Our own site scored 60. We followed our own report to 100.
On August 3, 2026 we pointed AnswerWorthy at answerworthyai.com, paid attention to every finding, shipped the fixes the same day, and re-scanned. Both reports are live, so you can verify every claim on this page yourself.
Published: August 3, 2026 · By Federico De Faveri
What did our own audit find?
A 60/100, and it was earned. Freshness scored 10: no dates anywhere, in markup or on the page. Evidence and provenance scored 17: no author, and zero outbound links behind a hero that said “built on research.” Entity clarity scored 33: no Organization schema, no sameAs profiles. The AI analyst pass was blunter than the checklist: it flagged that our research claim named no proof point a reader could follow, and that the page promised a specific, personalized report without demonstrating that specificity anywhere.
What did we change?
- Organization schema with our real legal entity, founder, and sameAs links to profiles that actually exist. We did not invent any.
- A visible FAQ answering the buyer questions the report predicted, mirrored verbatim into FAQPage markup.
- Named research sources with outbound links: the GEO paper (KDD 2024), Google's AI features documentation, and the Semrush AI Visibility Index.
- A named author, published and modified dates, and a visible “last updated” line.
- An llms.txt file and a tighter meta description.
Nothing on that list is exotic. It is the same fix plan the report hands any site, applied without shortcuts.
What happened to the score?
| Category | Before | After |
|---|---|---|
| Crawl access | 94 | 100 |
| Structured data | 67 | 100 |
| Entity clarity | 33 | 100 |
| Passage extractability | 100 | 100 |
| Evidence & provenance | 17 | 100 |
| Freshness | 10 | 100 |
| Buyer-prompt coverage | 50 | 100 |
| Click-worthiness | 80 | 100 |
| Overall | 60 | 100 |
Can your team or your coding agent do the same?
Yes. Every unlocked report is the same artifact we followed: every verified issue with the evidence behind it, a fix plan ordered by impact with effort labels, and the buyer prompts your page should be answering. Each report also includes an agent-ready fix brief: one click copies the whole thing as markdown you can paste straight into a coding agent like Claude Code or Cursor and say “work through this.” That is exactly how these fixes shipped in a day.
What we won't claim
A 100 readiness score is not a promise that assistants recommend you, and we refuse to pretend otherwise. Our own report says so: at scan time, the live answer checks truthfully showed web-search assistants naming competitors, not us, for our own buyer prompts. Readiness is the part you control: it removes every observable blocker to being found, understood, cited, and trusted that published research identifies. The mention itself is earned over time, and measuring it honestly is the point of the product. This page is about our own site, which we control end to end; platform constraints can make some fixes harder on yours.
See what your page scores
The scan is free, takes under a minute, and shows your score plus one personalized AI insight before you pay anything.
Run my free scan