160 pages
AI Search and GEO Glossary
Definitions of the terms behind AI search visibility: GEO, AEO, grounding, retrieval, AI crawlers, citation rate, share of voice and more.
#
A
Agentic search: definition and how it works
Agentic search defined: how AI agents plan, search repeatedly and read results before answering, and why brands should watch the third-party pages agents read.
What Is an AI Agent? Definition and Web Access
AI agent definition: a model that takes multi-step actions, like browsing pages for you. See why robots.txt may not apply and how pages can mislead agents.
AI citation: definition and how it works
AI citation definition: what counts as a cited source in ChatGPT, Claude, Perplexity and Google AI, how citation units differ, and what they prove.
What Is an AI Crawler? Training, Search and User Bots
AI crawler definition: the three types (training, search indexing, user-triggered), which bots do what, and why most cannot read JavaScript content.
AI Referral Traffic: Definition and Measurement
AI referral traffic is visits from links in AI answers. Learn how it is measured, why tagging is uneven, and why reported shares are upper bounds.
What Is an AI Search Engine? Definition and How It Works
AI search engine definition: how ChatGPT, Perplexity and Google AI Overviews retrieve pages and write answers, and why some answers skip the web entirely.
AI SEO: Definition of a Term With Two Meanings
AI SEO has two meanings: using AI tools for SEO work, or making a site visible in AI answers. Get a clear definition and the right guide for your question.
AI Visibility: A Working Definition
AI visibility defined: whether and how AI systems mention, cite and describe your brand. Its four parts, how engines differ, and what you can control.
AI Visibility Score: Definition and What It Hides
AI visibility score defined: what the composite measures, why it mixes retrieval-on and retrieval-off answers, and which components to read before the total.
AI Visibility Tracking: Definition and How It Works
AI visibility tracking defined: repeated measurement of how AI engines name and cite your brand under fixed conditions, and how it differs from a snapshot.
Amazonbot: Definition, User Agent and IndexNow Role
Amazonbot is Amazon's web crawler. Learn how it identifies itself, how robots.txt and firewalls control it, and what its IndexNow listing does and does not mean.
Answer engine: definition and how it works
Answer engine defined: how it rewrites queries, retrieves pages, selects sources and writes cited answers, plus which steps a publisher can and cannot influence.
Answer Engine Optimization (AEO): What the Term Means
Answer engine optimization (AEO) explained: what the term means, how it overlaps with GEO, and what Google says its AI features actually require.
Answer variability: why AI gives different answers
Answer variability is when the same prompt gets different AI answers. Learn why it happens and how to measure brand visibility fairly across repeated runs.
Applebot: Definition, User Agent and Rendering
Applebot is Apple's web crawler. Learn how it renders JavaScript with a browser-based crawler, why that matters for AI visibility, and what it does not fix.
What Is an Atomic Fact? Definition for AI Accuracy
Atomic fact: a single checkable statement pulled from an AI answer. See how fact decomposition makes brand accuracy measurable, one claim at a time.
Average position: definition and why AI has none
Average position defined: how classic search reports it, why Google exposes no position data for AI answers, and what to track in AI search instead.
B
Backlinks: Definition and AI Search Relevance
Backlinks defined: links from other sites to yours. See how they relate to AI answers, why brand size confounds the data, and what evidence actually supports.
Bingbot: Definition and Why It Matters for AI Answers
Bingbot is the crawler that builds Bing's index. Learn how it works, why Bing's index backs Copilot's AI answers, and what is and isn't documented beyond that.
Bot management: definition and AI crawler impact
Bot management classifies and challenges automated traffic. See how challenges, CAPTCHAs and crawler categories decide which AI crawlers reach your site.
Brand conflation: when AI mixes up companies
Brand conflation is when AI merges facts about two similarly named companies. Learn how it happens, why it matters, and what you can and can't control.
Brand facts: what they are and why consistency matters
Brand facts are the checkable details about your company that AI answers repeat. Learn what counts, how engines use them, and why contradictions create risk.
