Glossary
The terms our guides use, defined plainly. Each entry says what the term means for a site owner, cites the primary source, and links to the guide or tool that goes further.
AI search
- AI citationA link to a source page inside an AI-written answer. Being cited is not the same as being mentioned: an answer can name a business without linking it, and link a page without naming the business.
- AI OverviewsGoogle's AI-written summaries at the top of some search results, with links to the pages they draw on. They come from Google's ordinary index, so ordinary SEO applies.
- AI visibilityHow often AI assistants name or cite a business when people ask the questions its buyers ask. There is no official score: it is measured by asking the same questions on a schedule.
- Generative engine optimization (GEO)Work aimed at getting a business named or cited in answers written by AI assistants. The name comes from a 2023 research paper; AEO and LLMO are other labels for the same thing.
- Query fan-outA technique where an AI search system splits one question into several related searches, runs them, and combines what it finds into one answer. Google says AI Overviews and AI Mode may use it.
- Share of voiceYour brand's mentions as a share of all brand mentions in the same set of answers. It shows whether competitors are gaining on you, which mention rate alone does not.
AI crawlers
- AI crawlerA program an AI company runs to fetch web pages: to build its assistant's search index, to collect training data, or to open a page a person asked about.
- ChatGPT-UserThe agent ChatGPT uses to open a page when a person asks it to. OpenAI says robots.txt rules may not apply to it, and it does not decide what appears in ChatGPT search.
- ClaudeBotAnthropic's training crawler. Anthropic runs two others: Claude-SearchBot for Claude's search results, and Claude-User for pages people ask Claude to open.
- Google-ExtendedA robots.txt name, not a crawler, that controls whether Google may use your pages to train Gemini and ground its answers. It has no effect on Google Search or AI Overviews.
- GPTBotOpenAI's training crawler. Blocking it keeps your pages out of OpenAI's model training and has no effect on whether ChatGPT search shows you.
- llms.txtA Markdown file at the root of a site that lists its most useful pages for AI tools. A proposal, now in version 2, used mostly for software documentation.
- OAI-SearchBotOpenAI's search crawler. It builds the index ChatGPT searches, so blocking it keeps a site out of ChatGPT's search answers.
- PerplexityBotPerplexity's search crawler. It builds the index behind Perplexity's answers and, according to Perplexity, is not used to train models.
- robots.txtA plain text file at the root of a site that tells crawlers which paths they may fetch. It is a request, not a lock.
Links
- Anchor textThe clickable words of a link. Google reads them to understand the page being linked to, so they should say what is there.
- BacklinkA link to your site from another site. Google uses links to find pages and as one signal of which pages are relevant.
- DisavowA Search Console tool for asking Google to ignore specific links to your site. Google says most sites will not need it.
- Link exchangeTrading links with another site: you link to them, they link to you. Google's spam policies list excessive link exchanges as link spam.
- nofollow, sponsored and ugcValues for a link's rel attribute that tell Google what kind of link it is. Since 2019 Google treats all three as hints rather than rules.
Search basics
- Canonical URLThe address Google treats as the main version of a page that can be reached at several. You can suggest one; Google makes the choice.
- E-E-A-TExperience, expertise, authoritativeness and trustworthiness: the qualities Google's quality raters look for in content. Google says trust matters most and that E-E-A-T is not itself a ranking factor.
- Scaled content abuseGoogle's spam policy against publishing many pages mainly to manipulate rankings rather than to help people. It applies however the pages were made, with AI or without.
- Structured dataCode that describes a page's content in the schema.org vocabulary, so search engines can read facts such as the author, price or rating directly. Google recommends the JSON-LD format.