Know what
we actually measured.
The technical score and the answer test describe different things. Keeping them separate makes the results more useful.
1. A bounded public-page audit
We inspect the HTTPS entry page and up to two same-host internal pages, plus robots.txt, sitemap.xml and llms.txt. No sign-in, form submission or browser execution is performed on the target site. JavaScript-rendered content can be missed.
Access, page metadata, structured data and entity clarity are grouped into weighted categories. Passed checks contribute to the technical readiness score; unknown observations remain visible. Every finding includes the page and evidence used. A missing signal in a three-page sample is not proof it is missing across the entire domain.
2. Separate provider observations
The optional Claude visibility test sends a neutral buyer question to the Claude API with web search. Your brand and target URL are excluded from that prompt, then compared with the returned answer and citations. The report distinguishes a literal brand mention from a citation to your domain. Gemini separately provides a grounded buyer-answer preview with Google Search suggestions and original source links; that preview is not automatically analyzed for visibility.
A mention is not necessarily a recommendation, a source link is not necessarily an endorsement, and one response is not a rank. API answers can differ from consumer chat interfaces. ChatGPT and Perplexity are not tested. Both tests show their actual model, question and timestamp.
Score calculation
| Category | Weight |
|---|---|
| Crawl & discovery | 35% |
| Page identity | 25% |
| Structured data | 20% |
| Answer clarity | 20% |
Within each category, pass earns 1, review earns 0.5 and fail earns 0, weighted by the check’s importance. Unknown and informational checks do not contribute to the denominator. Unavailable categories are excluded and the remaining category weights are normalized. A high score means these limited checks passed; it is not a site-quality certification.
3. Summary and paid fix package
The free summary shows the technical score, category totals, issue counts and the exact scan limits. The one-time paid package contains the complete findings for those sampled pages, source evidence, prioritized fix instructions, verification steps and technical report exports. It does not expand this three-page sample into a complete-domain crawl or include us implementing changes.
Recommendations follow from observed gaps. Research questions are suggestions, not measured keyword volume. Review suggested changes with your full site, verified business facts and editorial standards before publishing. Unknown observations require verification before a change is made. The report can be retrieved only after payment is confirmed by Airwallex; arriving at a payment return page is not proof of payment.
What this does not establish
- No visibility, indexing, rich results, revenue or citation guarantees.
- No complete SEO crawl or browser-based Core Web Vitals measurement.
- No proof of search-engine index coverage from a successful fetch.
- No assessment of proprietary ranking factors or private account data.
- No independent fact-check of every statement, review or offer on a target page.
Beta limits
Per visitor, per UTC day: 20 technical audits and 3 answer tests. Shared service capacity is 500 audits and 50 answer tests per UTC day. Each fetch, page sample and response body is bounded. Requests are rejected if quota enforcement is unavailable.