The agent successfully assembled a comprehensive overview of Ardoq by combining direct site navigation with targeted web searches. The site itself is moderately agent-friendly but fragmented: homepage, about, solutions, features, and pricing pages are accessible and cited, but they render largely as JavaScript-heavy templates that don't expose detailed product positioning or pricing tiers in machine-readable form. The agent filled gaps by searching for competitor comparisons and use-case information, then synthesizing across sources to explain what Ardoq does, who it serves, and how it differs—though it correctly flagged that concrete pricing numbers are deliberately withheld from the public website.

• ›The site's main pages (/, /about, /solutions, /features, /pricing, /plans) returned HTML but with heavy obfuscation via JavaScript (evident from RRRR ASCII art placeholders in responses), making content extraction unreliable. The agent had to rely on search results and blog content to extract meaningful details.
• ›Specific pricing information was absent from the public site; the /pricing and /plans pages do not list actual costs, only frameworks (app-based tiers, add-ons, unlimited users). The agent correctly identified this and noted sales contact is required—a common enterprise software pattern but poor for agent discoverability.
• ›The agent found actionable product details (graph technology, AI agents, scenario modeling, use cases, target industries, customer examples) scattered across /features, /solutions, /about, and especially the blog post 'What Is Ardoq' and /platform-overview, requiring synthesis across multiple pages rather than a single product spec.
• ›Web search proved essential: competitor comparisons came from external sources (Catio, Gartner, Archilu), not from Ardoq's own site. The site does not publish a clear competitive positioning or alternatives comparison.
• ›The /customers page and blog post confirmed the 400+ customer count and provided industry context (financial services, insurance, utilities), but this social proof is not programmatically structured on the site.
• ›Two naive URL guesses (/product, /platform) returned 404s, indicating the site's navigation structure is not intuitive for agent exploration—the agent had to rely on search and prior knowledge to find the right paths.