The agent gathered a partial understanding of eCréa by fetching the homepage and a demo WordPress subdomain, but hit significant friction discovering key details about pricing, service scope, and training programs. The site is not agent-readable for comprehensive evaluation: pricing is intentionally hidden behind contact forms, service details are scattered or absent, and standard navigation paths (tarifs, services, formations) all returned 404s, forcing reliance on search and inference rather than published information.

• ›Step [1] (homepage) contained high-level positioning (webdesign, formations, Bordeaux) via meta tags and title, but the actual page content was truncated and did not include details on pricing model, service packages, or differentiators. Step [17] (demowp.ecrea.fr) was similarly truncated and revealed only that they use WordPress for training/demo, not their core offering.
• ›The agent attempted direct navigation to /services, /tarifs, and /formations (steps [3], [4], [5]) and received 404s on all three—standard information architecture is missing or broken, indicating the site does not surface core business info via standard URL paths. The agent then relied on web search to find references to the same homepage repeatedly, confirming the site has minimal discoverable structure.
• ›Pricing and service scope were entirely absent from what the agent could fetch. The agent inferred a 'custom quote' model from meta descriptions and search results, but never retrieved actual pricing information, service lists, or training course details in machine-readable form. The site appears intentionally opaque for automated evaluation—no FAQ, no pricing page, no service matrix, no structured data (schema.org).
• ›The agent successfully identified the founder (Raphaël Lang) and geographic focus (Bordeaux) from homepage metadata and search results, and inferred a boutique positioning from the single-founder profile, but could not verify differentiators, competitor analysis, or training certification details from the site itself—all were assembled through search and prior knowledge.
• ›The site is not agent-ready: it blocks standard navigation paths with 404s, publishes no pricing or service taxonomy, contains no client testimonials or case studies accessible via fetch, and provides no structured contact mechanism or CTA in the fetched content. The agent had to bridge the gap with 33% web search and 17% prior knowledge.