# EarnCitations > EarnCitations is building a citation engine for AI search: it checks whether AI crawlers can read a website, measures how AI answer engines (ChatGPT, Gemini, Claude, Perplexity) answer buyer questions, and finds why a business is or isn't named. Status: pre-launch. Available now, free: the AI-readiness audit, fix kit, question doctor and question map, what AI sees, answer check, delivery check, docs check, watch, AI log analyzer, Fact Ledger tools and the MCP server. Planned: fixes shipped as pull requests or CMS drafts, help earning third-party sources, and billing per verified citation. Principles: earned, not gamed — no hidden prompt injection, cloaking, fake reviews or undisclosed posts. Boundaries: the free checks read public pages the way an AI crawler does (no JavaScript, robots.txt honoured) and return findings and drafted files. They don't log in, change a site or publish anything; the site owner reviews and deploys the fixes (IndexNow pings and alerts are sent only on request). Whether an AI engine cites a site is that engine's decision: EarnCitations measures it and doesn't control it. Markdown: every page listed here also has a Markdown copy at its URL plus .md (https://earncitations.com/pricing.html.md, or /pricing.md; the homepage is /index.html.md), rebuilt with the page: the page's summary on top, its links and tables kept as Markdown. The whole site as one file: https://earncitations.com/llms-full.txt ## Product - [Home and free AI-readiness audit](https://earncitations.com/): what EarnCitations does, the Citation Stack, pricing - [Pricing](https://earncitations.com/pricing.html): free audit; Starter $149, Growth $319 and Scale $999 a month (50, 400 and 1,000 tracked buyer questions, checked weekly, every other day and daily; 14-day free trial); custom Enterprise; or Earned, $0 base plus $2–$60 per verified net-new AI citation, capped monthly - [Compare](https://earncitations.com/compare.html): honest comparison with Jam, AI-visibility trackers, SEO agencies and DIY, including when each is the better choice - [Research: how AI-ready are 124 software companies? (2026-11-21)](https://earncitations.com/research.html): median score 87/100; none block OpenAI, Anthropic or Perplexity search crawlers in robots.txt; 76 of 117 publish llms.txt; schema and comparison pages are the common gaps - [Research: do software companies answer their own buyers' questions?](https://earncitations.com/research.html#questions): 799 questions across 116 sites; 23% say plainly whether there's a free trial, 16% whether they're SOC 2 compliant - [Research: Markdown copies for AI agents](https://earncitations.com/research.html#markdown-copies): 34 of 110 software pricing pages send a Markdown copy when asked; 13 of those leave out prices the page shows - [Research: would a pricing-page fix reach AI crawlers?](https://earncitations.com/research.html#delivery): Delivery check on /pricing for 123 software sites; 104 of 111 live pages are self-canonical, but only 56 of the 93 in a sitemap carry a lastmod, and up to 37 of 89 sites with llms.txt don't link their pricing page from it (a floor for those that do: links to /pricing.md weren't counted then) - [Research: robots.txt names that match no crawler](https://earncitations.com/research.html#robots-names): lookalike dashes and invisible characters in User-agent lines; none in 115 software robots.txt files, one known case outside the sample - [Research: could a coding agent set these products up from their docs?](https://earncitations.com/research.html#docs): docs check from the homepage of 123 software sites; developer docs found for 109; 1 passes all twelve checks (median 8); 17 of 94 quickstarts have no code in the HTML, and 26 of 109 have an OpenAPI contract we could find - [AI crawler directory](https://earncitations.com/crawlers.html): 26 AI crawlers and robots.txt tokens from 12 operators, what each does per the operator's own docs, which ones may bypass robots.txt, and copy-paste rules - [AI crawler dataset (JSON)](https://earncitations.com/crawlers.json): the same 26 agents with purpose, documentation URL and a dated changelog - [llms.txt generator and validator](https://earncitations.com/llms-txt.html): draft an llms.txt from your own pages; validate any llms.txt in the browser - [Organization schema generator](https://earncitations.com/schema.html): build Organization or LocalBusiness JSON-LD with sameAs, logo, address and hours; optional Wikidata lookup - [AI crawler log analyzer](https://earncitations.com/logs.html): drop in a server access log to see which AI agents visited, which pages people asked assistants about, which requests failed, whether agents read your llms.txt and Markdown copies (or ask for ones you don't have), and how many people who clicked through from an AI answer reached a signup, demo or contact page; runs in the browser, nothing uploaded; also a GA4 "AI assistants" channel regex built from the same assistant list (logs.html#analytics, `earncitations logs --channel`); and, for sites without server logs, a reader for a GA4, Plausible, Fathom or Matomo export (one row per source) that sets AI-assistant sessions, key events and revenue against everyone else's (logs.html#export, `earncitations logs --analytics export.csv`) - [Research: hotels and restaurants](https://earncitations.com/research-local.html): 150 independent US hotel and restaurant websites from a reproducible Wikidata sample; which load, which answer "what time is check-in?" and "what are your hours?", with method and data (research-local.json) - [What AI sees](https://earncitations.com/view.html): one page as an AI crawler that doesn't run JavaScript reads it: the text, the