AI and SEO terms, defined in plain language.
A
AI Citation
A reference to your website or content that appears within an AI system’s response. AI citations are how assistants like ChatGPT, Gemini, and Perplexity credit sources for information they present to users. Unlike traditional links, citations don’t drive traffic directly — they signal to users that your content is trustworthy enough for an AI to reference. Citation patterns are emerging as a key visibility metric in AI search. See how AI citation systems work and learn about AI citation optimization.
AI Crawler
Automated bots operated by AI companies to discover, retrieve, and index web content for training AI models and populating AI assistant responses. Major AI crawlers include GPTBot (OpenAI), ClaudeBot (Anthropic), PerplexityBot (Perplexity), Google-Extended (Google), and OAI-SearchBot. These crawlers respect robots.txt directives — you can block them if desired, though blocking may reduce your visibility in their AI systems. Understanding which crawlers access your site and configuring robots.txt accordingly is part of AI SEO strategy. Learn more in our AI crawlers guide.
AI Search
Search systems powered by large language models (LLMs) that generate answers or summaries instead of returning a ranked list of links. Examples include ChatGPT Search, Google AI Overviews (in Google Search), Perplexity, Microsoft Copilot, and Claude search. AI search prioritizes relevance and comprehensiveness over traditional ranking signals, requiring a different optimization approach than classic SEO. Zero-click search is more common in AI systems — users get their answer without visiting your site. See AI Search Explained and how AI search differs from Google Search.
AI SEO
The practice of optimizing content so it gets found, understood, and cited by AI-powered search systems — Google AI Overviews, ChatGPT search, Perplexity, Gemini, and Copilot. AI SEO builds on classic SEO foundations but adds new priorities: entity clarity, retrievability (how easily AI systems can extract answers from your content), citation-worthiness, and compatibility with LLM reasoning. Unlike traditional SEO, which focuses on ranking pages, AI SEO emphasizes content authority and factual accuracy. Start with the AI SEO cluster.
AI Visibility Tracking
The practice of measuring how often, and how favorably, a brand or website appears in AI-generated answers — citations in ChatGPT, mentions in Google AI Overviews, and appearances in Perplexity or Copilot responses. Unlike classic rank tracking, which watches a position in a list of blue links, AI visibility tracking watches whether you’re cited at all, and how you’re framed when you are. It’s an emerging discipline: most tools repurpose citation scraping and manual prompt testing rather than offering official APIs. See AI search analytics and how to track AI Overviews.
Answer Engine Optimization (AEO)
A subset of AI SEO focused on optimizing content to be the answer that answer engines (AI assistants) deliver. AEO prioritizes direct answers, clear structure, authority signals, and citation-worthiness. Unlike traditional SEO which rewards keyword optimization, and GEO which rewards topical depth, AEO rewards content that directly answers the user’s question in a way an AI can extract and present. The distinction matters: an AEO-optimized article might lead with a clear, bold definition or answer before diving into supporting detail. See the Answer Engine Optimization guide.
C
Canonical Tag
An HTML tag (rel="canonical") that tells search engines which version of a page is the “master” copy when duplicate or near-duplicate content exists at multiple URLs. Canonical tags consolidate ranking signals onto one URL instead of splitting them across near-identical pages — common with URL parameters, print versions, or syndicated content. Getting canonicals wrong (pointing them at the wrong page, or omitting them) is one of the most frequent causes of duplicate-content and indexing problems in Search Console. See canonical tags explained and fixing “duplicate without user-selected canonical”.
Content Cluster
A group of related articles built around a central pillar page, interlinked so search engines and readers can move between the overview and its supporting depth. Content clusters are the practical mechanism behind topical authority — instead of one page trying to rank for everything, a pillar covers a topic broadly while cluster articles cover each sub-question in depth, all linking back to the pillar. Well-built clusters distribute link equity, reduce keyword cannibalization, and make a site easier for AI systems to parse as a coherent knowledge base. See content clusters and topical authority.
Content Decay
The gradual loss of organic traffic, rankings, or relevance a page experiences over time as competitors publish fresher content, facts go stale, or search intent shifts. Content decay is normal — not every page is meant to rank forever — but left unmanaged it quietly erodes a site’s total organic traffic. Catching it early means monitoring for declining pages in Search Console or GA4 and refreshing (rather than abandoning) content that once performed. See finding declining content with GSC and content refresh strategy.
