Search by acronym, full term, alternate wording or topic. Every entry includes a definition and an instruction for using the concept responsibly.
Definition: A software system that can plan or perform actions toward a goal, often by calling tools, browsing pages or interacting with services.
Practical instruction: Use accessible labels, predictable controls and clear confirmation steps when an agent may interact with a website.
Definition: A Google Search experience designed for longer, exploratory and comparative questions using generative AI and supporting web links.
Practical instruction: Maintain the normal SEO foundation; Google does not require separate AI Mode markup.
Definition: A broad, non-standard label for work intended to improve content, data or business information for AI-powered systems.
Practical instruction: Define the scope every time the term is used because AIO can also mean AI Overviews.
Definition: Google-generated summaries that may appear in Search with supporting web links for questions where an overview may help.
Practical instruction: Do not create a separate hidden AI page. Keep important information crawlable, textual and accurately structured.
Definition: A search experience that uses AI to interpret a request, retrieve information and often generate a direct or synthesized response.
Practical instruction: Treat AI search as a group of different products rather than one universal engine.
Definition: A search or information system designed to provide a direct response rather than only a list of possible sources.
Practical instruction: Write the essential answer first, then add the evidence, conditions and context needed to use it correctly.
Definition: The practice of presenting accurate information in a direct, structured form that can satisfy a specific question.
Practical instruction: Use descriptive question headings, answer immediately and keep the answer visible in ordinary HTML.
Definition: The act of identifying the source, author or organization responsible for information, research, media or a claim.
Practical instruction: Link to original or authoritative sources and identify who produced original work.
Definition: The preferred URL selected for a page when the same or very similar content can be reached through more than one address.
Practical instruction: Use a self-referencing canonical on unique pages and avoid publishing copies across several domains.
Definition: A software interface that accepts conversational input and returns generated or scripted responses.
Practical instruction: Do not assume every chatbot searches the live web; retrieval access varies by product and session.
Definition: The process of dividing a document into smaller passages that can be indexed and retrieved independently.
Practical instruction: Use clear sections and headings so passages retain meaning when retrieved away from the full page.
Definition: A reference identifying a source used to support or expand an answer.
Practical instruction: A citation is not the same as an endorsement, ranking or guarantee that the source will be reused later.
Definition: The amount of text, data or tokens a language model can consider within a single interaction.
Practical instruction: Put essential definitions close to the relevant heading instead of relying on distant context.
Definition: The ability of a crawler to reach and request a page and its important resources.
Practical instruction: Return successful HTTP responses, use crawlable links and avoid login walls for public material.
Definition: Automated software that requests pages and resources so a service can discover, index, analyze or use them.
Practical instruction: Check robots.txt and server logs, but also confirm that the firewall, CDN and bot protection allow legitimate crawlers.
Definition: How current the available information is relative to the question being asked.
Practical instruction: Update changed services, hours, prices, locations and contact details rather than leaving contradictory versions online.
Definition: A schema.org type used to describe a word, phrase, acronym or concept and its meaning.
Practical instruction: Use it to describe visible definitions, not to insert hidden meanings that the page never explains.
Definition: The public introduction connecting a business identity with its services, locations, contact details and official links before a customer makes contact.
Practical instruction: Make the identity, offer, service area and next step clear on the same public page.
Definition: A quality-evaluation concept referring to experience, expertise, authoritativeness and trust, with trust as the central consideration.
Practical instruction: Show who created the content, why they know the subject and what evidence supports important claims.
Definition: A numerical representation used to capture relationships or similarity among words, passages, images or other data.
Practical instruction: Remember that semantic similarity does not prove factual accuracy or business identity.
Definition: A distinct person, organization, place, product, event or concept that a system can identify and connect with information.
Practical instruction: Use consistent names and official links so the same entity is not mistaken for several unrelated ones.
Definition: The process of deciding whether different records or references describe the same real-world entity.
Practical instruction: Correct conflicting names, addresses, phone numbers and websites where the business has control.
Definition: AI designed to create new text, images, audio, code or other outputs from learned patterns and supplied context.
Practical instruction: Review generated material for accuracy, originality, relevance and unsupported claims before publication.
Definition: Work intended to improve the usefulness or visibility of content and entities in generated answers.
Practical instruction: Add original value, clear authorship, direct definitions, supporting evidence and a strong technical foundation.
Definition: The use of retrieved or supplied evidence to anchor a generated response in external information.
Practical instruction: Grounding can reduce unsupported answers, but the source and retrieval process can still be wrong.
Definition: A generated statement that is false, unsupported or presented with more confidence than the available evidence justifies.
Practical instruction: Provide verifiable facts and monitor important brand information, but do not claim hallucinations can be eliminated by publishing one page.
Definition: A processed collection of pages, records or passages that a search or retrieval system can query.
Practical instruction: A crawler reaching a page does not guarantee that the page will be indexed.
Definition: A protocol that lets participating search engines know when a URL has been added, updated or deleted.
Practical instruction: Submit changed URLs after deployment, but understand that receipt does not guarantee indexing.
Definition: The process of locating relevant documents, records or passages in response to a query.
Practical instruction: Write sections that remain understandable when retrieved as individual passages.
Definition: A JSON-based format commonly used to add linked structured data to a web page.
Practical instruction: Keep JSON-LD valid, page-specific and consistent with what users can see.
Definition: A structured network of entities and the relationships among them.
Practical instruction: Connect official names, people, organizations, places and products consistently rather than repeating disconnected keywords.
