Check the foundations.
We fetch a small page sample and inspect crawl rules, metadata, structure and identity.
A clearer view of AI search
Be the answer.
Start with the evidence.
Understand what your website makes easy to discover.
Find the gaps. Test an answer. Know what to fix next.
Answer testingClaude observed answer & citations✦ Gemini grounded answer previewChatGPT · Perplexity not tested
Your evidence, organized
Website audit
Technical readiness—out of 100 · not an AI rank
Evidence before a promise.
This is a bounded public-page check. It shows what we measured, what needs attention and where the sample ends.
What the pages tell us
Your next steps, in order.
Review each finding against your full site before changing a template.
Questions worth answering
Suggested research prompts. Add a topic to make them more specific; search demand has not been measured.
A separate observation
What does a grounded answer say?
Test a neutral buyer question. Claude web search shows observed mentions and citations; Gemini shows a grounded answer preview. The brand and audited URL are excluded from the question sent to either provider.
ChatGPT and Perplexity: not tested. A brand mention and a link to your domain are different outcomes. No score or trend is inferred from one answer.
A before and after you can inspect.
Opt in to keep up to 10 reports in this browser. We do not create an account or store report bodies on our server.
Scope and limitations
Clarity over guessworkA score is a start.
A score is a start.
The evidence is the point.
Move from a vague “optimize for AI” to a specific page, a specific gap and a practical next step.
See how we measure ↗01 Discoverability02 Clear entities03 Useful answers
Evidence reportIllustrative example
72/100
Technical readiness4 areas · every finding has a source
✓
Page is publicly accessibleHomepage · HTTP 200
Passed!
Clarify the publisherOrganization identity not found
ReviewEvidence
Organization: not found in sampled JSON-LDAdd verified organization details to the site template, matching visible content.
One fix. A clear reason. A page to check. ↗
From question to actionThree steps.
Three steps.
No black box.
A real test of public pages, followed by recommendations you can inspect.
Read the evidence.
Every result points to what was observed. Priorities and practical fixes follow from those findings.
Test. Improve. Repeat.
Try a grounded answer, download the report and compare another audit after making changes.
Built for useful workNot another
Not another
unexplained AI score.
Observed
Public-page findings with URLs, timestamps and a visible sample size.
Separated
Technical readiness and provider answers are different kinds of evidence.
Actionable
A prioritized fix plan, research questions and portable JSON or Markdown reports.
Good questions
Clear answers.
Does the score tell me where I rank in ChatGPT?
No. The score measures technical readiness on a bounded sample of public pages. A separate Claude web-search test shows one observed answer, brand mention and domain citations with its prompt and time. Gemini provides a grounded buyer-answer preview. ChatGPT and Perplexity are not tested.
What does an audit actually check?
We fetch your entry page and up to two useful internal pages, plus robots.txt, sitemap.xml and llms.txt. We inspect access, metadata, page structure, entity details and structured data, then attach evidence to each recommendation.
Can I use it without an account?
Yes. The beta includes 20 technical audits and 3 answer tests per visitor per UTC day, subject to shared capacity. Reports can be downloaded. Optional history stays in this browser.
Will the tool make changes to my website?
No. It produces a prioritized action plan. You review and implement the changes in your own site. Suggested questions are research starting points, not measured search demand or verified business claims.
Is a schema or llms.txt file enough to get cited?
No. Structured data can help describe your visible content, and llms.txt is an optional discovery file. Neither guarantees indexing, a citation, a rich result or a sale. Useful content and trustworthy sources still matter.
Can I compare results after making improvements?
Yes. Choose to save reports on this device, run another audit and compare readiness scores and findings. This is a local report history, not continuous rank tracking.
Make your next
improvement count.
20 technical audits and 3 answer tests per visitor per UTC day.
Shared capacity: 500 audits and 50 answer tests per day.