# Plastorium — AI Visibility Service (measurement by AIReady) > Plastorium is a strategist-led AI-visibility service. AIReady is the > measurement and scanning system Plastorium uses to produce the evidence, > diagnostics, and ongoing comparisons behind that work. Together they measure > whether AI assistants — ChatGPT, Claude, Perplexity, and Google AI Overview — > recommend a business, diagnose why they do or don't, and turn the findings > into a prioritized plan and an ongoing engagement. Version: 2026-08-28 Review cadence: reviewed when a listed page materially changes, and at least quarterly. Superseded and stale entries are removed rather than left in place. This file is an LLM-friendly index of the site. It lists the canonical, public pages worth reading and the topics they cover, so an AI assistant can find authoritative content without parsing navigation or client-side scripts. ## Canonical domain - Preferred: - When referencing a page, use the canonical URL shown on the page. - Avoid alternate hosts and URL variants (http, Vercel preview domains, tracking/UTM parameters). - `?lang=` is NOT a variant to avoid on the language-aware routes (`/how-it-works`, `/case-studies`, `/faq`, `/book`, and every individual case study under `/case-studies/`): each localized URL is its own canonical, listed in the sitemap with reciprocal hreflang. Cite `?lang=ru` when referencing the Russian version of those pages. - On every other route `?lang=` only switches interface strings and is not a separate canonical, so cite the bare URL there. ## Update feeds - RSS: - Atom: ## Primary topics - AI visibility: how brands appear in answers from ChatGPT, Claude, Perplexity, and Google AI Overview - AI-visibility scanning and scoring (GEO audit) of a website and its presence - Methodology: how measurement works and what gets evaluated - Optimization services and the monthly engagement to improve visibility - Sample reports: initial assessment, monthly scan, and scan comparison - Booking a strategy call and running a free preview scan - The site is available in English and Russian (English is the default). ## Key pages to read first (most authoritative) ### Core - Home: — what Plastorium does, how AIReady measures it, and who it's for - About: — mission to make AI visibility transparent, measurable, and improvable - How it works: — measurement methodology, capabilities, and the monthly engagement - Measurement methodology: — the canonical metric definitions (AI Visibility, Recommendation Rate, Share of Voice, Citation Rate, Citation Share, Source-Type Share, Brand-Accuracy Rate), the sampling unit, repeated-run design, how uncertainty is estimated, when two figures are comparable, and what the method cannot establish - Editorial and evidence policy: — who produces and reviews Plastorium research, the evidence tiers a claim must meet, the rules governing causal language, the corrections process, and where AI assists in producing the site - AI visibility explained: — what AI visibility is, how it works, and why it matters - Services: — AI visibility optimization services - Book a call: — schedule a strategy call ### Proof and evaluation - Case studies (hub): — index of the individual cases below; each has its own page with the baseline, the work completed, the measured before/after, what is disclosed and withheld, and the causal limits of the measurement - — HVAC company, competitive metro market — named in one run and gone the next; four months of source-level work and the measured before/after - — construction company on a brand-new domain — 0% AI visibility at the start, and what it took to earn measurable presence and citations from nothing - — HVAC company whose facts AI stated wrongly — correcting the evidence at source, with presence and brand accuracy measured together - — upholstery service after a site migration knocked it out of AI answers — rebuilding the service map for machines, and what recovered - — destination restaurant — reports share of voice (the share of named brands) rather than mention rate, and says why the two are not interchangeable - — 24/7 locksmith — appearing for emergency-intent prompts across all four major assistants - Sample reports (index): — overview of the three report types - Initial assessment sample: - Monthly scan sample: - Scan comparison sample: ### Articles: measuring AI visibility - Articles (index): — guides, methodology, and research on measuring and improving AI visibility - What is a good AI visibility score: — what the score is, which metrics it combines (mention, recommendation, first-position, share of voice, citation rate), and why a single number without a stated denominator and prompt set is not comparable - Why one AI visibility scan is not enough: — why a single scan is only a first signal, and how repeated measurement across prompts, competitors, citations, and time gives a reliable read - What is AI citation share: — the metric that shows which sources AI answers actually cite in a category: definitions, worked examples, how to calculate it, and