Fix guides: how to resolve what the scan flags
The free scan grades your site against the checks that decide whether AI answer engines (ChatGPT, Claude, Perplexity, Google AI Mode) can crawl, understand, and cite you. When a check fails, it links here.
Each guide explains three things: what the check looks at, why it matters for AI search visibility, and exactly how to fix it with real code you can paste. They are written for the person who has to make the change, not for a strategy deck.
Can the engines even read the page?
Section titled “Can the engines even read the page?”These are the render-time floor. If they fail, nothing else you do matters, because the bot never sees your content.
- Returns HTTP 200 to AI crawlers (T1)
- Answer content in raw HTML, no JavaScript (T2)
- Server-rendered HTML on the homepage (S1)
- JSON-LD present in raw HTML (T3)
- Indexable, no noindex (T5)
Crawler access: robots.txt, WAF, llms.txt
Section titled “Crawler access: robots.txt, WAF, llms.txt”- robots.txt sanity, the four never-dos (S2)
- Allow search and user-triggered AI bots (R2)
- The bot-wall: WAF blocking crawlers robots.txt allows (R5)
- Publish an llms.txt (L1)
- Author the llms.txt, don’t dump a sitemap (L2)
Content structure for extraction
Section titled “Content structure for extraction”- Open with a 40 to 60 word answer block (C1)
- Make your H2s question-shaped (C3)
- Add a comparison table (C4)
- Keep paragraphs under ~150 words (C13)
- Add an FAQ section (C16)
- Match FAQ schema to the visible text (C17)
Entity and structured data
Section titled “Entity and structured data”- Ship site-wide Organization schema (E1)
- Add a Wikidata URI to sameAs (E3)
- Build a 5+ entry sameAs mesh (E5)