Brand Mention in AI Answers: Definition
Brand mention defined: what it means when an AI answer names your brand, how memory-only and retrieval-backed mentions differ, and what you can influence.
Brand prominence: definition and the confounder
Brand prominence explained: what it means in AI search, the 67.3% citation concentration finding, and why it confounds most correlational GEO studies.
Bytespider: What This Crawler Is and How to Handle It
Bytespider is a high-volume web crawler often grouped with AI bots. Learn what its user agent means, why it skips JavaScript, and how to decide on it per bot.
C
Canonical tag: definition and how it works
Canonical tag defined: how rel canonical names the preferred URL among duplicates, and why canonical conflicts can change which URL AI answers show.
ChatGPT-User: Definition of OpenAI's User Agent
ChatGPT-User is OpenAI's fetcher for user-initiated page requests. Learn what it does, why robots.txt may not apply, and how it differs from OAI-SearchBot.
Chunking: Definition and Who Actually Does It
Chunking is how AI retrieval systems split documents into segments inside their own index. Learn who controls it and why Google says you needn't chunk.
Citation rate: definition and how it's calculated
Citation rate is the share of sampled AI answers that link to your domain. See how it's calculated, when it's undefined, and why the denominator must be stated.
Citation share: definition and how to read it
Citation share is your citations as a fraction of all citations for a query set. Learn how it works, why it can fall as citations rise, and how to read it.
Claude-SearchBot: Definition and Role in Claude Search
Claude-SearchBot is Anthropic's crawler for Claude search quality. Learn what it does, how blocking it differs from ClaudeBot, and what to check first.
Claude-User: what this user agent does
Claude-User is Anthropic's agent for fetches a person triggers in Claude. Learn how it differs from ClaudeBot and what blocking it means for your pages.
ClaudeBot: definition and what blocking does
ClaudeBot is Anthropic's training crawler. Learn how it differs from Claude-SearchBot and Claude-User, what blocking it does, and why IP blocks backfire.
Click-through rate (CTR): definition and AI search limits
Click-through rate (CTR) is clicks divided by impressions. Learn how search CTR works and why it can't be computed for most AI answer surfaces today.
Client-side rendering (CSR): definition and risk
Client-side rendering (CSR) builds page content in the browser with JavaScript. Learn why most AI crawlers miss it and how to check your own pages.
Cloaking: Definition and AI Crawler Risk
Cloaking means showing crawlers different content than users see. Learn how it works, why cloaking for AI crawlers is a spam risk, and what isn't cloaking.
What Is a Consideration Set in AI Search?
Consideration set: the pages an AI engine retrieves as candidates before it writes an answer. Learn how it forms, why entry is hard, and what evidence says.
Content delivery network (CDN): definition
Content delivery network (CDN) defined: how a CDN serves and protects a site at the edge, and how its bot defaults can block AI crawlers without you editing a thing.
Content freshness: definition and evidence
Content freshness defined: what update dates mean for AI answers, what the evidence shows, and why bumping dates without real changes proves nothing.
Position bias in LLM answers: definition
Position bias: why LLMs favor documents placed earlier in their context, what C-SEO Bench and a 252,000-trial study found, and why publishers can't set it.
Context window: definition and why it matters
Context window: the text an AI model can read for one answer, including retrieved pages. Learn how it works and why most GEO studies skipped retrieval.
Conversational search: what it is
Conversational search is multi-turn querying where context carries across turns. See how it rewrites queries and why logged-in results are hard to reproduce.
Conversational SEO (C-SEO): definition and evidence
Conversational SEO (C-SEO) defined: rewriting content so conversational AI engines cite it, and why C-SEO Bench found most methods ineffective or harmful.
What Is Crawl Budget? Definition and AI Crawler Caveats
Crawl budget explained: how many URLs a crawler fetches from your site, when it matters (mostly very large sites), and why AI crawler access matters more.