facts it can take (prices, hours, phone, address, structured data), widgets whose content isn't in the HTML, blocked agents, buttons and fields an agent can't name, WebMCP form problems, images with no size, and a visible "Updated" date that disagrees with dateModified - [Delivery check](https://earncitations.com/delivery.html): after you ship a fix, confirm each step separately: live, the new text in the HTML without JavaScript (and in the page's Markdown copy, if it has one), allowed in robots.txt, indexable, in the sitemap with a fresh lastmod, linked, in llms.txt - [Docs check](https://earncitations.com/docs.html): could a coding agent (Claude Code, Cursor, Codex) set your product up from your public docs? Twelve checks read as an AI crawler would: a quickstart it can find, code as text, install command, authentication, what success looks like, OpenAPI, errors, versions, llms.txt; it doesn't run an agent. Compare with a competitor's docs, check by check - [Answer check](https://earncitations.com/claims.html): paste an AI answer and its cited links; every claim is checked against the pages it cites (supported, contradicted, unaddressed or unknown), with prices, durations and percentages matched exactly - [Answer log](https://earncitations.com/track.html): save the AI answers you check and see how often they name you, recommend you and cite your pages over time; discovery and diagnostic questions apart, rates per topic (a "## Pricing" line in your set), over all answers or the last 7, 30 or 90 days, every denominator shown, attempts that got no answer counted next to every rate; competitors counted the same way; load your frozen question set to see what to ask next on each engine; compare answers before and after a change (improved, unchanged, worse, inconclusive); kept in your browser, exportable as JSON lines for `earncitations track report` - [Method](https://earncitations.com/method.html): how to measure what AI answers say about you, step by step: freeze the questions (discovery and diagnostic apart), ask on a schedule, count per answer four ways, check every claim against its sources, classify and fix, confirm delivery, re-ask; each step with its free tool - [Question map](https://earncitations.com/map.html): the buyer or guest questions a site is likely asked, proposed from its own pages, each with a verdict (answered, weak, missing, crawl problem) and the next step - [Question doctor](https://earncitations.com/question.html): for one buyer question, whether the site's answer is blocked from AI crawlers, missing, or there but weak, with the page to fix and the off-site sources to work (own, claim, earn, contest); add a competitor's domain to see what their page gives an assistant that yours doesn't - [AI price check](https://earncitations.com/prices.html): the prices an AI crawler reads in a pricing page's HTML (no JavaScript) next to the prices in its Offer markup, and where they disagree - [Head-to-head](https://earncitations.com/vs.html): the same AI-readiness audit on two sites, showing only the checks and crawler rules where they differ - [How to use EarnCitations, in order](https://earncitations.com/routine.html): the free tools in order: the first hour (can AI read you?), the first week (what buyers ask), then every week and month - [AEO glossary](https://earncitations.com/glossary.html): 34 AI-search terms (robots.txt groups, Google-Extended, sameAs, llms.txt, grounding), each linked to its primary source - [AI-readiness checklist](https://earncitations.com/checklist.html): every check the free audit runs, with fixes - [Fact Ledger specification](https://earncitations.com/ledger.html): open standard for signed, evidence-linked business facts at /.well-known/facts.json - [Answer Index methodology](https://earncitations.com/answer-index.html): how the public benchmark of AI recommendations is measured - [Manifesto](https://earncitations.com/manifesto.html): why we refuse to game AI engines - [About](https://earncitations.com/about.html): what exists today (pre-launch) and how to contact us - [What's new](https://earncitations.com/changelog.html): new checks and changes by day, newest first - [Data and security](https://earncitations.com/security.html): what the free checks keep (in memory only, minutes to hours), what stays in your browser, and the third parties involved - [Our crawler](https://earncitations.com/bot.html): what the EarnCitationsAudit user agent fetches, how it honours robots.txt, and how to block it - [Developers](https://earncitations.com/developers.html): CLI, MCP server (audit_site, check_prices, check_robots, analyze_logs, analyze_analytics_export, answer_log_report, training_presence, diagnose_question, question_map, crawler_view, check_answer_claims, check_delivery, check_docs, draft_fixes, ledger_verify_domain, ledger_validate, ledger_draft), HTTP API, and IndexNow submission from the CLI - [OpenAPI contract](https://earncitations.com/openapi.json): the HTTP API as OpenAPI 3.1, with every error status and whether retrying is safe (also at developers.html#errors); listed in the RFC 9727 API catalog at https://earncitations.com/.well-known/api-catalog - [API limits](https://earncitations.com/developers.html#limits): requests a minute, cache time and time limit for each endpoint ## Facts - [EarnCitations Fact Ledger](https://earncitations.com/.well-known/facts.json): signed canonical facts about EarnCitations - [Fact Ledger public key](https://earncitations.com/.well-known/facts.pub) ## Optional - [Full text of every page](https://earncitations.com/llms-full.txt): all pages as one markdown file - [MCP manifest](https://earncitations.com/.well-known/mcp.json): stdio MCP server description