Public Scoring Methodology

How GEO Scoring calculates your AI visibility score

Name: GEO Scoring
Author: GEO Scoring

Our model is built to explain why a website is easy or hard for AI systems to extract, trust, and cite. The final score combines weighted dimensions, blocker penalties, and a small set of AI-facing bonus signals into a single 100-point framework.

Run a free scan Read scoring FAQ

Model

100 points

A weighted score built for AI search and answer-engine visibility.

Dimensions

Report categories spanning structured data, content, authority, and readability.

Adjustments

Penalties + Bonuses

Critical blockers reduce the score, while select AI-facing implementation patterns can add a small lift.

Output

Grade + fixes

Every report pairs a final grade with prioritized remediation guidance.

Weighted dimensions

Each dimension measures a different part of AI citation readiness. Higher-impact areas receive more weight.

Total weight: 100 points

Schema Tier-1 Coverage

14 pts

Verify the core Schema types that match the detected website type while keeping this category worth 14 points.

Personal sites: Person + Organization + Article/BlogPosting (Product, FAQ, HowTo, ItemList are not required and their weight is moved to the remaining types)
Commercial sites: Organization + Product/Offer + FAQ + Article + HowTo + ItemList
SaaS / corporate / blog sites: Organization + content schemas relevant to the site (FAQ, Article, HowTo, ItemList), with non-applicable types reallocated

Schema Attribute Completeness

6 pts

Check whether the applicable Schema entities include the attributes AI systems need to trust them.

Organization or Person identity fields according to site type
Product attributes only for sites that sell products or offers
Article freshness and author fields for non-personal content sites
@id references across separate <script> blocks and @graph containers count as present attributes — multi-script JSON-LD is treated the same as a single inline entity

Schema Nesting Structure

5 pts

Assess whether applicable Schema entities connect into a coherent graph instead of isolated snippets.

Product to Organization links for commerce sites
Article to publisher/author links — personal sites are checked against Article → Person specifically
Cross-script @id references (e.g. one <script> declares Person with @id, another references it from Article.author) and @graph arrays are resolved before scoring, so multi-script sites get the same credit as a single inline entity
Multiple complementary Schema types with non-applicable checks removed and points reallocated

Open Graph & Social Metadata

5 pts

Ensure shared URLs are understandable across crawlers and downstream AI systems.

og:title / og:description
Image availability
Share-ready page metadata

Content Richness & Freshness

20 pts

Reward useful content assets and freshness signals without overvaluing raw page count. The 20 points are split across FAQ, Blog, Knowledge pages, and Freshness, with weights redistributed per website type so the total always reaches 20.

FAQ answer quality and Schema coverage — counted only for site types that publish FAQs (commerce / SaaS / corporate / blog / unknown)
Core content pages with Article Schema — content URLs like /columns/ , /news/ , /press/ also count as article pages, and pages without a keyword URL fall back to Article-schema detection
HowTo / technical documentation / about pages — counted only when applicable to the site type
datePublished, dateModified, and recent-update signals — coverage is measured against Article-class schemas (Article / BlogPosting / NewsArticle), not against every WebSite or WebPage entry
Personal sites: FAQ and HowTo/Tech points are moved into Blog (8) and Freshness (12) since solo creators rarely run those formats

Knowledge Graph Authority

15 pts

Assess brand/entity signals that make your site easier to trust and cite.

sameAs links and knowsAbout areas
Founding date and entity history
External authority references
Internal entity links and location signals

Content Citability & Absorption

20 pts

Measure whether content can be extracted as definitions, data, comparisons, steps, and direct answers.

Self-contained content blocks
Definition blocks that explain concepts clearly
Data blocks with percentages, amounts, years, counts, or measurable facts
Comparison and step blocks
Direct answers plus summary, table, list, and FAQ answer assets

E-E-A-T Signals

10 pts

Check for expertise and trust markers that support high-confidence citations. Sub-weights are reallocated per website type so every site type can reach a full 10.

Author Schema and a visible byline — the byline check matches HTML class/rel/itemprop hints AND, as a fallback, looks for the Schema-resolved author name in the page content (so framework sites without semantic byline classes still get credit)
Required trust pages depend on site type: personal = contact only; corporate / blog = privacy + contact; commerce / SaaS / unknown = privacy + contact + terms — points for non-required trust pages are reallocated to expertise/about so the category sums to 10 either way
Experience and expertise cues — credentials, founding history, professional bios
About page presence (weighted higher for personal sites)

AI Readability

5 pts

Look for clean HTML, sensible semantic structure, and machine-readable layout.

