Is your page ready for AI?
Audit one page the way AI crawlers read it: crawler access, structured data, headings, questions and answers, and how directly it answers. A score out of 100 and a fix for each gap. Free, no sign-up.
What the audit checks
The audit reads one page and its site's robots.txt, as an AI crawler would, without running JavaScript. Each section below counts towards a score out of 100. A problem scores nothing, a suggestion half and a pass in full.
Crawler access (35 points)
Whether AI search and assistant crawlers may read the page: its status, the robots.txt rules for its path, any noindex setting, where its canonical link points, and how much text arrives before any JavaScript runs. If they cannot read it, the audit says so instead of giving a score.
Structured data (20 points)
Whether the page has JSON-LD, whether it parses, and whether it has the properties Google documents for its types. The schema checker shows the same check node by node.
Headings (15 points)
One h1 for the page's topic, subheadings that name each section, no skipped levels and no empty headings. A clear outline lets a reader, or a program, find the part that answers a question.
Questions and answers (10 points)
Whether the page answers the questions people ask, as headings ending in a question mark, and whether FAQPage markup pairs each question with its answer.
Answer-ready writing (20 points)
Whether the page answers up front: an opening paragraph of 20 to 80 words, a direct first sentence under each question heading, paragraphs that average no more than 100 words with few over 150, at least one list or table, a date on the page or in its structured data, and an author or publisher in the structured data.
Common questions
What does the score measure?
How ready one page is to be read and quoted by AI search and assistants, on five things we can check from its HTML and its site's robots.txt. It does not measure whether assistants mention you; that depends on far more than one page, and it is what Vouch itself tracks.
Why is there no score for my page?
When AI search and assistant crawlers cannot read the page, nothing else on it matters yet. That happens when the page does not answer with status 200, when robots.txt blocks a crawler such as OAI-SearchBot, PerplexityBot or Claude-User from its path, or when a robots meta tag or X-Robots-Tag header says noindex.
Does blocking GPTBot lower my score?
No. GPTBot, ClaudeBot, CCBot and Meta-ExternalAgent collect pages for AI training. Blocking them keeps your pages out of that training data but does not stop AI search crawlers, so the audit notes it and leaves the score alone.
Where do the word limits come from?
They are our rules of thumb, not figures any assistant publishes: an opening paragraph of 20 to 80 words with a first sentence of at most 30, and an answer under a question heading that starts with a sentence of at most 40 words. A passage that short can be lifted whole as an answer. Missing one is a suggestion, never a problem.
Does FAQ markup still matter?
Google stopped showing FAQ rich results on 7 May 2026, so FAQPage markup will not change how the page looks in Google. Write a questions section because readers like one and assistants can quote it; the markup then states each question and answer as a pair for any program that reads it.
What does the audit store?
A page's result is kept in memory for an hour, so repeat checks and shared result links do not hit the page again. Nothing about the page is saved to a database. The address of the person checking is only counted, hashed, for the rate limit.