Measurement
How to read your grade: what each score band means
14 August 2026 5 min read
A single number is only useful if you know what it's made of. Beacon's grade is 100 points split across four measured categories, with AI readiness reported separately because it mixes measured checks with model estimates and shouldn't be blended into a hard score.
The four graded categories
Each is scored from checks we can observe directly from a crawl.
- Performance (30) — server response time, page weight, JavaScript dependency, reachability and internal linking.
- SEO (30) — titles, descriptions, headings, canonicals, sitemap, robots, social tags, indexability and alt text.
- Mobile (30) — viewport, language declaration, semantic landmarks, content depth and error handling.
- Security (10) — HTTPS, HSTS and security.txt.
AI readiness, reported separately
This covers llms.txt, AI crawler access, structured data, entity naming, answer depth, freshness, pricing clarity and comparison content. It is shown as a percentage rather than folded into the 100 so you can see technical health and AI positioning move independently.
Score bands
Rough guidance for what a total means in practice.
- 90-100 — ahead of most sites we scan; remaining fixes are marginal.
- 75-89 — solid foundations, a handful of targeted fixes left.
- 50-74 — buyers and assistants are hitting friction; work the priority list.
- Under 50 — structural gaps costing visibility in both search and AI answers.
Measured vs estimated
Anything derived from fetching your site is labelled measured. Anything derived from asking a model is labelled estimated and carries a count of how many engines agreed. Keep that distinction when you report internally — it protects you from defending a number you can't reproduce.
See where your own site stands
Run the free grader for a 100-point score across performance, SEO, mobile and security, plus an AI readiness rating and a prioritised fix list.
Grade my site