Home/Glossary
Glossary
SEO, GEO and AI terms, short and practical: what it means, how it works, where people get it wrong.
| Term | Meaning |
|---|---|
| SEO12 | |
| E-E-A-TExperience, Expertise, Authoritativeness, Trust | Experience, Expertise, Authoritativeness and Trust — the frame Google raters use to judge page quality. |
| Canonicalrel=canonical | The tag by which a page declares its primary address. |
| Core Web VitalsLCP, INP, CLS | Google’s three field metrics: LCP, INP and CLS. |
| Crawl budget | How many pages a bot is willing to fetch on your site over a period. |
| Keyword set | The collected and grouped list of queries a site targets. |
| Query clustering | Splitting queries into groups by how much their result pages overlap. |
| Anchor text | The visible text of a link. |
| Link profile | The whole set of external links: who links, with what anchors, how fast. |
| Linking domain | A site that links to yours. |
| Yandex fastbot | A crawler that revisits pages almost immediately after publication. |
| Behavioural fraud | Faking user behaviour to lift rankings. |
| Structured data (Schema.org)JSON-LD | A machine-readable description of what is on the page. |
| GEO3 | |
| GEOGenerative Engine Optimization | Generative Engine Optimization — getting your material cited inside AI answers. |
| AI Overviews | The AI answer block above Google organic results. |
| llms.txt | A plain-text file at the site root describing the project for AI crawlers. |
| AI16 | |
| LLMLarge Language Model | A large language model — a network trained to predict the next chunk of text. |
| Hallucination | A fact invented by a model and delivered in the same tone as a true one. |
| Embedding | A numeric vector a text is turned into, so meanings can be compared arithmetically. |
| RAGRetrieval-Augmented Generation | Retrieval-Augmented Generation: the model fetches documents first and answers from them. |
| Prompt | The instruction given to a model: role, task, constraints, output format. |
| Token | The chunk a model works with: a word, a piece of a word or a punctuation mark. |
| Text watermark | A hidden mark inside the text itself: not a character and not a file signature, but a slight bias in word choice. |
| Z-score | How far a measurement sits from chance — the number a detector uses to decide whether a mark is present. |
| Zero-width charactersinvisible Unicode | Unicode characters that exist in the text but render as nothing. |
| AI detector | A service that guesses from style how likely it is that a text was machine-written. |
| Watermark detector | A check of one specific mark rather than of style: the detector knows the secret the mark was made with. |
| Metadata | Service data inside a file: camera, date, software, sometimes an “AI-generated” flag. |
| C2PACoalition for Content Provenance and Authenticity | An industry standard that attaches a signed history to a file: what made it, what edited it, when. |
| Content CredentialsC2PA in practice | The human-facing side of C2PA: a badge on the image and a panel with its edit history. |
| SynthID | Google’s technology that embeds a mark into the pixels themselves rather than into metadata. |
| Provenance | The verifiable history of a file: what created it, what happened to it, who vouches for that. |