Page test

Page test answers one question: would an AI engine quote this page when answering this prompt? Open Page test in a project, enter a URL and a prompt, and click Test page. It takes about 20 seconds.

What's checked

Access (30% of the score). The page is fetched the way an AI crawler sees it — no JavaScript — and checked for:

• HTTP 200 and no noindex
• robots.txt allowing AI search bots (OAI-SearchBot, ChatGPT-User, PerplexityBot, Claude-SearchBot, Googlebot) to fetch that URL
• Main text in the raw HTML, and at least ~300 words
• A title, one H1, structured data, a visible date and question-style headings

Content (70% of the score). A language model reads the page text and grades how directly, specifically and credibly it answers the prompt.

What you get

• Score out of 100, split into content and access
• Verdict — answers the prompt, partly answers it, or doesn't answer it
• Missing from the page — what a complete answer needs that the page lacks
• Edits, most impactful first
• Suggested opening — a 40–70 word paragraph that answers the prompt using only facts from the page (fill in anything marked [confirm])
• Most quotable passage today — the sentence an engine is most likely to lift now

A good workflow

1. From a content gap in the action plan, click Test a page against this prompt.
2. Test your page, then the competitor page engines cite for the same prompt.
3. Apply the top edits, publish, and test again.
4. Mark the action done so aeotime measures the impact.

Recent tests are listed at the bottom of the page so you can compare before and after.