Crawl payload limit: definition and the 2MB cap
Crawl payload limit explained: Googlebot reads up to 2MB of HTML per URL and ignores the rest. Learn how the cap works, who it affects and when to check.
Crawl-to-Referral Ratio: Definition and Caveats
Crawl-to-referral ratio explained: crawler requests per referred visit, why bot purpose drives the number, and why old published figures mislead you.
D
data-nosnippet: definition and how it works
data-nosnippet is an HTML attribute that keeps a section out of snippets and AI summaries while the page stays indexed. See how it works and where it misfires.
Deindexing: Definition and Its Effect on AI Answers
Deindexing explained: what it means when an engine drops your pages, why Google's AI features stop citing them, and what the Otterly test showed about Bing.
Disallow directive: definition and common mistakes
Disallow directive explained: what a robots.txt Disallow rule does, a correct per-bot snippet, and how a blanket AI block removes you from ChatGPT search.
Domain Authority: Definition and What It Isn't
Domain authority is a Moz metric built from link data. No search engine computes it. Learn what a DA score measures and why it can't predict AI citations.
E
E-E-A-T: Definition and Why There Is No Score
E-E-A-T defined: Google's rater-guideline concept of experience, expertise, authoritativeness and trust, why it isn't a ranking factor, and why no score exists.
What Are Embeddings? Definition for AI Search
Embeddings defined: numeric representations of text meaning that AI search uses to retrieve pages. Learn how they work and why meaning beats exact keywords.
What Is an Entity in Search? Definition for AI Search
Entity, defined: the distinct things (companies, people, products) search systems track, and why consistent facts across your own site matter for AI answers.
F
FActScore: Definition and How It Works
FActScore definition: the share of atomic facts in an AI answer that a knowledge source supports. How it is computed, its equal weighting, and its blind spots.
Factual accuracy in AI answers: definition
Factual accuracy is the share of an AI answer's claims that sources support. Learn how sound metrics count claims, and why summing penalties gives broken scores.
Faithfulness (RAG Metric): Definition and Formula
Faithfulness is supported claims divided by total claims, scored against retrieved context. Learn how it's measured and why faithful is not the same as correct.
FAQPage Schema: Definition and Status After Deprecation
FAQPage schema defined: what the markup does, why Google deprecated FAQ rich results in May 2026, and what the evidence says about AI citations.
Featured Snippet: Definition and the AI Link
Featured snippet explained: what the answer box is, how snippet controls also govern AI Overviews and AI Mode, and why snippet eligibility comes first.
Fine-tuning: what it is and what it isn't
Fine-tuning is extra training a model owner runs on a base model. Learn how it works, why brands can't fine-tune ChatGPT, and what you can control instead.
G
What Is a Generative Engine? Definition and How It Works
Generative engine defined: an LLM that composes answers from retrieved sources. How it works, why it matters for AI visibility, and what the GEO paper tested.
Generative Engine Optimization (GEO): Definition
Generative engine optimization (GEO) defined: where the term came from, what it tries to change, and what the evidence says actually works for AI visibility.
Google-Extended: Definition and What It Controls
Google-Extended is the robots.txt token for Gemini training and grounding. See what it controls, what it doesn't, and why it leaves AI Overviews untouched.
Googlebot: Definition and Its Role in AI Features
Googlebot is Google's search crawler. Learn how it renders JavaScript, why it is the only robots.txt control for AI Overviews, and which snippet tags matter.
GPTBot: definition and what it controls
GPTBot is OpenAI's training crawler. Learn what its user agent does, what blocking GPTBot changes, and why it does not control ChatGPT search visibility.
Grounding in AI: Definition and How It Works
Grounding in AI means tying an answer to sources retrieved at answer time. Learn how grounded and ungrounded answers differ and what controls grounding.
What Is a Grounding Query? Definition and Examples
Grounding query: the search an AI system writes and runs to fetch sources. Learn how it differs from the prompt and where Bing shows a sample of real ones.
H
AI hallucination: definition and causes
AI hallucination defined: when a model states something false or unsupported. Learn the four causes behind wrong brand facts and what you can check today.
