METHODOLOGY · UPDATED 6 OCTOBER 2026
Evidence you can inspect. Limits you can understand.
Citequay separates technical website readiness, AI answer observations and business outcomes. Each answers a different question.
Technical readiness is a weighted diagnostic of checked website evidence. AI citation research is a limited sample of sourced API answers. Business impact requires your own measurement of enquiries, customers or revenue. None of these is interchangeable with the others.
What does the website audit read?
The scanner retrieves the submitted public HTML page and discovers up to seven supporting pages on the same origin. It prioritises relevant service, product, about, contact, location, menu and pricing links. It also checks robots.txt and sitemap evidence. Report coverage lists the pages read, failed requests and pages omitted from the bounded scan.
The scanner respects its robots.txt access rules. It does not sign in, access private networks or submit checkout forms. URL queries and fragments are removed before crawling. Pages that require JavaScript to reveal their content may provide incomplete initial HTML evidence.
Which technical checks are included?
| Check | Evidence examined |
|---|---|
| Secure website | Whether the checked page uses HTTPS. |
| Page title | The title element in the fetched HTML. |
| Search description | Meta description content. |
| Main heading | The number and text of H1 headings. |
| Readable content | Text available in the initial HTML. |
| Business identity | Recognised structured business data, rather than unrelated Product or WebPage markup alone. |
| Indexing directives | Explicit noindex directives in HTML and response headers. |
| Preferred page address | Canonical URL evidence. |
| Contact information | Accessible contact signals on the checked pages. |
| Sitemap | An accessible XML sitemap, not just a robots.txt declaration. |
| Crawler access | Observable search-crawler rules in robots.txt. |
How is technical readiness calculated?
Each check has a diagnostic weight. Page-level checks receive proportional credit based on how many scanned pages pass. Unknown checks are excluded from the denominator. The report retains the evidence behind passed, action-needed and unverified findings.
The score is a technical heuristic. Its thresholds are Citequay diagnostic rules, not Google requirements. A higher score does not prove better rankings, more citations or more customers. Different crawl coverage or scoring methods can invalidate a direct comparison; Citequay withholds incompatible score comparisons.
What does the audit not test?
It does not verify search-engine index status, render every page with JavaScript, measure real-user Core Web Vitals, complete a purchase, assess every backlink or perform a security audit. A missing signal in the sample does not prove that it is missing everywhere on the site. Use dedicated tools and professional review where those questions matter.
How are AI answers evaluated?
A paid research run uses six buyer intents and one configured web-enabled API model. One optional repeat uses the same questions and setup. The report preserves the provider, model, timestamps, answer text and sources. Failed answers are separated from successfully completed answers, rather than silently treated as zero visibility.
Website citations require a source link to the target domain or its subdomains. Third-party mentions do not count as target-domain citations. Business names in answers are matched conservatively. Other cited websites may be directories, publications, suppliers or unrelated sources; they are not automatically competitors.
The AI citation research guide explains how to interpret the sample. API answers differ from consumer apps, and the sample does not measure a whole market.
How should an improvement be verified?
- Read the affected page and confirm the finding is relevant.
- Keep the original, review the proposed change and publish through an authorised editor.
- Inspect the live page. Marking an action “implemented” is a work record, not automatic proof.
- Repeat an audit of the same address and comparable scope.
- Measure commercial outcomes separately over comparable periods. A before/after change alone does not prove causation.
Who is responsible for this methodology?
Citequay is operated by Synapsek Labs PTY LTD. Reports expose their methodology and evidence so customers can question the result. Send a suspected error, affected page address and check date to support@citequay.com. Do not email passwords or private access tokens.