AI answers · measured, not guessed
When buyers ask AI for a recommendation, what does it say about you?
More people now ask ChatGPT, Perplexity or Google’s AI “who’s the best … near me?”. Cited shows you what those engines can read on your site today, and the paid audit records what they actually answer for your buyer questions, with the exact prompts and the sources they cited.
“Best family dentist near me that’s open on Saturdays?”
Three practices stand out: Your Practice[1], open 9–2 on Saturdays with a children’s plan, Practice B[2] and Practice C[3]…
- yourpractice.example: opening hours, FAQ
- practice-b.example
- a local directory
Easy, we’re pacing requests
You’ve sent a lot of requests in a short time. You can try again in 30 seconds.
What the free check covers
- Access. Whether robots.txt and your firewall let in the crawlers ChatGPT search, Perplexity, Claude, Gemini and Copilot use to fetch pages they cite (search bots and training bots are scored separately).
- Readability. Whether your words are in the HTML without JavaScript, with a clear title, description, main heading and canonical address.
- Clear facts. Contact details, a plain description of what you do and where, and structured data (low weight: studies found no measurable effect on AI citations).
- Your questions. The buyer questions worth testing, drafted from your site, ready for the paid audit’s live answer checks.
- Shopify stores. Product markup, GTINs, descriptions, image alt text and policy pages. How the Shopify check works.
The full rules and weights are on the method page.
The $149 audit
Up to 20 buyer questions you choose, asked live on every engine that is switched on: each answer, whether you were named, who was named instead, the sources cited, and the exact prompt. Plus a 10-page readiness review and best-practice suggestions you can paste. 3 runs in 30 days, so you can compare.
Questions, answered
What does the free check look at?
Your homepage, robots.txt and sitemap, fetched the way AI search crawlers fetch them: whether robots.txt or your firewall shuts out the crawlers ChatGPT search, Perplexity, Claude and Bing use, whether your words are in the HTML without JavaScript, whether the page is indexable, and whether contact details and a clear description are there. It also drafts the buyer questions worth testing. Stores get a product-data check.
Do you ask ChatGPT and Perplexity about me?
Yes, in the paid audit, for every engine that is switched on: each of your questions is sent as written, through the engine’s API with web search, and you get the answer, whether you were named, who was named instead, and the sources cited. The pricing page lists which engines are on right now. Answers vary from run to run and the consumer apps can differ from the APIs, so the report is a dated sample.
Will this get me recommended by ChatGPT?
We don’t promise that and nobody honestly can. Engines decide for themselves. One thing is clear-cut: if robots.txt or a firewall blocks an engine’s search crawler, that engine can’t read your pages. Beyond access, we suggest best-practice fixes (clear facts, direct answers), but 2026 studies found that llms.txt and schema markup on their own made no measurable difference to AI citations. That is why the audit measures, and why you get 3 runs to see whether anything changed.
Is blocking AI crawlers bad?
It depends which ones. Search crawlers (OAI-SearchBot, PerplexityBot, Claude-SearchBot, Googlebot, Bingbot) fetch pages so engines can cite them. Training crawlers (GPTBot, ClaudeBot, Google-Extended, CCBot) collect training data; blocking them is a fair business choice and, according to the engines’ own documentation, doesn’t remove you from their search answers. The report treats them separately.
Should I add llms.txt?
Only if you want to. An Ahrefs study of 137,000 sites (June 2026) found 97% of llms.txt files got no requests at all, and no major answer engine says it reads them. The check doesn’t score it.
Do you store my site?
No. We fetch public pages, build the report and return it to your browser. Nothing is written to a database. See privacy.