Core Web Vitals
A set of three Google metrics measuring real-world page experience: Largest Contentful Paint (loading speed), Interaction to Next Paint (responsiveness), and Cumulative Layout Shift (visual stability). Core Web Vitals are a confirmed Google ranking factor, and — because slow or janky pages get abandoned before an AI crawler or reader finishes loading them — they matter for AI visibility too. Sites are scored per-URL and per-device, meaning a fast desktop page can still fail on mobile. See Core Web Vitals explained and designing a Core Web Vitals experiment.
Crawl Budget
The number of pages a search engine’s crawler is willing and able to fetch from a site within a given time period. Crawl budget matters most on large sites — a small blog rarely runs out, but a site with tens of thousands of URLs can have important pages missed if budget is wasted on redirects, duplicate parameters, or low-value pages. Optimizing crawl budget means pruning what shouldn’t be crawled so what matters gets found and indexed faster. See crawl budget optimization and the crawlability guide.
D
Duplicate Content
Substantially identical or near-identical content that appears at more than one URL, either within the same site or across different domains. Search engines don’t penalize duplicate content by default, but they do have to pick one version to rank and may split signals — or waste crawl budget — across the copies. Common causes include URL parameters, HTTP/HTTPS or www/non-www variants, staging environments left indexed, and syndicated articles. See duplicate content (technical SEO) and fixing duplicate content in WordPress.
E
Engagement Rate
A GA4 metric measuring the percentage of sessions that lasted longer than 10 seconds, included a conversion event, or had two or more page/screen views. GA4 replaced Universal Analytics’ bounce rate with engagement rate (its inverse-ish counterpart), shifting the framing from “did they leave immediately” to “did they actually engage.” For SEO, engagement rate by landing page is a useful, if imperfect, proxy for whether organic traffic is finding what it came for. See GA4 engagement rate explained.
Entity (SEO)
A distinct, uniquely identifiable thing — a person, place, organization, product, or concept — that search engines and AI systems can recognize and connect to other entities, independent of the exact words used to describe it. Entity SEO means making sure a brand, product, or topic is clearly and consistently identified (through structured data, consistent naming, and authoritative mentions) so search and AI systems know unambiguously who and what you are. It’s the foundation underneath knowledge graphs and much of modern AI search. See entity SEO and how knowledge graphs work.
F
Featured Snippet
A highlighted answer box shown above traditional results on a Google search page, pulled directly from a page’s content and typically formatted as a paragraph, list, or table. Featured snippets long predate AI Overviews and remain a distinct surface — some queries show both, some show only one, and Google increasingly draws AI Overviews from many sources at once rather than a single snippet-worthy page. Whether snippets still matter as much as they used to is a live debate. See AI Overviews vs featured snippets and do featured snippets still matter?.
G
Generative Engine Optimization (GEO)
A narrower branch of AI SEO focused specifically on earning citations inside AI-generated answers and overviews. GEO tactics include entity optimization (being recognized as a trusted authority for your topic), retrieval optimization (making content easy for AI systems to extract from), and semantic clarity (phrasing content in ways LLMs understand and prioritize). GEO differs from traditional SEO by valuing comprehensiveness and topic depth over keyword frequency. See the GEO guide and GEO/AEO cluster.
Google Analytics 4 (GA4)
Google’s event-based analytics platform, and the successor to Universal Analytics (which stopped processing data in 2023). GA4 tracks user interactions as events rather than sessions-and-pageviews, which changes how metrics like bounce rate (retired, replaced by engagement rate), conversions, and traffic sources are measured and reported. For SEO, GA4 is where organic-traffic performance, landing-page behavior, and — increasingly — AI-referral traffic get analyzed. See GA4 for SEO and tracking AI traffic in GA4.
Google Search Console (GSC)
Google’s free tool for monitoring how a site performs in Google Search — impressions, clicks, average position, indexing status, and crawl activity — reported directly from Google’s own systems rather than inferred from third-party tools. GSC is the primary place to diagnose why a page isn’t ranking or isn’t indexed at all, request re-crawls, and (increasingly) see a dedicated report for how content performs in AI-powered Search features. See the GSC Performance Report explained and the URL Inspection Tool walkthrough.
H
Hreflang
An HTML attribute that tells search engines which language and regional version of a page to serve to users in a given locale, used on sites with multiple language or country versions of the same content. Implemented incorrectly — missing return tags, wrong language codes, or conflicting canonicals — hreflang is one of the more error-prone parts of international technical SEO, and mistakes can cause the wrong version of a page to rank in the wrong country. See the hreflang guide.
I
Indexing
The process by which a search engine adds a crawled page to its searchable database, making it eligible to appear in results. Crawling and indexing are distinct steps — a page can be crawled but not indexed (often flagged in Search Console as “crawled, currently not indexed”), usually because the engine judged it low-value, duplicate, or thin. Indexing is the prerequisite for ranking: a page that isn’t indexed can’t show up in search results at all, no matter how well-optimized it is. See indexing best practices and the Page Indexing Report explained.
