Glossary

An AI system that takes autonomous actions on behalf of a user — searching, retrieving, synthesising, and acting — rather than simply answering a single prompt. In the context of clinics, agentic AI includes Google AI Mode (which can now book appointments autonomously) and AI agents that research providers before recommending them to patients.

A composite score out of 100, developed by Iris by AdChoreo, measuring how well a clinic's digital infrastructure enables AI platforms to find, verify, and recommend the practice. Calculated across six dimensions: GBP completeness, on-page AI readability, schema markup coverage, NAP citation consistency, root-level AI files, and active AI surfacing. The average across 1,000+ independent clinics in India: 47 out of 100.

The practice of structuring digital content so that AI answer engines — ChatGPT, Perplexity, Google AI Overviews, Claude — can extract, trust, and cite a practice as the direct answer to a user query. AEO is distinct from SEO: SEO optimises for ranking position in a list of links; AEO optimises for being named in a synthesised AI response.

A root-level JSON file that provides AI retrieval systems with a machine-readable entity profile for a clinic: name, specialty, services, practitioners, location, hours, and usage permissions. Works alongside schema markup and llms.txt to establish a reliable entity signal that AI systems cite accurately. Deployed by Iris as a standard deliverable in every engagement.

Google's conversational AI search interface, unified with AI Overviews at Google I/O 2026. AI Mode surpassed 1 billion monthly users by mid-2026. It produces synthesised responses rather than ranked link lists, draws on GBP data for local recommendations, and can autonomously book beauty and wellness appointments. Medical clinic booking is the next expansion category.

The AI-generated summary that appears at the top of Google search results for most health and wellness queries. AI Overviews draw from multiple indexed sources, cite each one, and are trusted by 37% of patients as the most authoritative section of Google search results (rater8, 2026). Being cited in an AI Overview is currently the highest-value visibility outcome in healthcare local search.

The accumulated tendency of an AI system to include a specific clinic in responses for relevant local queries, built through repeated accurate citations over time. Citation confidence is a compounding asset: early citation leads to more citation, which leads to stronger recommendation patterns. Displacing an incumbent with high citation confidence requires both correct infrastructure and time.

The limit on how many pages on a website an AI crawler or search engine will index in a given crawl cycle. Poorly configured robots.txt files that permit AI crawlers to visit non-indexable booking platform pages (Zenoti, Vagaro, Boulevard) waste crawl budget that could have been spent indexing actual service pages. Iris manages robots.txt to direct crawl budget to the highest-value pages.

Google's framework for evaluating the quality of health and wellness content. The signals include: clinical credentials and fellowship training (Expertise), documented patient outcomes and case studies (Experience), consistent citation across authoritative sources (Authoritativeness), and accurate, verifiable entity data (Trust). E-E-A-T is the strongest predictor of Google AI Overview citation for medical content.

The structured facts that define a business as a specific, verifiable entity: name, address, phone number, specialty, services, practitioners, hours. AI systems cross-reference entity data from multiple sources (your website, GBP, Healthgrades, Yelp, directories) to construct a unified picture of your clinic. When those sources disagree — because of an old address, a rebrand, or a discontinued service — AI citations become inaccurate.

A schema.org structured data type that marks up question-and-answer content in a machine-readable format. FAQPage schema makes patient FAQ sections directly extractable by AI systems, enabling those answers to appear in Google AI Overviews and ChatGPT responses for pre-consultation research queries. One of the most reliable AI citation surfaces available to a clinic website.

The primary local data asset for any clinic — the profile that powers Google Maps results and feeds AI Mode's local recommendations. A clinic's GBP primary category is one of the highest-impact configuration decisions for local AI visibility. Iris manages GBP on an ongoing basis: weekly posts, Q&A, review monitoring, photo updates, and category audits.

The discipline of structuring a clinic's digital presence so that AI generative engines — ChatGPT, Gemini, Perplexity, Google AI — can find, understand, and cite the clinic. GEO deliverables include schema markup, llms.txt, robots.txt AI crawler configuration, ai-index.json, and content written in the inverted pyramid format for extraction. SEO and GEO are complementary, not competing.

A writing structure where the most important information appears first, followed by supporting detail. AI systems extract direct answer sentences — they look for content that opens with the key fact rather than building toward it. A procedure page that begins 'Melasma is a common pigmentation condition caused by sun exposure, hormonal changes, and genetic predisposition' will be cited by AI. A page that begins 'At [Clinic Name], we understand skin pigmentation concerns...' will not.