Hallucination rate: definition and how it's measured
Hallucination rate is the share of AI statements or answers that are false or unsupported. Learn both definitions, the right denominator, and benchmark limits.
Hidden text: definition and why it's a risk
Hidden text is content users can't see but machines can read. Learn how it works, why Google treats it as spam, and how it carries prompt injection into AI answers.
Hybrid Search: Keyword Plus Vector Retrieval
Hybrid search combines BM25 keyword scoring with vector similarity in one retrieval step. Learn how it works and why exact facts belong in your raw HTML.
Hydration: Definition and the AI Crawler Gap
Hydration explained: how JavaScript attaches to server HTML, why content added after hydration can be invisible to AI crawlers, and how to run a hydration diff.
I
Impressions: definition for search and AI features
Impressions defined: what counts as a search impression, what AI Overview impressions in Search Console show, and what they cannot tell you about traffic.
Index eligibility: definition and how it works
Index eligibility means a page meets the conditions to be shown in search and AI answers. See what Google requires and why eligible does not mean shown.
Indexing: Definition and Why AI Answers Depend on It
Indexing defined: how search engines store fetched pages, why each engine keeps its own search index, and how index status decides which AI answers can cite you.
IndexNow: definition and how it works
IndexNow defined: how the URL submission protocol works, which engines consume it, what a 200 response means, and why it is not a direct AI search lever.
Inline Citation: Definition in AI Answers
Inline citation defined: how AI answers attach sources to claims, span-level quotes vs numbered links, and why citation counts from different engines don't add up.
IP allowlisting: definition and crawler caveats
IP allowlisting explained: what it is, how published crawler IP ranges like OpenAI's searchbot.json support it, and why IP-only rules need verification.
J
JavaScript rendering: definition for AI crawlers
JavaScript rendering defined: how crawlers execute JS to build the DOM, which ones render (Googlebot, Applebot), and why most AI crawlers miss JS-only content.
JSON-LD: Definition and How It Works
JSON-LD defined: the script format Google recommends for structured data, a short correct snippet, and what the evidence says about JSON-LD and AI answers.
K
Keyword stuffing: definition and why it backfires
Keyword stuffing defined: why Google treats it as spam, what the GEO study found about AI answers, and what to write instead for AI search visibility.
Knowledge cutoff: definition and why it matters
Knowledge cutoff explained: the date a model stopped seeing training data, why one AI app can mix cutoffs, and what that means for your brand's AI answers.
Knowledge graph: definition and how it works
Knowledge graph defined: a database of entities and their relationships. How it works, what it means for AI answers, and the myths to ignore about it.
L
What Is a Large Language Model (LLM)? Plain Definition
Large language model (LLM) explained for marketers: how next-token prediction works, memory vs live search, and why it changes what AI says about your brand.
lastmod: Definition and Why It Proves Little
lastmod is the self-declared last change date in an XML sitemap. Learn how it works, why it is easy to spoof, and how to use it as a date consistency check.
What Is LLM Optimization (LLMO)? A Plain Definition
LLM optimization (LLMO) explained: what it means, how a model's frozen training memory differs from live retrieval, and which levers have real evidence.
llms.txt: Definition of the Proposed AI Text File
llms.txt is a proposed Markdown file that lists a site's key content for AI models. See how it works, what Google says, and why it is optional.
Long-tail keywords: definition and AI search
Long-tail keywords defined: specific, lower-volume queries. See how AI engines rewrite them, what Google says about variations, and where to focus instead.
M
Manual action: definition and how it differs
Manual action: Google's human-applied penalty, reported in Search Console. See how it differs from silent algorithmic removal and what it means for AI search.
max-snippet: Definition, Syntax and AI Feature Effect
max-snippet caps how much of a page search engines show as a snippet. Learn the syntax, what max-snippet:0 does to AI Overviews, and common mistakes to avoid.
