AEO is the work of getting your brand into direct AI answers rather than a ranked list of links. It overlaps heavily with GEO, and many people use the two terms interchangeably. The slight difference: AEO is often used for any answer engine, including AI Overviews and Perplexity, while GEO is most often used for the generative chat assistants. Both aim at the same prize, the one answer a person reads instead of clicking through results.
- AI visibility
- AI visibility is how often, how prominently, and how favorably an AI assistant mentions a brand, product, person, or topic when people ask about that category. It is the outcome that GEO and AEO try to improve. Saidly turns AI visibility into numbers: a 0 to 100 sentiment score per model, a representative quote, recurring themes, the sources each model cited, and share of voice against competitors. See what ChatGPT says about your brand for a single-model example.
- AI brand monitoring
- AI brand monitoring is tracking what AI assistants say about a brand over time, on a schedule, rather than checking once by hand. It is the AI-era version of media monitoring or search-result monitoring. A monitoring tool re-asks the models on a cadence you set, records each answer, and flags when sentiment, themes, or sources shift. Saidly runs monthly, weekly, or daily by plan, plus on-demand reports, and delivers them to your inbox and dashboard.
- Share of voice is the percentage of relevant AI answers that mention you versus your competitors. If five names come up in a category and you appear in two of them, your share of voice is roughly 40 percent. It tells you whether the models even know you exist in your space, separate from whether they like you. Saidly tracks share of voice per model, so you can see exactly where you lead and where a rival owns the answer. Read more on competitor share of voice.
- AI sentiment
- AI sentiment is how positive, neutral, or negative an AI assistant's answer about you is. It is the tone of the description, not just whether you were named. A brand can have high share of voice and poor sentiment if the models mention it but describe it in a bad light. Saidly is sentiment-first: it reads each model's full answer and scores the tone. For a deeper look, see AI brand sentiment.
- Grounding / web search
- Grounding is when an AI assistant searches the live web before it answers, then bases its reply on the pages it just read. A grounded answer reflects today's internet, including recent news and reviews. An ungrounded answer relies only on the model's training, which has a cutoff date and can lag reality by months. All four models Saidly tracks (Claude, ChatGPT, Gemini, Grok) are queried with grounding on, so reports match what real people see.
- Citations (cited sources)
- Citations, also called cited sources, are the web pages a grounded AI answer links to as its evidence. They tell you which sites are shaping the model's view of you, so you know where to publish, earn coverage, or correct the record. Saidly captures the cited sources behind each grounded answer, so improving your AI visibility becomes a concrete list of pages to influence rather than a guess.
- Model-only vs compare
- Model-only is a report run with web search turned off, showing what a model believes from its training alone, with no live pages and no citations. Compare runs the same question twice, once grounded on the web and once model-only, and shows the gap between the two. A wide gap is the useful signal: it means the live internet has moved but the model's built-in view has not caught up yet. Both lenses are available on Saidly's Pro and Business plans. The FAQ walks through when to use each.
- Sentiment score
- A sentiment score is a single number, from 0 to 100, that summarizes how positive an AI assistant's answer about you is. Saidly produces one score per model and shows the change since the last report, so a drop or a climb is obvious at a glance. The score sits next to a representative quote so the number always has the words behind it.
- Recurring themes
- Recurring themes are the ideas that keep showing up across the models' answers, such as "affordable," "hard to set up," or "great support." They explain why a sentiment score is where it is and tell you what to reinforce or fix. Saidly pulls recurring themes out of every report so you are not left reading raw answers to find the pattern.
- Named entity
- A named entity is whatever you track: a company, a product, a person (a founder, executive, or creator), an organization or non-profit, a political campaign, or a topic. AI visibility is not just for websites. Saidly tracks any named entity and frames the question by type, so a person is scored differently from a product.
How is GEO different from SEO?
SEO and GEO share a goal, being found, but they target different surfaces. SEO competes for rank on a page of links. GEO competes for inclusion in a single generated answer that most people never click past. The table sums up the practical differences.
| SEO | GEO / AEO | |
|---|---|---|
| Surface | Page of ranked links | One generated answer |
| Where people read it | Search results | Inside an AI chat |
| Success looks like | Ranking in the top results | Being named, cited, and described well |
| What you measure | Position and clicks | Sentiment, share of voice, citations |
| How you check it | Rank trackers | AI brand monitoring (Saidly) |
Where does Saidly fit in?
Saidly, built by Woodfire Digital, LLC, is an AI brand monitoring tool that measures every term on this page across four AI assistants: Claude, ChatGPT, Gemini, and Grok. It scores AI sentiment 0 to 100 per model, captures a quote and the cited sources, surfaces recurring themes, and tracks share of voice against competitors, on a schedule or on demand. Pricing is fully public: Starter $2.99/mo, Pro $13.99/mo, Business $49.99/mo, and a custom Enterprise tier, with a 30-day free trial and no credit card to start. For the full method behind the numbers, read the methodology.
See these terms on your own name
The fastest way to learn the vocabulary is to watch it applied to a name you care about. Add your company, product, or your own name and Saidly will run the four models, score the sentiment, and lay out the themes, citations, and share of voice for you.