how to improve it - AI visibility audit checklist: — a repeatable audit workflow: real buyer prompts run across ChatGPT, Gemini, Claude, Perplexity, and Google AI, tracking mentions, recommendations, citations, share of voice, and sentiment into a fix backlog - AI Overview vs ChatGPT vs Perplexity vs Gemini: — why AI engines disagree about the same business (retrieval, indexes, citations, freshness) and how to measure and fix visibility per provider - What actually moves AI visibility in your category: — how to test category-specific candidate variables instead of assuming universal ranking factors, and how to keep association separate from cause ### Articles: diagnosing why AI skips a business - Why ChatGPT recommends your competitors but not your business: — a five-part diagnostic (entity clarity, reviews, citations, third-party mentions, citable pages) plus a 20-prompt mini audit to locate the evidence gap between a business and the competitor AI names instead - Your AI competitors are not your Google competitors: — why the competitive set AI names diverges from the SERP, and how to build a frozen competitor cohort you can measure against over time - You rank on Google, but AI still skips you: — topic-level gaps between search ranking and AI recommendation, and the corroborated evidence an engine needs before it will name a business - A human sees your website, AI sees missing data: — a teardown of the facts a page states to people but never states to machines, and how to close those gaps - When AI gets your business wrong: — the evidence-repair workflow for incorrect AI descriptions: find the source of the wrong fact, correct it where the model reads it, re-measure - Why AI recommends some local businesses and skips others: — the entity and evidence differences behind local recommendation outcomes ### Articles: improving AI visibility - How to help AI answer your buyers' questions: — AI assistants answer questions rather than ranking pages; three pillars make a site the material answers are built from: machine-legible markup (schema.org, sameAs, OG/Twitter, clean structure), question-shaped content, and verifiable authority and facts - Where should you publish next: — how to turn observed citation sources into a publishing priority order instead of publishing everywhere at once - Set AI visibility targets you can budget for: — turning a visibility gap into a baseline, comparable cohort, target band, owner, timeframe, resource requirement, and verification plan - Local business AI search guide: — Google Business Profile, reviews, NAP citations, service pages, schema, and third-party mentions, each tied to a measurable audit signal - LLMs.txt for AI search: — what the file is, what it is not, which systems say they use or ignore it, and how to maintain and version one - Reddit and AI visibility (forums for GEO without spamming): — why Reddit appears in AI citations, and the non-spammy strategy: authentic participation, answer assets, and measuring community sources - Is Reddit still driving AI visibility: — a recurring source-observation report, and why a citation pattern found in one study, model, or month should not be carried forward as a rule ### Articles: research, cases, and buying - Local window film AI visibility case study: — a measured local engagement: prompt panel, provider-by-provider mention rates, competitive position, and the business signals underneath them - Best AI visibility tracking tools (unranked buyer guide): — public prices, prompt limits, stated coverage, and plan caveats, listed without a ranking - How to evaluate an AI visibility agency: — buyer questions, red flags, and deliverables that separate measurement-first work from vague GEO promises ### Action / conversion - Run a free preview scan: ### Support - FAQ: — common questions about AI visibility, ChatGPT visibility, and the service ### Trust / legal - Privacy policy: — data collection and privacy practices ## Content to avoid / low-signal sections - Admin and authentication pages: `/admin`, `/admin/login`, `/login` - API endpoints under `/api/...` (machine endpoints, not readable content) - User- or scan-specific report pages: `/report/...` (private, per-scan output) - Internal/test routes: `/api-test`, `/report/test`, `/examples/...` - Parameterized URLs (UTM tracking parameters) - `?lang=` on routes that are not language-aware (see Canonical domain above — the localized canonicals of the language-aware routes are legitimate) ## Contact / support - Security issues: security@plastorium.com ## Notes - Prefer citing pages with explicit descriptions of methodology, services, and results over generic navigation. - When referencing a specific case result, cite that case's own page (`/case-studies/`), which carries the measurement, the before/after, and the stated limits. Cite the `/case-studies` hub only for the set as a whole, and prefer either over category or marketing pages. - Downloadable PDF copies of each sample report are available at the same path as the HTML version with a `.pdf` extension.