The agent successfully explained o11.ai's core offering, target market, and differentiation, but pricing information was incomplete and not directly sourced from the o11.ai site itself. The agent discovered what o11 does primarily through the Y Combinator launch post and external company profiles, and had to work around a JavaScript-rendered frontend that blocked content extraction from o11.ai's own blog and pricing pages. The site itself is poorly navigable for automated discovery—key pages exist but serve empty shells, forcing the agent to rely on third-party sources and prior knowledge to construct a complete answer.

• ›The Y Combinator launch post [11] was the most content-rich source, providing the core value proposition, use cases, and founder names. This was essential to understanding what o11 does and who it's for. The Welcome.AI company profile [10] and Tracxn profile [16] corroborated this but added little new detail.
• ›Pricing was never extracted from o11.ai's own site—the /pricing URL [1] returned only a JavaScript shell with no readable content. The agent cited pricing tiers (Free, Pro $20, Premium $200, Enterprise custom) but disclosed in the final response that these 'weren't prominently displayed on their website' and came from third-party sources not shown in the trajectory. This represents a critical gap in site navigability.
• ›Blog comparison articles [14, 15] (o11 vs. Copilot, alternatives to Copilot in Excel) were fetched but returned only JavaScript shells—the agent could not extract their content, limiting its ability to ground the differentiation section in primary sources from the site itself.
• ›The agent used web search 42% of the time, relying heavily on external discovery because the o11.ai site does not publish machine-readable, static content. The homepage and key pages are React/Svelte SPAs that require JavaScript rendering—a structural impediment to agent navigation.
• ›The /demo, /solutions, and /blog pages all returned 200 status but were similarly empty shells, suggesting the entire site is a client-side rendered application with no server-side content. This is the primary cause of friction—the site exists but is not agent-readable.