What Is Mention Rate? AI Brand Mention Rate Defined
Mention rate is the share of sampled AI answers that name your brand, linked or not. Learn how it differs from citation rate and why prompt choice drives it.
Meta-ExternalAgent: Definition and How It Works
Meta-ExternalAgent is Meta's AI training crawler. Learn what it fetches, why it skips JavaScript, and how it differs from Meta-WebIndexer in robots.txt.
Meta-WebIndexer: definition and how it works
Meta-WebIndexer is Meta's crawler for Meta AI search. Learn what it does, how robots.txt controls it, and what allowing it can and cannot do for citations.
Model routing: definition and why AI answers vary
Model routing explained: how AI apps send each request to one of several models, why cutoffs and search behavior differ, and what that means for your brand.
N
Navigational link: definition in ChatGPT search
Navigational link, defined: how ChatGPT can still link to a site opted out of OAI-SearchBot, and why that differs from being cited in a search answer.
noindex: definition and how it works
noindex is the directive that keeps a page out of search indexes and Google's AI features. See the meta tag, the header form, and how templates leak it by accident.
nosnippet: Definition and Effect on AI Answers
nosnippet is a robots directive that stops Google showing text snippets. Learn how it works, how it keeps pages out of AI Overviews, and when to use it.
O
OAI-SearchBot: Definition and How It Works
OAI-SearchBot is the crawler behind ChatGPT search. Learn how it works, OpenAI's two-part eligibility rule, and what blocking it in robots.txt really does.
Organic Ranking: Definition and AI Citations
Organic ranking defined: a page's position in unpaid search results. See what Ahrefs' March 2026 data says about top 10 rankings and AI Overview citations.
P
Parametric knowledge: what an AI knows without search
Parametric knowledge is what an AI model learned in training, frozen at its cutoff. See how it shapes search-off answers and what you can and can't change.
People-first content: definition from Google
People-first content is Google's term for pages made to help readers, not to game rankings. See the definition, Google's AI-content stance, and what it means.
Perplexity-User: Definition and robots.txt Behavior
Perplexity-User is Perplexity's fetcher for user-initiated requests. Learn how it differs from PerplexityBot and why robots.txt blocking is unreliable for it.
PerplexityBot: Definition and How It Works
PerplexityBot is Perplexity's search indexing crawler. Learn what it does, how robots.txt and firewalls affect it, and what allowing it does and does not get you.
Position-Adjusted Word Count (PAWC): definition
Position-adjusted word count (PAWC) defined: how the GEO paper's metric works, why it partly counts added words, and where the 40% uplift figure comes from.
What Is Prerendering? Definition for AI Search
Prerendering explained: how HTML snapshots are built ahead of requests, why AI crawlers that skip JavaScript need them, and how to avoid cloaking.
Prompt: definition in AI search
Prompt, defined for AI search: how a prompt differs from the queries an AI engine runs, why wording changes answers, and why trackers can only sample prompts.
Prompt injection: definition and brand risk
Prompt injection explained: how planted instructions steer AI search agents, what the research shows, and why third-party pages can misrepresent your brand.
What Is a Prompt Set? AI Visibility Tracking Defined
Prompt set: the sample of questions an AI visibility tracker asks. Learn how its size, wording and brand mix shape every metric, and what makes a set fair.
Prompt tracking: definition and how it works
Prompt tracking defined: how running fixed prompts on AI engines over time works, what it records, and why it measures a proxy for the queries engines run.
Q
Query fan-out: definition and how it works
Query fan-out explained: how AI engines split one prompt into several searches, why the real retrieval queries stay hidden, and what that means for tracking.
What Is Query Rewriting in AI Search?
Query rewriting explained: how ChatGPT, Google and Gemini turn a prompt into their own search queries, and why prompt tracking only measures a proxy.
R
Reranking: Definition and How It Works in AI Search
Reranking is a second scoring pass that reorders retrieved results before an AI model reads them. Learn how it works and what publishers can control.
Retrieval in AI search: definition and how it works
Retrieval in AI search is the answer-time step that picks which pages enter the model's context. See how it works and what publishers control upstream.
