GEO checker
Can AI assistants find, trust and cite your site?
What the GEO check covers93 scored checks in 6 groups and 17 suggestions, each with its source.
- Answer-engine bots allowed in robots.txt
- Content in the server HTML, no bot challenge
- Entity clarity: Organization schema, sameAs, one brand name
- Citability: dated content, sourced statistics
- Static checks only, no AI answers sampled
GEO checker results, explained
- Strong (85 to 100)
- AI assistants can fetch your pages, read them without JavaScript, tell who runs the site and find dated, sourced content worth citing.
- Ready with gaps (70 to 84)
- Access works, but entity or citation signals are thin: no Organization markup, mixed brand names, or statistics without a source.
- Important blockers (50 to 69)
- Part of the site is hard for AI to use. Often a CDN challenges some bots, or key pages render their text only in the browser.
- Likely to struggle (under 50)
- AI search crawlers are blocked in robots.txt or by the firewall, so assistants cannot read the pages they would cite.
- Not scored
- Practices the AI engines have not confirmed are kept as labelled suggestions, so you can weigh them yourself. They never move the score.
Common problems with generative engine optimization
Blocking the bots that cite you
A block-all-AI list in robots.txt includes OAI-SearchBot, Claude-SearchBot and PerplexityBot next to the training crawlers.
Fix: Decide on training separately. Allow the search and user-triggered crawlers if you want to be cited in AI answers.
Text that needs a browser
The article body loads from an API after the page starts. Several AI crawlers read the HTML without running scripts and see an empty shell.
Fix: Serve the main text in the first HTML response through server rendering or static generation.
An unclear entity
The site calls itself Acme, Acme Inc and AcmeHQ on different pages, and has no Organization markup linking to its official profiles.
Fix: Use one name, publish Organization JSON-LD on the homepage with url, logo and sameAs, and keep the about page consistent.
Claims without sources or dates
Statistics appear with no link to where they came from, and articles show no published or updated date.
Fix: Link each figure to its source and show a visible date that matches the Article markup.
Organization markup that leaves AI guessing, and markup that does not
sameAs ties the site to the same company elsewhere, which helps an assistant tell it apart from others with a similar name.
{
"@context": "https://schema.org",
"@type": "Organization",
"name": "AcmeHQ"
}{
"@context": "https://schema.org",
"@type": "Organization",
"name": "Acme",
"url": "https://acme.example/",
"logo": "https://acme.example/logo.png",
"sameAs": [
"https://www.linkedin.com/company/acme-example",
"https://github.com/acme-example"
]
}Questions about generative engine optimization
- Does the GEO checker ask ChatGPT or Perplexity about my brand?
- No. It runs static checks on your pages: access for AI crawlers, server HTML, entity markup and citability. It never sends prompts to an AI model.
- What is the difference between GEO and AEO?
- AEO is about being the direct answer in search features. GEO, generative engine optimization, is about being read, trusted and cited by AI assistants that write their own answers.
- Can I block AI training and still be cited?
- Yes. Training crawlers such as GPTBot and ClaudeBot have their own robots.txt names, separate from OAI-SearchBot, Claude-SearchBot and PerplexityBot, which fetch pages for answers.
- Why does the checker fetch my pages as several AI bots?
- CDNs and firewalls often treat bots differently from browsers. Fetching as Bingbot, Claude-SearchBot, PerplexityBot and OAI-SearchBot shows whether any of them gets a challenge, an error or a noindex the browser version lacks. The requests come from our servers, not the bots' own IP ranges, so treat a block as a lead to confirm in your CDN settings.