Asked of a language model without live browsing — what the model has absorbed about this market, not a search result fetched today.
Put to a live web search — the pages an assistant reads before answering. Absent from these results it cannot cite you; present, it still might not.
Le point fort majeur est l'AI crawlability avec un score de 98/100, garantissant que les modèles de langage peuvent lire intégralement chaque page. Le principal frein est le GEO à 0/100, indiquant une incapacité totale à fournir des passages citables aux IA, ce qui fait perdre des opportunités de visibilité éducative. La mesure la plus efficace consiste à rédiger des passages autonomes et citables, en ajoutant des données structurées FAQPage et des dates publiées lisibles par machine, afin d'améliorer immédiatement le GEO et l'AEO.
Ask it what a customer would ask you. It answers only with what is written on the pages of croix-rouge.fr.
Hello! I’m the assistant for Croix-Rouge française. What would you like to know?
A Tadaaah demo. This is an AI assistant: it can make mistakes.
$29 CADa month, on top of PRO
No paragraph in the ~100-200 word "citable passage" range was found; AI engines prefer to lift complete, context-free passages.
Do: Restructure key sections into standalone paragraphs (~130-170 words) that make sense without surrounding context.
No FAQPage/QAPage JSON-LD with answered questions was found on the page.
Do: Wrap your most-asked questions in FAQPage JSON-LD with the full question and answer text, matching what a visitor can actually read on the page.
No outbound links to authoritative domains (.gov/.edu, research, major publishers) were found.
Do: Support factual claims with links to primary, authoritative sources.
No <time datetime> or article:published_time/modified_time meta was found.
Do: Add datePublished/dateModified to Article schema and a visible "Last updated" date.
Automated snapshot generated by Tadaaah from public data. Re-run the audit for the latest version.