What is Claude compound trust multiplier?
Claude applies compounding trust to JSON-LD structured data. 1 schema type is baseline. 2 types give a small boost. 3 types cross a trust threshold. 4+ types create compounding trust. Sites with Organization plus FAQPage plus Article plus BreadcrumbList get exponentially higher Claude scores than sites with Organization alone. Tidio hit this threshold — that's part of why it scored 63.
Why does Claude penalize missing llms.txt?
Claude evaluates llms.txt on a four-level quality rubric: basic existence, structured URLs, detailed descriptions, and llms-full.txt link. No llms.txt means a trust penalty — Claude interprets it as a signal you haven't considered AI accessibility. Sites with high-quality llms.txt files get proportionally higher Claude scores.
How does ClaudeBot directive affect Claude scores?
Claude applies a specific trust penalty when ClaudeBot isn't explicitly mentioned in robots.txt — even if the site allows other crawlers like GPTBot. Explicitly allowing ClaudeBot signals willingness to be cited by Claude. Block ClaudeBot but allow GPTBot? Reduced Claude citation rates.
How does Claude handle entity disambiguation?
Claude cross-references Organization schema, sameAs links, address data, and domain registration signals to distinguish similarly-named businesses. Without clear entity signals, Claude will confuse you with competitors or refuse to cite you to avoid inaccuracy. HelpSquad's Claude score of 42 — 5 points below its ChatGPT score of 47 — was partly driven by thin entity signals.
What content licensing signals does Claude look for?
Claude evaluates licensing to determine citation willingness. Clear CreativeCommons licenses, TDM Reservation Protocol headers, or ai.txt permissions explicitly granting citation rights increase citation rates. Ambiguous or missing licensing makes Claude more conservative about quoting content directly.
How does Claude evaluate semantic HTML hierarchy?
Claude performs deeper semantic HTML analysis than other engines. It evaluates strict H1→H2→H3 nesting, ARIA landmark roles, semantic sectioning elements (main, article, section, nav), and content-to-boilerplate ratio. Pages with clean semantic hierarchy get a measurable Claude scoring boost.
What are citation-ready fact blocks for Claude?
Claude preferentially cites content structured as fact blocks: a named statistic or claim, followed by its source, followed by context. A specific number, company attribution, date, market context. These match Claude's internal reasoning patterns and get extracted at higher rates than narrative prose.
Why did Tidio score higher on Claude than expected?
Tidio got a +14 Claude bonus over what technical metrics alone would predict. Why? A 251-line llms.txt file. 4+ JSON-LD schema types triggering compound trust. Explicit ClaudeBot allow in robots.txt. Clean semantic HTML hierarchy. Claude's governance-first evaluation rewards sites that invest in machine-readable signals — and Tidio nailed all of them.