Retrieval-Off Answer vs Retrieval-On, Defined
Retrieval-off answer: what it means when an AI answers from training data alone, how it differs from search-on answers, and why you often can't tell which.
Retrieval-augmented generation (RAG): definition
Retrieval-augmented generation (RAG) explained: how retrieval and generation work as separate stages, and why most GEO studies only tested the second one.
Retrieved sources: definition and the cited subset
Retrieved sources are every page an AI engine consulted for an answer. Learn how they differ from citations and why counting only citations undercounts you.
Reverse DNS Verification for Crawlers: Definition
Reverse DNS verification confirms a crawler's IP maps to the operator's domain and back. Learn how the rDNS lookup works and where IP lists fit for AI bots.
Rich results: definition and how AI answers differ
Rich results explained: what they are, how structured data powers them, and why rich result eligibility is separate from inclusion in Google's AI answers.
Robots Exclusion Protocol (RFC 9309) Explained
Robots Exclusion Protocol explained: what RFC 9309 standardized in 2022, how group matching and Allow/Disallow work, and why compliance is voluntary.
Robots meta tag: definition and directives
Robots meta tag explained: what it is, which directives (noindex, nosnippet, max-snippet) affect AI answers, and how it differs from robots.txt crawling rules.
robots.txt: Definition and How It Works for AI Crawlers
robots.txt defined: the root-level file of per-bot crawl rules, how AI crawlers read it, and why an allowed bot can still be blocked or ignore the file.
S
Scaled content abuse: definition and Google's policy
Scaled content abuse is Google's spam policy on mass-produced pages made to manipulate rankings. See the definition, how removal works, and AI search impact.
Schema Markup: Definition and What the Evidence Shows
Schema markup defined: what schema.org markup is, how it works, and what Ahrefs' matched test of 1,885 pages found about schema and AI citations.
Search crawler: definition in AI search
Search crawler defined: the bots that build the index AI search retrieves from, how they differ from training bots, and why allowing them buys only eligibility.
Search engine optimization: a definition for AI search
Search engine optimization defined: how SEO makes pages crawlable, indexable and relevant, and why Google says the same rules apply to AI Overviews and AI Mode.
Search intent: definition and how AI answers vary
Search intent is the goal behind a query. Learn the four intent types, how AI engines handle them, and why brand visibility should be measured across intents.
Search Quality Rater Guidelines: Definition
Search Quality Rater Guidelines defined: what Google's manual for human raters covers, why ratings never rank a page, and what that means for AI visibility.
What Is Semantic Search? Meaning-Based Retrieval Explained
Semantic search explained: how meaning-based retrieval works, why Google says you needn't chase every long-tail variant, and what it means for AI visibility.
AI Answer Sentiment: Definition and Measurement
AI answer sentiment is the tone an AI answer takes toward your brand. Learn how it is scored, why it is a judgment, and why to read it alongside accuracy.
SERP: Definition in the Age of AI Features
SERP definition: what a search engine results page contains, how AI Overviews appear on it, why they vary by query and location, and what it means for you.
Server-side rendering (SSR): definition for AI search
Server-side rendering (SSR) defined: how server rendered HTML works, why it is the best-evidenced access fix for AI crawlers, and what it does not do for citations.
AI Share of Voice: Definition and Limits
AI share of voice defined: how brand mentions across a prompt set are counted against competitors, why prompts and brand size skew it, and how to read it.
Source Attribution in AI: Definition and How It Works
Source attribution in AI links generated claims to the sources behind them. Learn how it works in RAG, why citations differ from use, and what you control.
Static site generation (SSG): definition
Static site generation (SSG) builds HTML at deploy time. Learn how it works and why it puts content in the raw HTML that non-rendering AI crawlers read.
Structured data: definition and AI search role
Structured data explained: what the markup is, how engines read it, and what Google and Ahrefs' study show about its role in AI search answers as of 2026.