A plain-text file at yourwebsite.com/llms.txt that provides AI crawlers with a curated, prioritised map of a clinic's most important content — analogous to sitemap.xml for search engines but written specifically for AI comprehension. Proposed at llmstxt.org in September 2024. Perplexity and Anthropic's Claude have confirmed their crawlers reference llms.txt files. Iris deploys llms.txt as a standard deliverable in every engagement.

The set of three local business listings (with stars, reviews, and distance) that appear at the top of Google search results for 'near me' and location-specific queries. Local Pack ranking is determined primarily by GBP data — completeness, category accuracy, review volume, and posting frequency. Appearing in the Local Pack is the direct path from a patient's search query to a call or visit.

A schema.org structured data type that formally identifies a website as a medical practice entity — specifying the practice type (Dermatologist, Dentist, Ophthalmologist, etc.), location, services, and practitioners. MedicalBusiness schema is critical for AI systems to correctly classify and recommend a practice for specialty-specific queries. Absent in over 80% of independent clinic websites audited by Iris.

A schema.org structured data type that formally declares each treatment or procedure a clinic offers, including procedureType, bodyLocation, and indication. Each procedure page on a clinic website should have its own MedicalProcedure schema block. This enables AI systems to match the clinic to specific procedure-level patient queries ('best LASIK surgeon near me', 'hair PRP treatment in Bengaluru') rather than only to generic specialty queries.

The requirement that a clinic's name, address, and phone number are identical across all online directories — Google, Yelp, Healthgrades, Practo, Justdial, and any local directories. NAP inconsistency (caused by address changes, rebrands, or auto-populated directory errors) is the most common cause of inaccurate AI citations. It is also one of the most consistently fixable problems in a clinic's digital infrastructure.

A citation-first AI search engine with over 100 million monthly active users as of mid-2026, growing faster than any other AI search platform. Perplexity launched dedicated healthcare search capabilities in July 2026. It crawls the open web using PerplexityBot — independently of Google's index. A clinic can appear in Google AI results and be invisible on Perplexity if PerplexityBot is blocked or if the website has no llms.txt.

A schema.org structured data type that formally identifies a practitioner — including fellowship training, board certifications, medical school, and subspecialty credentials — in a machine-readable format. When a patient asks an AI chatbot for the best board-certified LASIK surgeon or the most experienced IVF specialist in their city, Person schema is what enables the AI to weight physician credentials as a quality signal rather than guessing from unstructured biography text.

The technical mechanism used by ChatGPT, Perplexity, and other AI search tools to answer questions by first retrieving relevant content from indexed web sources, then generating a synthesised response citing those sources. RAG is why the structured content on your website — schema, llms.txt, AI-readable text — directly determines whether your clinic appears in AI-generated patient recommendations.

A plain-text file at yourwebsite.com/robots.txt that tells web crawlers which parts of the site they can and cannot access. Most clinic robots.txt files were configured before AI-specific crawlers existed — they may permit Googlebot and Bingbot while silently excluding ChatGPT-User, PerplexityBot, ClaudeBot, and Google-Extended. When AI crawlers are excluded, no amount of schema or content quality compensates. Iris audits and corrects robots.txt as the first step of every engagement.

Machine-readable structured data added to a website — typically in JSON-LD format within a script tag — that formally describes the entities on the page: the type of business, its location, its services, its practitioners, and its reviews. Schema is the primary mechanism through which AI systems formally classify and cite a clinic. Absent or invalid in over 80% of independent clinic websites. Iris deploys a full @graph JSON-LD schema stack including MedicalBusiness, MedicalProcedure, FAQPage, Speakable, Person, and BreadcrumbList types.

A schema.org property that marks specific HTML sections by CSS class name and signals to voice assistants (Siri, Google Assistant, Alexa) which content should be read aloud in response to voice search queries. Without Speakable schema, voice assistants default to reading only a clinic's name and address — providing no spoken description of services or reason for the patient to choose that clinic. Fewer than 2% of independent clinics have implemented Speakable schema.

An umbrella term for any standardised format for providing explicit information about a webpage's content to AI systems and search engines. Schema markup (JSON-LD) is the most common form of structured data used in healthcare AI visibility. Structured data tells AI systems what a page is about rather than requiring them to infer it from unstructured text.

Search queries initiated by speaking to a voice assistant — Siri, Google Assistant, or Alexa — rather than typing. 27% of mobile searches in India are voice-initiated. Voice search returns a single spoken result, not a list of links, making it a winner-take-most visibility surface. The infrastructure required for voice search visibility includes Speakable schema, LocalBusiness schema with geo-coordinates, and a correctly configured GBP.