Internal Linking
The practice of linking between pages on the same site, used to help both users and search engines navigate a site’s structure and understand which pages matter most. Internal links pass authority between pages, establish topical relationships (a pillar linking to its cluster articles, for example), and give crawlers a path to discover deeper content that might otherwise sit orphaned. It’s one of the highest-leverage, lowest-cost technical SEO levers because it requires no new content — just better connections between what already exists. See the internal linking strategy guide and designing an internal linking experiment.
J
JSON-LD
A JavaScript-based format for embedding structured data (schema.org markup) in a webpage, written as a single script block rather than scattered through HTML attributes. JSON-LD is Google’s recommended structured-data format because it’s easy to generate, validate, and update independently of a page’s visible markup. Nearly all modern schema markup — Article, FAQPage, Organization, Product, and more — is implemented as JSON-LD rather than the older microdata or RDFa formats. See JSON-LD vs microdata vs RDFa and how to test and validate schema markup.
K
Knowledge Graph
A database of entities (people, places, organizations, concepts) and their relationships, used by search engines and AI systems to understand meaning beyond keywords. Knowledge graphs connect the dots — they know that “Apple” can mean a fruit or a company, and which interpretation is correct based on context. For SEO, this means optimizing for entity recognition: making it clear who you are (your brand), what you do (your services), and how you relate to broader topics. Appearing in knowledge graphs increases visibility to both traditional search and AI systems. See how knowledge graphs work and knowledge graph optimization.
L
llms.txt
A proposed standard file (placed at /llms.txt in a website’s root) that websites can use to provide explicit guidance to AI crawlers about which content should be included in AI training or responses. Similar to robots.txt, llms.txt would let you exclude certain pages, set crawl rules, and specify how AI systems should cite your content. The standard is still emerging and adoption is early, but some AI companies are exploring it as a way to respect publisher preferences. Whether to implement llms.txt is a strategic choice — blocking crawlers may reduce AI visibility but protects proprietary content. Read the llms.txt guide.
M
Multimodal Search
Search that accepts and reasons across more than one input type in a single query — text combined with an image, a photo combined with a spoken question, or video content analyzed alongside text. Tools like Google Lens and multimodal AI assistants let users search “with a picture and a sentence” rather than typing keywords, which changes what “ranking” even means: relevance now has to hold across modalities, not just match text. See multimodal search.
O
On-Page SEO
Optimization work done directly on a page’s content and HTML — title tags, headings, body copy, internal links, image alt text, and URL structure — as opposed to off-page factors like backlinks or technical/crawling factors at the site level. On-page SEO is the most controllable layer of SEO because it doesn’t depend on other sites linking to you or on infrastructure changes; it’s also foundational for AI search, since clear headings and direct answers are exactly what AI systems extract and cite. See the on-page SEO checklist.
Organic Traffic
Visits to a website that arrive from unpaid search engine results, as distinct from paid search, direct visits, referral links, or social traffic. Organic traffic is the metric most SEO work ultimately aims to grow, and analyzing it by landing page, query, and device reveals which content is actually working versus which merely ranks. As AI-generated answers increasingly satisfy queries without a click, some organic-traffic strategies now also track AI-referral traffic as a related but distinct channel. See analyzing organic traffic in GA4.
P
Permalink Structure
The format WordPress uses to build a page or post’s URL — for example, a clean /category/post-name/ structure versus a default query-string format like ?p=123. Permalink structure affects both readability (clean URLs are easier for users to parse) and technical SEO (changing an established structure without redirects breaks every existing link to the site). It’s a one-time decision worth getting right early, since revisiting it later means a full redirect mapping project. See WordPress permalink structure for SEO.
Programmatic SEO
An approach to content creation that generates large numbers of similar, template-based pages from a structured data source — pricing pages by city, product comparisons, or tool-specific landing pages — rather than writing each page individually. Done well, programmatic SEO captures long-tail search demand at scale; done poorly, it produces thin, near-duplicate pages that search engines treat as low-quality or even spammy. The line between the two comes down to whether each page offers genuinely unique value. See programmatic SEO.
Prompt Engineering
The practice of writing and refining instructions given to an AI system to reliably produce a desired output. In an SEO context, prompt engineering covers both using AI tools effectively for research and content work (keyword clustering, briefs, technical audits) and understanding how the prompts users type into AI assistants shape which content gets surfaced and cited in response. See prompt engineering for SEO and the 100+ SEO prompts library.
