GEO Audit

Optimized for Google. Quotable by machines?

A GEO audit measures whether generative engines can access your pages, lift your passages, and attribute your claims. Essel scores it deterministically against your real site — GEO, AEO, and AI-crawler access are about a quarter of your total score — with a fix for every finding. Free for any page, no sign-up.

0% of the score is AI-native
Both GEO + AEO scored
Crawl site-wide citability
Free single page, no sign-up

From unquotable to cited.

Three steps, measured at each one.

01

Paste any page

The free scan scores GEO, AEO, and AI-discoverability for one URL — every finding and fix visible.

02
72

Read the GEO verdict

Passage citability, answer structure, citable evidence, crawler access — measured, not vibes.

03
1
2
3

Go site-wide for $10

Citability coverage across the crawl, site-wide crawler gating, and generated llms.txt + JSON-LD.

Coverage

What a GEO audit actually measures.

Built on the measured levers — passage structure, citable evidence, attributable identity — not the folklore. The report is honest about what doesn't move the needle, including llms.txt.

GEO — can engines lift your content?

5/5 passed

Passage citability

whether your sections sit in the word window a generative engine can quote whole

Citable evidence

quantified claims — percentages, figures, dates — a model will attribute to you

Publisher identity

Organization and Person schema plus a named author, the attributable identity engines cite

Heading outline

the clean hierarchy a model uses to map your document

Site-level citability

what share of your whole site carries question headings, lists, and tables — measured across the crawl

AEO — can engines answer with you?

4/4 passed

Question-shaped headings

headings phrased as the questions people actually type

Answer-first openings

direct answers sized for lifting, with preamble and hedging openers flagged

Extractable structure

lists and tables — the shapes answer engines quote verbatim

FAQ & QA schema

FAQPage and QAPage markup, scored honestly as opportunity, not as a retired rich-result win

The access layer underneath

4/4 passed

AI crawler access

17+ AI crawlers checked against your robots.txt and your edge's actual behavior

Server-rendered content

generative engines read HTML, not your JavaScript bundle

Snippet eligibility

directives that silently exclude you from AI Overviews

Freshness signals

machine-readable dates, scored by age, not mere presence

One score. Five pillars.

Here's what a real report looks like.

Your shape

TechnicalAI readinessContentPerformanceAuthority
0 / 100 · C

Strong technical foundations, thin where AI engines look.

Your scores

Technical & crawl /100

Search engines can crawl and index it

AI readiness /100

AI engines can find, read, and cite it

Content /100

Clear, complete, well-structured

Performance /100

How fast it loads and responds

Authority /100

Web presence and links

What AI thinks of you

Ask an AI assistant about your site and it says:

A free AI resume builder that imports LinkedIn profiles and turns them into ATS-friendly resumes in minutes.

AI resume builder For job seekers and career changers Somewhat known

Search foundations vs AI readiness

11 categories · renormalized 0–100
0
Search foundations F grade
Indexability & technical –/20
Title & meta –/15
Headings & structure –/10
Images & accessibility –/10
Content quality & E-E-A-T –/15
Core Web Vitals –/15
Authority & web presence –/10
0
AI readiness F grade
Structured data –/15
AI discoverability –/12
AI search (GEO) –/10
Answer engines (AEO) –/5

Page speed

Perf –
Loading
Responsiveness
Visual stability
Interactivity

Authority & web presence

Web presence
Internal links
Outbound links

Fix these first

Highest-impact, with the exact steps
Move structured data into the <head> Your schema sits in <body>. AI crawlers that don't run JavaScript often skip it entirely.
Open with a one-line definition A direct answer up top is what ChatGPT, Perplexity, and AI Overviews lift as the source.
Add question-and-answer sections Mirror how people ask. Answer engines quote Q&A blocks directly.
Add alt text to 6 images Improves accessibility and gives AI more to read from the page.

Also in your report

What AI thinksSite modelCompetitor benchmarksContent AI readsUnderstanding curveEvery check, explained

A report tells you what's broken. Essel also writes what's missing.

The same engine behind this audit researches, writes, and publishes GEO-optimized articles for your site — on a cadence, autonomously.

Start free
SEO, AI search (GEO) and answer engines (AEO) The free score takes seconds, no sign-up

The comparison

Deterministic audit vs. GEO consulting.

GEO consulting is young and methodologies vary wildly. A deterministic audit gives you the same checks, the same weights, and a re-run button.

Essel site audit

$0

in the whole-site audit · free for one page

  • Scope: GEO + AEO + access, scored per page and site-wide
  • Method: Deterministic checks on your real crawled pages
  • Evidence base: Built on measured levers (passage structure, quantified claims, identity)
  • Price: In the $10 whole-site audit
  • Output: Findings with pasteable fixes, weighted ~26% of your total score

GEO consulting

$1,000+

per engagement · methodology varies

  • Scope: Usually a content-marketing add-on
  • Method: Manual review, methodology varies
  • Evidence base: Often optimizes unmeasured signals
  • Price: $1,000+ engagements
  • Output: Slide deck

GEO, AEO, and AI-discoverability carry 37 of 142 possible points in Essel's scoring model — roughly 26% of the total, reflecting where discovery is actually shifting.

FAQ

GEO, explained.

The questions people ask before optimizing for generative engines.

A GEO audit measures whether generative engines — ChatGPT, Perplexity, Google's AI Overviews — can access, understand, and cite your site when composing answers. It checks the access layer (AI-crawler permissions, server-rendered content), the structure layer (passages sized for quoting, question headings, direct answers), and the evidence layer (quantified claims and an attributable identity worth citing). Essel scores all of it per page and across your site, with a fix for every gap.

SEO optimizes for ranked lists of links; GEO optimizes for being quoted inside a composed answer. The foundations overlap — crawlability, structure, authority — but GEO adds requirements classic SEO never tested: passages an engine can lift whole, claims specific enough to attribute, an identity to attribute them to, and access rules for AI crawlers that robots.txt files written in 2015 silently block. Essel's audit scores both in one report; the AI-native categories carry roughly a quarter of the total.

The measured levers are structural: sections sized so an engine can quote them, answer-first openings under question-shaped headings, quantified claims with a named author and organization behind them. The commonly oversold levers are llms.txt and schema markup on their own — publishing either without fixing structure moves little, which is why Essel's audit weights llms.txt near zero and says so in the report. An honest audit tells you where the weight actually is.

Answer-engine optimization is its own scored category in the audit, separate from GEO. AEO asks a narrower question: when someone types a question your page answers, can an answer engine lift the answer directly? The audit checks question-shaped headings, answer-first paragraph structure, extractable lists and tables, and FAQ markup — and flags hedging openers that bury the answer a paragraph deep.

The single-page GEO audit is free: paste a URL and the scan scores GEO, AEO, and AI-discoverability for that page with every finding and fix visible, no sign-up. The $10 whole-site audit runs the same checks across up to 100 crawled pages and adds the site-level view — citability coverage, site-wide crawler gating, and generated artifacts like llms.txt and JSON-LD blocks.

See your GEO score in the next minute.

Paste a URL and get the generative-engine verdict for that page free — citability, answer structure, crawler access, and every fix. Whole site, $10.