Technical Readiness

Technical readiness assesses whether AI crawlers can access and reliably process the website. It contributes 15% to the GEO score.

Composition

The technical score consists of three parts:

• Crawler access, 50%: Important AI crawlers are evaluated against robots.txt rules. Higher-priority crawlers carry more weight.
• Page technology, 35%: The average technical score across all analyzed URLs.
• llms.txt, 15%: The structure and completeness of an available llms.txt file.

A missing robots.txt means that access is unrestricted under the web standard. It therefore does not automatically reduce the score.

The tier of an AI crawler

The crawler table in the report lists a tier for every crawler. It states how strongly a block affects visibility:

• Tier 1 - feeds AI search products where people actively look for answers, such as GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot and PerplexityBot. Blocking these costs the most.
• Tier 2 - belongs to the wider AI ecosystem, such as Google-Extended, Applebot-Extended or Amazonbot. Access has an indirect effect.
• Tier 3 - pure training crawlers such as CCBot or Bytespider. They collect material for future models but do not answer questions. Blocking these barely costs visibility.

In the score, tier 1 counts triple, tier 2 double and tier 3 once. The crawler access value is the share of allowed crawlers weighted this way. So excluding training crawlers for legal reasons costs little - unlike a blanket block for all AI systems.

Technical page checks

Each page starts at 100. Deductions include:

• HTTP errors, noindex or almost no content in the raw HTML
• strong dependence on JavaScript for visible content
• very little page content
• a missing title, H1, canonical link, language declaration or meta description
• multiple H1 headings

llms.txt

The audit checks the title, description, useful sections, absolute URLs, entry descriptions, key facts and contact information. llms.txt is a supporting signal and does not replace accessible web pages or structured data.

How to improve the score

1. Allow required AI crawlers to access relevant content.
2. Provide essential content in the delivered HTML.
3. Resolve HTTP errors and unintended noindex directives.
4. Add clear titles, H1 headings, canonical links and language declarations.
5. Maintain an accurate and current llms.txt file.