GPT Ready Check v5.0 — Scoring Methodology & Technical Limitations

Transparent breakdown of how our 100-point AI Search Technical Readiness Score is calculated and what a static HTML audit can and cannot measure.

← Back to Scanner

1. Score Architecture (10 Categories • 100 Points Total)

Our scanner evaluates 37 technical signals across your homepage, /robots.txt, /sitemap.xml, and up to 3 discovered internal subpages (/contact, /about, /services):

27 Scored Core Checks (100 Points Total): Every scored check is grounded in verifiable HTTP headers, RFC 9309 robots.txt rules, XML sitemap entries, HTML metadata, or Schema.org syntax.
7 Advisory Best-Practice Checks (0 Points Max): Emerging GEO/AEO signals (extended 15-bot matrix, snippet permission flags, sameAs links, OpenGraph tags, HTML5 landmarks, freshness timestamps, and FAQPage schema) are reported for visibility but never deduct points from your 0–100 score.
1 Unmeasured Static-HTTP Limitation Notice (Check 10.4 — 0 Points): Explicitly discloses that visual mobile layout & tap-target rendering requires a full browser engine and is not scored in a static HTTP check.
2 Experimental Checks (EXP.1 /llms.txt & EXP.2 /llms-full.txt — 0 Points): Community proposals that are not required by OpenAI, Google, or Bing.
Pro-Rata Scaling for Not-Tested Checks: If a page has 0 <img> tags (Check 10.2) or is a client-side JS shell (Check 6.3), that check is marked NOT_TESTED and excluded from the denominator so your site is never unfairly penalised.

Category	Max Points	Scored Checks & Justification
1. AI Crawler Accessibility	14 pts	`1.1` Search & Answer Bots (`OAI-SearchBot`, `Googlebot`, `PerplexityBot`) — 10 pts; `1.2` Training vs Search Separation — 4 pts; `1.3` Extended 15-Bot Matrix — 0 pts (advisory). RFC 9309 allow-by-default receives full PASS points; explicit `Allow:` lines are never required unless overriding a broader `Disallow`.
2. Robots.txt & Bot Governance	8 pts	`2.1` Valid `/robots.txt` HTTP & syntax — 3 pts (clean 404 earns 2/3 WARNING); `2.2` Sitemap directive in `robots.txt` — 2 pts; `2.3` Static CSS/JS asset access — 3 pts.
3. XML Sitemap & URL Discovery	10 pts	`3.1` Reachable XML Sitemap — 5 pts; `3.2` Valid `<loc>` URL entries — 3 pts; `3.3` Sampled internal link health (no dead 404s) — 2 pts.
4. Indexability & Canonicalisation	12 pts	`4.1` No `noindex` block — 6 pts; `4.2` Clean self-referencing `<link rel="canonical">` — 4 pts; `4.3` Direct URL resolution (≤2 redirects) — 2 pts; `4.4` Snippet permissions — 0 pts (advisory).
5. Structured Data & Entity Schema	12 pts	`5.1` JSON-LD syntax validity & block presence — 4 pts; `5.2` Semantic validity & zero conflicting Organization entities — 4 pts; `5.3` Entity attribute completeness — 4 pts; `5.4` `sameAs` links — 0 pts (advisory). Sites without JSON-LD that have clean HTML identity receive partial credit (7/12 WARNING) rather than failing as broken.
6. Metadata & Content Structure	12 pts	`6.1` Title & summary description presence (no arbitrary character-count failure) — 5 pts; `6.2` Heading hierarchy (`<h1>`/`<h2>`) — 3 pts; `6.3` Server-rendered prose (≥80 words combined or ≥50 homepage words = full 4/4 PASS) — 4 pts; `6.4`/`6.5` OpenGraph & landmarks — 0 pts.
7. Entity & Business Identity Clarity	10 pts	`7.1` Distinct brand name — 4 pts; `7.2` Crawlable phone/email across homepage + `/contact` — 3 pts; `7.3` Geographic/service region clarity — 3 pts.
8. Answerability & Citation Readiness	10 pts	`8.1` Question/FAQ headings — 4 pts; `8.2` Scannable lists or tables — 3 pts; `8.3` Clear opening summary paragraph — 3 pts; `8.4`/`8.5` Freshness & FAQ schema — 0 pts.
9. Trust, Security & Legitimacy Signals	6 pts	`9.1` HTTPS encryption — 3 pts; `9.2` About, Contact & Privacy/Terms links — 3 pts.
10. Baseline Technical & Accessibility Signals	6 pts	`10.1` Responsive `<meta name="viewport">` tag — 2 pts; `10.2` Image `alt` attribute coverage (excluded if 0 images) — 2 pts; `10.3` Initial server response health (≤2500ms PASS, 2501–5000ms WARN, >5000ms FAIL) — 2 pts; `10.4` Visual mobile layout — 0 pts (Not Tested limitation disclosure).

2. Five Things Our Static HTML Scanner Cannot Measure

Actual AI Recommendation Frequency: Passing all technical checks confirms that AI search crawlers can access and parse your site, but does not measure or guarantee how often ChatGPT, Perplexity, or Google AI Overviews recommend your business.
Visual Mobile Layout & Tap-Target Spacing: We inspect raw HTML and HTTP headers without rendering CSS in a mobile browser viewport. Use Google Lighthouse for visual tap-target and Cumulative Layout Shift (CLS) testing.
Client-Side JavaScript Hydration: If your website renders text exclusively inside a client-side React/Vue bundle after page load, our static HTTP scanner marks body text as NOT_TESTED rather than guessing.
Off-Site Authority & Backlinks: AI engines weigh third-party citations, reviews, and backlinks across the web, which a single-domain technical scan does not measure.
Multi-Region User Latency: Check 10.3 measures a single HTTP response from our server to catch severe timeouts (>2.5s–5.0s), not global CDN edge latency.