R
Rank Tracking
The practice of monitoring a site’s search engine ranking position for a defined set of keywords over time, typically using dedicated software rather than manual searching (which is skewed by personalization and location). Rank tracking tells you whether a page is moving up or down for a target query, but it’s an increasingly incomplete picture on its own — it doesn’t capture whether a page is being cited in an AI Overview or an assistant’s answer, which is what AI visibility tracking is built to measure. See the Ahrefs guide and Semrush guide, two of the most widely used rank-tracking platforms.
Retrieval-Augmented Generation (RAG)
A technique where AI systems retrieve relevant documents or passages from the web (or a database) before generating a response, rather than relying solely on information in their training data. RAG allows AI assistants to provide current, sourced, and cite-able answers. From an SEO perspective, RAG systems need to find and rank your content highly in order to retrieve and cite it — making relevance, content organization, and topical depth critical. Most modern AI assistants use retrieval-based approaches, making traditional ranking-like signals (authority, clarity, comprehensiveness) as important to AI search as they are to Google.
S
Schema Markup
Structured data added to a page’s code, using the shared vocabulary defined at schema.org, that explicitly labels what a piece of content is — a recipe, a product, an FAQ, an organization — rather than leaving search engines to infer it from surrounding text. Schema markup is what powers rich results (star ratings, FAQ dropdowns, breadcrumbs) in Google Search, and it’s increasingly important for AI search too, since clearly labeled entities are easier for an LLM to extract and cite correctly. See the schema markup complete guide and schema for AI search.
Semantic SEO
An approach to content and site structure that optimizes for meaning and topical relationships rather than exact-match keywords. Semantic SEO leans on the fact that modern search and AI systems use vector embeddings and entity recognition to understand that related concepts, synonyms, and questions all belong to the same topic — so a page can rank for (and be cited for) queries it never uses the literal words for. See semantic SEO for AI and vector search.
Structured Data
A standardized way of organizing information on a page so machines — search engines, AI crawlers, voice assistants — can read it unambiguously, rather than inferring meaning from unstructured prose. Schema.org markup (usually written as JSON-LD) is the most common form of structured data on the web, but the term also covers things like well-formed tables, consistent heading hierarchies, and semantic HTML. Structured data is foundational to both rich results in classic search and citation accuracy in AI search. See the structured data guide.
T
Topical Authority
A measure of how completely and deeply a website covers a subject, used by search engines to judge trustworthiness and relevance. Rather than optimizing individual pages for individual keywords, topical authority means building a cluster of interlinked, comprehensive content around a central topic — so your site becomes a recognized authority for that topic as a whole. Both traditional search and AI systems reward topical authority because it signals expertise. See the full topical authority breakdown and learn how to build it with content clusters.
V
Vector Search
A search method based on semantic similarity rather than keyword matching. Instead of looking for exact words, vector search converts content and queries into mathematical vectors (embeddings) and finds the most semantically similar results. This allows search to understand meaning — so “tall building” and “skyscraper” are recognized as similar even though they share no keywords. AI systems and modern retrieval methods use vector search heavily, which means keyword-stuffed content is less valuable than semantically coherent, well-organized content. For SEO in the AI era, this means writing for clarity and topical cohesion rather than optimizing for keyword density.
Voice Search
Search conducted by speaking a query aloud to a device — a smart speaker, phone assistant, or in-car system — rather than typing it, typically answered with a single spoken result rather than a list of links. Voice search queries tend to be longer and more conversational than typed queries (“what’s the best pizza place near me that’s open now” versus “pizza near me”), which pushed early SEO toward natural-language and question-based content long before AI search made that mainstream. See voice search evolution.
W
WordPress SEO
The set of SEO practices, plugins, and settings specific to sites built on WordPress — permalink structure, sitemap generation, schema output, caching, and image optimization — as distinct from general SEO principles that apply to any platform. Because WordPress powers a large share of the web, plugin choices (Rank Math, Yoast, and caching/sitemap tools) meaningfully shape a site’s technical SEO baseline before any content work even begins. See the Rank Math setup guide and WordPress SEO without plugins.
Z
Zero-Click Search
A search where the user gets a complete answer directly on the results page (or in an AI assistant’s response) and never clicks through to any website. Featured snippets, knowledge panels, and especially AI Overviews and chat-based answers have pushed zero-click rates higher over time, which means traffic is no longer a reliable proxy for a page’s SEO performance — a page can be the source behind a satisfying answer and receive zero visits. See zero-click AI search and zero-click search beyond AI Overviews.
Every term above links to a dedicated reference page with schema markup, and each definition points to the in-depth guide behind it. Put the vocabulary to work with the free tools and resources, or go deeper in the blog.