How the DigiJaws AI-Readiness Score Works: The 18 Signals

By Beau Thoutt · Updated September 2026

The DigiJaws AI-readiness score is a 0-to-100 number, plus a letter grade, built from eighteen weighted signals grouped into three categories: Discovery, Understanding, and Trust. Each signal is a concrete, checkable fact about your page or its machine-facing files. Higher-impact signals carry more weight, so the score reflects what actually moves AI visibility — not a checklist of equal boxes.

How the score is built

Every scan fetches your live page plus your robots.txt, llms.txt, and sitemap.xml, then evaluates each of the eighteen signals as pass, partial, or fail. Signals are weighted by impact and rolled up first into three category scores, then into one overall score from 0 to 100. The overall score is summarized as a letter grade so you can read it at a glance.

The three categories

Discovery — can AI reach and index your content?

Whether crawlers and agents can physically get to your pages and know where everything lives.

Understanding — can AI parse what your page means?

Whether your titles, metadata, and structured data let a model extract your meaning cleanly.

Trust — can AI rely on you as a source?

Whether identity and security signals mark your site as a credible, citable source.

The eighteen signals

All eighteen are listed below, grouped into the three categories. Exact field names, value types and the JSON the API returns are in the Data API reference.

Discovery

• AI-crawler access — does robots.txt allow GPTBot, ClaudeBot, PerplexityBot, Google-Extended and friends? Blocking them makes you invisible to AI search.
• Sitemap.xml — a valid, reachable sitemap so every URL is discoverable.
• llms.txt — the standard file that tells language models what your site is and points to your key content.
• Canonical tag — one clear canonical URL so engines index the right version and do not split signals across duplicates.

Understanding

• Title tag — a specific, descriptive title element engines can use as the headline.
• Meta description — a concise summary engines can lift as a snippet.
• Open Graph — OG tags for clean titles, descriptions, and previews when your link is shared or cited.
• Structured data — JSON-LD that labels what your content is, so models do not have to guess.
• Answer-first content — passages that state the answer up front, so an engine can lift them without rewriting.
• Content depth — enough substance on the page for an engine to have something worth quoting.
• Single H1 — exactly one top-level heading, so the page has one unambiguous subject.
• Subheading structure — a logical H2/H3 order, so models can follow what each section covers.

Trust

• HTTPS — a secure connection, a baseline credibility signal.
• Organization schema — machine-readable identity: who you are, your name, logo, and links.
• FAQ schema — structured questions and answers that answer engines can extract directly.
• Author & date signals — a named author and a visible published or updated date, so an engine can judge who wrote this and when.
• Entity consistency — the same brand name, description and identifiers across your site and your public profiles, so engines resolve you to one entity.
• Citations & sources — outbound references to primary sources, which raise the odds an engine treats your page as citable itself.

How weighting works

Signals are not equal. Being blocked at robots.txt or missing HTTPS caps your ceiling no matter what else you do, so those carry heavy weight. Schema and llms.txt are high-leverage because they unlock machine understanding. Smaller polish items still count, but the engine surfaces the highest-impact gaps first in your action plan, so you fix what matters most before the details.

From score to grade

The overall 0-to-100 score is summarized into a letter grade that bands AI-readiness from A (genuinely AI-ready) down to F (largely invisible to AI engines). To see your own exact score and grade — and which specific signals are passing or failing — run your URL through the SEO Agent. It returns the number, the grade, the per-signal breakdown, and the fixes in one pass.

How to improve each category

Discovery gaps are usually fixed in minutes: allow AI crawlers in robots.txt, publish a sitemap, and add an llms.txt (the AI-Fix Generator drafts one from your site). Understanding gaps are solved with clean titles, a meta description, Open Graph, and JSON-LD. Trust gaps come from HTTPS plus Organization and FAQ schema. Every one of these has a generated, copy-paste fix — see how DigiJaws works.

FAQ

What is a good AI-readiness score?

Higher is better, and an A grade means a site is genuinely ready for AI engines and agents. Most sites that have never optimized for AI score in the failing range because llms.txt, AI-crawler rules, and machine-readable schema are missing. Run the SEO Agent to see where you stand.

Why did my score change between scans?

Because the scan is live. A redeploy that strips schema, a robots.txt edit, or a dropped canonical tag will move the number. That is exactly why account monitoring alerts you on a drop of eight points or more.

Which signal should I fix first?

Whatever your action plan lists first — the engine orders fixes by impact. In practice, unblocking AI crawlers, adding llms.txt, and adding structured data give the biggest early gains.

Does this replace my SEO score?

No. It is a separate, complementary measure focused on AI engines and agents. See what AI search readiness is.