The agent assembled a comprehensive understanding of doola by working around a JavaScript-heavy main site that did not return readable content. It successfully retrieved actionable information from a help center article and a third-party review guide, then supplemented with prior knowledge to answer what doola does, who it's for, how it's priced, and how it compares to alternatives. The site's own domain was largely inaccessible to the agent; doola.com's core content required external sources or cached knowledge to explain.

• ›Step [9] (help.doola.com) returned structured content explaining core services (LLC/C Corp formation, bookkeeping, tax compliance, e-commerce analytics) — this was the only first-party source that yielded readable product information.
• ›Step [12] (headwestguide.com third-party review) provided pricing tiers, competitive comparisons, and clarifications on annual subscription vs. one-time fees — information the agent could not extract from doola.com itself.
• ›Steps [1], [2], [4] fetched doola.com's main pages but returned only HTML scaffolding with truncated JavaScript, making the site unreadable to the agent; the agent had to rely on web search results and prior knowledge for 50% of its sourcing.
• ›The agent correctly identified key confusions (state fees not included in advertised pricing, bookkeeping scope unclear, long-term cost transparency) by cross-referencing multiple external sources — doola.com itself did not surface these caveats prominently.
• ›Site navigation was poor for agent discovery: direct URL guesses for /features and /platform returned 404s, forcing the agent to pivot to search and external resources rather than following an internal information hierarchy.