GeoTest — Methodology · GEO v3.0
How the GEO Score works
A 0–100 rating built from six weighted categories, calibrated against citation behavior observed across ChatGPT, Perplexity, Google AI Overview, Claude, and Gemini (2024–2026).
Six weighted categories · 100 points total
| Category | Weight | What it measures |
|---|---|---|
| Content Quality | 25 pts | Depth, heading hierarchy, cornerstone pages of 1,500+ words. |
| Citation Worthiness | 20 pts | Factual density, quotable definitions, comparison tables, freshness. |
| Structured Data | 18 pts | Organization, WebSite, Product, Article, Breadcrumb, FAQPage JSON-LD. |
| Authority Signals | 17 pts | Author bylines, About/Team pages, founding date, contact info. |
| FAQ & Q&A Format | 12 pts | Visible Q&A sections plus FAQPage schema with 5+ entries. |
| Crawler Accessibility | 8 pts | Viewport meta, canonical, no noindex, server-rendered text. |
Sites scoring 80+ are reliably cited by Perplexity and Google AI Overview for their core keywords. Sites under 50 are usually invisible to generative engines — regardless of their traditional SEO authority.
What happens during a scan
- Fetch — crawl & render. We fetch your page the way an AI crawler does — checking viewport, canonical, noindex, and server-rendered text.
- Parse — extract signals. Headings, factual density, definitions, tables, schema blocks, bylines, FAQ structure — everything the rubric weighs.
- Score — apply the rubric. Each category is scored against calibrated thresholds and live citation outcomes.
- Fix — rank the fixes. Issues are ranked by point impact and effort, with copy/paste schema and content rewrites ready to ship.
Pro Site Analysis goes further: it crawls your site from the homepage, analyzes up to 200 pages (larger sites are sampled), and runs live citation tests against Perplexity (sonar-pro).
The full technical write-up is published on our methodology page.