Supporting link: definition in Google AI features
Supporting link definition: Google's term for a source shown in AI Overviews or AI Mode, the exact eligibility rule, and what it does and doesn't guarantee.
What Is a System Prompt? Definition for AI Search
System prompt: the hidden instructions an app gives an AI model. Learn how they steer when assistants search, and why one model answers differently across apps.
T
Temperature in LLMs: Definition and Effect on AI Answers
Temperature in LLMs controls how random each answer is. Learn how sampling temperature works and why one AI answer about your brand is a sample, not a trend.
Thin Content: Definition and Risks for AI Search
Thin content means pages that add little value for users. Learn how it links to scaled content abuse, silent deindexing, and lost AI Overviews citations.
What Are Tokens in an LLM? Definition and Cost
Tokens are the units of text an LLM reads and writes. Learn how tokenization sets context limits and API cost, and what that means for AI answer tooling.
Tool use (function calling): definition
Tool use, or function calling, is how an LLM calls search or fetch. Learn how it works and why skipping the search call means an answer from memory alone.
Topic authority: definition, scope and limits
Topic authority means being a known go-to source on a subject. See how Google scopes it, why it is not a site score, and what it does not do for AI answers.
Topical relevance: definition and the evidence
Topical relevance is the fit between a page's substance and the query an AI engine runs. See what the research shows, why it is not keyword matching, and what to do.
Training crawler: definition and what blocking does
Training crawler defined: the AI bots that collect pages for model training, what blocking them changes, and why it does not remove you from AI search.
Training data: definition and the AI visibility link
Training data is the text an AI model learns from before release. See how GPTBot and other crawlers collect it, and what the evidence says about brand recall.
U
Unlinked Mention: Definition and AI Search Relevance
Unlinked mention: what it means when a site names your brand without a link, how AI engines use those mentions, and what the evidence does and doesn't show.
Unverifiable claim: definition in accuracy scoring
Unverifiable claim defined: an AI statement no source can confirm or contradict. Learn how it differs from a false claim and why scoring it is a judgment call.
User agent: definition and AI bot names
User agent defined: the name a client sends with each request, how robots.txt rules target AI bot names, and why spoofing means you verify with IP signals.
User-Triggered Fetcher: Definition and robots.txt
User-triggered fetcher defined: the AI agent that loads a page when a user asks, why robots.txt may not apply to it, and how it differs from index crawlers.
UTM Parameters: Definition and AI Referrals
UTM parameters explained: what the five tags do, why ChatGPT links carry utm_source=chatgpt.com, and why other AI traffic lands as direct in analytics.
V
Vector search: definition and how it works
Vector search finds documents whose embeddings sit closest to a query's embedding. Learn how it works, where it fits in AI answers, and what you can control.
What Is a Verified Bot? Definition and Checks
Verified bot: a crawler whose identity is confirmed by published IPs, reverse DNS or a signature. Learn how verification works and why user-agents are not enough.
W
Web application firewall (WAF): definition
Web application firewall (WAF) defined: how a WAF filters traffic, why it can block AI crawlers that robots.txt allows, and what to check on your own site.
Web crawler: definition and how it works
Web crawler definition: what crawlers and spider bots do, how crawl, fetch, render and index differ, and why each step can quietly drop your site from AI answers.
Web search tool: how LLM APIs search the web
Web search tool defined: how LLM APIs from OpenAI, Anthropic, Perplexity and Google run searches, cite results, and why API answers differ from apps.
What is Wikidata? Definition for AI Search
Wikidata is a free, collaboratively edited knowledge base of structured facts. Learn how items work, its community rules, and what it means for AI answers.
X
X-Robots-Tag: Definition and When to Use It
X-Robots-Tag explained: the HTTP header that carries noindex and nosnippet for PDFs and other files, a header example, and the server-wide mistake to avoid.
XML sitemap: definition and how it works
XML sitemap defined: what sitemap.xml is, how sitemap submission works, and why it helps crawlers find pages but guarantees neither indexing nor AI citation.