Semantic heading hierarchy
Readable HTML structure
Low-friction parsing

Website-type aware scoring

GEO analysis first classifies the site as personal, commerce, SaaS/software, corporate/service, blog/content, or unknown. Three categories then redistribute their sub-weights based on that classification so every site type can reach the full module total: Schema (14 + 6 + 5), E-E-A-T (10), and Content Richness (20). When a check is not applicable for the site type, its points are reallocated across the remaining applicable checks as whole numbers instead of being silently lost.

Personal / creator

Schema

Person + Organization + Article/BlogPosting. Article → Person nesting is the primary structural check.

E-E-A-T

Trust pages: contact only. Author Schema + visible byline still required, but expertise (3) and About (2) carry more weight.

Content Richness

FAQ and HowTo/Tech docs do not apply. Their weight moves to Blog (8) and Freshness (12) so the 20-point total is reachable.

Commerce

Schema

Organization, Product/Offer, FAQ, Article, HowTo, ItemList. Product → Organization nesting is checked.

E-E-A-T

Trust pages: privacy + contact + terms all required and counted.

Content Richness

Default split — FAQ (6), Blog (4), Knowledge (4), Freshness (6).

SaaS / software

Schema

Organization plus SoftwareApplication/WebApplication, FAQ, Article, HowTo, ItemList. Product Schema is only required when the site has a catalog.

E-E-A-T

Trust pages: privacy + contact + terms.

Content Richness

Default split — FAQ (6), Blog (4), Knowledge (4), Freshness (6).

Corporate / service

Schema

Organization, Article, FAQ, ItemList. Product and HowTo are skipped unless detected.

E-E-A-T

Trust pages: privacy + contact (terms not required). The unused terms weight is reallocated to expertise (up to 3).

Content Richness

Default split — FAQ (6), Blog (4), Knowledge (4), Freshness (6).

Blog / content

Schema

Organization, Article/BlogPosting, FAQ, ItemList. Product checks skipped.

E-E-A-T

Trust pages: privacy + contact. Terms weight reallocated to expertise (up to 3).

Content Richness

Default split — FAQ (6), Blog (4), Knowledge (4), Freshness (6).

How the final score is produced

We do not stop at a raw weighted sum. The final grade reflects both what your site does well and whether there are hard blockers that prevent AI systems from consuming the content reliably.

Scan

We fetch the submitted site, inspect page structure, detect rendering mode, and collect machine-readable signals.

Score

Each dimension receives a weighted score based on the evidence we find on the site and in the page markup.

Adjust

We subtract penalties for hard blockers (blocked AI crawlers, missing HTTPS, redirect issues) and add bonuses for proactive AI-facing signals (llms.txt, ItemList Schema). Bonus is capped at +10, and the final score is clamped to [0, 100] — bonuses act as a recovery buffer rather than an extension above 100.

Recommend

The report surfaces fixes that map back to the remaining points in each category. Suggestion point values are scaled to the site type so the sum of "current score + applied suggestions" reaches the module total — never leaving unreachable points on the table.

Penalty signals

Subtract

Major AI crawlers blocked in robots.txt

-15

Blocking GPTBot, ClaudeBot, PerplexityBot or similar AI crawlers suppresses visibility regardless of content quality.

Missing HTTPS

-10

Trust and accessibility signals drop when the site is not served securely.

HTTP does not redirect to HTTPS

-3

Inconsistent canonical delivery weakens consolidation and crawl confidence.

Redirect chain too long

-3

More than a few hops to reach the final URL slows crawlers and dilutes signal attribution.

Bonus signals

Add

llms.txt file at site root

An llms.txt file at /llms.txt makes high-value resources easier for AI systems to discover and prioritize.

ItemList Schema structured data

ItemList Schema (rankings, top-N lists, structured collections) improves synthesis for list-style AI answers.

Total bonus is capped at +10, and the final score after applying both penalties and bonuses is clamped to [0, 100]. A site that already scores 100 on the base 100 will not exceed 100 — bonuses act as a recovery buffer against penalties or partially earned dimensions.

Grade bands

Your final letter grade summarizes both weighted performance and adjustment effects.