Everything you
want to know

CiteScan is an AI discoverability audit tool. You enter any URL, we crawl the page and score it across 22 checks, and give you a Discovery Readiness Index (DRI) — a single number showing how likely your content is to appear in ChatGPT, Perplexity, Claude, and Google AI Overviews.

Yes. One page, 22 checks, no sign-up required. The free tier gives you your DRI score, all three pillar breakdowns, and your top 3 priority fixes. Pro features (competitor ranking, share of model voice, full site audit, PDF report) are paid add-ons.

Around 25 seconds. We run three AI scoring agents in parallel after the initial page crawl, so the total time is roughly: crawl (~3s) + parallel scoring (~18s) + summary generation (~4s).

No. CiteScan is entirely web-based. No plugin, no code snippet, no tag manager. Just enter a URL and go.

DRI stands for Discovery Readiness Index. It's a weighted composite of your Visibility (40%), Credibility (35%), and Technical Health (25%) scores — all on a 0–100 scale. A score above 70 means AI engines are actively considering and citing your content. Below 50 means you're largely invisible in AI-generated answers.

Grades map to five bands: A (65+) — your site is in the top band of AI-cited reference pages. B (53–64) — above the market median for AI citation. C (42–52) — mid-market. D (30–41) — below the majority of AI-cited pages. F (below 30) — below the citation threshold for most queries. For reference, Wikipedia's strongest pages score 61–67 on this index, and well-maintained long-form guides typically reach the low-to-mid 60s.

Google ranking and AI discoverability measure different things. Google ranks pages primarily on backlinks, keyword relevance, and content quality. AI engines also need structured data, clear entity attribution, answer-formatted content, and explicit crawler access — signals that most SEO-optimised sites are missing.

Traditional SEO doesn't require author schema, E-E-A-T signals, or FAQ markup — but AI citation does. A well-ranked page with no named author and no JSON-LD schema will score poorly on Credibility, because AI engines can't attribute the content to a credible entity.

Visibility measures whether your content contains citable claims and whether AI bots can read it. Credibility measures whether AI engines trust the source enough to cite it — authorship, schema, E-E-A-T. Technical measures whether the page is structurally accessible: speed, canonical tags, structured data markup, and robots.txt configuration.

Your DRI reflects your site at the time of the scan. Results are cached for 24 hours. After making changes — new schema, robots.txt edits, content rewrites — run a fresh scan to see the updated score. Note that even after your page is updated, AI engines take days to weeks to re-crawl and reflect the change in their answers.

No — GPTBot is OpenAI's training crawler, not its search crawler. ChatGPT search results are driven by OAI-SearchBot. If you've blocked GPTBot but allowed OAI-SearchBot, your pages can still appear in ChatGPT search answers. CiteScan specifically tests the search crawler, not the training crawler.

We check ChatGPT (via OAI-SearchBot), Claude (via Claude-SearchBot), and Perplexity (via PerplexityBot) for search-answer crawl access. We also check training crawler access separately and flag it as a 'partial' verdict when search access is allowed but training access is blocked.

Perplexity crawls live for every query — changes can reflect within days. ChatGPT Search uses OAI-SearchBot and updates over 2–4 weeks. Google AI Overviews use standard Googlebot and can reflect changes within hours to days. Model training data (GPTBot, ClaudeBot) is updated far less frequently — months to years.

Crawler access is just one of four Visibility sub-scores. The others — citation worthiness, entity clarity, and content freshness — also matter. A page that's crawlable but lacks specific statistics, named entities, or clear topical focus will still score poorly on citation probability.

It depends on your scan, but the most common quick wins are: (1) adding JSON-LD Organization or Article schema to your homepage, (2) adding a named author to blog posts with an author schema block, (3) allowing OAI-SearchBot and PerplexityBot in robots.txt if they're blocked. Each of these can be done in under an hour and typically moves your score by 4–10 DRI points.

Technical fixes (robots.txt, schema) are picked up by Perplexity in days, ChatGPT search in 2–4 weeks. Content changes (answer-first structure, new statistics, author attribution) follow similar timelines. Model training data — which affects what AI models 'know' about you — updates over months. CiteScan shows time-to-reflect estimates for each fix.

Some fixes require a developer — JSON-LD schema usually needs to go into your page's <head> tag, which may require CMS access. Others — like adding an author byline, updating a meta description, or editing robots.txt — can often be done through your CMS without technical help. CiteScan labels each fix by effort level.

No. Every DRI improvement is also a standard SEO best practice — faster pages, better schema, clearer authorship, structured content. The scoring model is designed so that anything that improves AI discoverability also improves or is neutral to Google ranking.

Free includes: DRI score, all three pillar scores and sub-scores, top 3 priority fixes, and the AI bot access check. Pro adds: competitor DRI ranking (your score vs the sites AI engines cite in your category), Share of Model Voice tracking (weekly), full site audit (up to 50 pages), and the full PDF remediation report.

Share of Model Voice (SOMV) is the percentage of AI answers — across ChatGPT, Claude, and Perplexity — that mention your brand vs competitors, when users ask relevant questions in your category. We run 50 representative queries per week and track brand mentions. It's the AI equivalent of share of voice in paid media.

We identify the sites AI engines most commonly cite alongside yours in your category, scan each of them with the same 22-check audit, and rank everyone by DRI. You can see exactly which pillar you're leading or trailing on — so you know where the citation gap is coming from.