Definition: An information panel generated by Google from its understanding of an entity and relevant sources.
Practical instruction: A business cannot create a knowledge panel directly; it can maintain accurate official information and relevant profiles.
Definition: A model trained on large amounts of language data to predict and generate text and perform language-related tasks.
Practical instruction: Do not treat an LLM’s internal knowledge as a live, complete or always accurate database.
Definition: A proposed plain-text website guide intended to summarize important pages for systems that choose to read it.
Practical instruction: Use it only as a supplement. Google states that special AI text files are not required for AI Overviews or AI Mode.
Definition: Work intended to improve how a business appears for location-related searches and local customer needs.
Practical instruction: Maintain the Google Business Profile, accurate business details, relevant location content and real customer information.
Definition: A protocol for connecting AI applications with external tools, data sources and services through standardized interfaces.
Practical instruction: MCP is an application integration concept, not a replacement for public website crawlability.
Definition: A local-search abbreviation for a business’s name, address and phone number.
Practical instruction: Use the current official details consistently while accounting for service-area businesses that do not publicly receive customers at an address.
Definition: A field concerned with enabling computers to process, interpret and generate human language.
Practical instruction: Write naturally and clearly; awkward keyword repetition is not required for machines to understand the topic.
Definition: A directive asking compliant search engines not to include a page in their search index.
Practical instruction: Allow the crawler to access the page if it must read the noindex directive.
Definition: OpenAI’s crawler used to discover public content for ChatGPT search experiences.
Practical instruction: Allow it in robots.txt and ensure the server or WAF does not block legitimate requests when search visibility is desired.
Definition: Unpaid visibility in search results rather than placement purchased through advertising.
Practical instruction: Separate organic visibility reporting from paid campaigns and referral traffic.
Definition: The process of locating a specific section of a document that may answer a question.
Practical instruction: Use headings, complete sentences and local context so a retrieved passage does not become misleading.
Definition: Perplexity’s crawler for surfacing and linking websites in Perplexity search results.
Practical instruction: Allow the crawler and, where necessary, verify official IP ranges in WAF rules.
Definition: Input provided to an AI system describing a request, question, context or desired behaviour.
Practical instruction: A user’s wording can change retrieval and output, so test realistic customer questions rather than only one exact prompt.
Definition: A technique in which an AI search system issues multiple related searches to address different parts of a complex request.
Practical instruction: Build a coherent topic structure instead of creating a thin page for every imagined wording variation.
Definition: The ordering or scoring of documents, passages, products or results according to a system’s relevance and quality criteria.
Practical instruction: Do not confuse technical eligibility with a guarantee of high placement.
Definition: The first REG pillar, focused on helping systems identify the correct business or entity and connect its legitimate public references.
Practical instruction: Standardize the official name, people, location, services, contacts and official links.
Definition: The second REG pillar, focused on making accurate information about an entity available and retrievable for relevant questions.
Practical instruction: Publish direct service, area, credential, policy and contact information in visible text.
Definition: An AI architecture that retrieves external information and supplies it as context while a model generates a response.
Practical instruction: Evaluate source quality, retrieval quality and generation quality separately.
Definition: A search result enhanced with additional visual or interactive information supported by eligible structured data.
Practical instruction: Valid structured data creates eligibility, not a guaranteed enhanced appearance.
Definition: A file at the site root that communicates crawl permissions to robots that follow the Robots Exclusion Protocol.
Practical instruction: Use User-agent: * and Allow: / when the goal is to permit all compliant crawlers.
Definition: Machine-readable vocabulary used to describe people, organizations, articles, products, events, terms and other entities on a page.
Practical instruction: Choose page-appropriate types and do not add unsupported or invisible claims.
Definition: The practice of helping search engines discover and understand content while helping people find useful pages through search results.
Practical instruction: Start with crawlability, useful content, internal links, page structure and accurate business details.
Definition: Search that attempts to understand meaning, relationships and intent rather than relying only on exact word matching.
Practical instruction: Use natural language, clear entities and complete explanations instead of forcing every possible keyword variation.
Definition: The geographic area in which a business provides services to customers.
Practical instruction: Name real cities, regions or distance limits and distinguish the service area from a storefront address.
Definition: A machine-readable list of important site URLs and optional update information used to support discovery.
Practical instruction: Submit the sitemap to search tools and keep it aligned with canonical, indexable pages.
Definition: Machine-readable information that labels the meaning and relationships of page content using a defined vocabulary.
Practical instruction: Ensure every material statement in the structured data is supported by visible page content.
Definition: A unit of text processed by a language model; a word may contain one or several tokens.
Practical instruction: Token counts affect model context and cost, but website quality is not measured by token volume.
Definition: The ability of an AI system to invoke external functions, searches, databases or services while handling a request.
Practical instruction: Design public interactions with clear labels, predictable inputs and confirmation before consequential actions.
Definition: A database designed to store and search numerical vector representations, often used for semantic retrieval.
Practical instruction: A vector database improves similarity search; it does not resolve inaccurate source data automatically.
Definition: A search initiated through spoken language, often using a phone, car system, smart speaker or assistant.
Practical instruction: Answer natural customer questions and keep local contact and location information current.
Definition: A search session in which the user receives enough information on the results page or answer interface without clicking a website.
Practical instruction: Measure more than website clicks; visibility, branded demand and later customer actions may occur elsewhere.
The same definitions are available as JSON and CSV. Reuse should preserve attribution and should not imply that platform companies endorse original terms such as REG.