kpmg.com
success
Claude Code
intent

What does kpmg.com do and who is it for? Explain it back to me.

15steps
64.3sduration
$0.9594cost
157,450tokens
18 steps3 reasoning steps6 searches
home
/us/en/home
/us/en/who-we-are
/us/en/services
/us/en/home/services/a…
/us/en/home/about.html
/us/en/home/careers.html
/us/en/insights.html
search
search
/xx/en/what-we-do/serv…
search
search
search
search
100%
on-site discovery
78%
reliability
67%
link following
path origin
  • previous resource67%
  • prior knowledge33%
insight

The agent assembled a comprehensive explanation of KPMG by relying almost entirely on web search results and prior knowledge, because KPMG's website itself does not transparently publish what it does, who it serves, pricing, or competitive differentiation. The agent successfully answered all four facets of the task (what they do, who it's for, pricing, differentiation), but had to work around opaque site navigation and heavy JavaScript rendering to do so. The website proved poorly structured for discovery—the agent could not extract core business information from direct site fetches and instead sourced nearly all substantive facts from external research.

  • ›Direct site fetches (steps 1–8) returned only HTML scaffolding and navigation chrome with no actual content visible in the truncated responses. Steps 2, 4, 5, 6, 8 returned 200 status but no rendered service or positioning information, forcing the agent to pivot to web search.
  • ›All substantive information in the final answer—the three service lines and their revenue breakdown, Big Four comparisons, regulatory/financial services specialization, pricing model, and KPMG's niche differentiation—came from web search results (steps 10, 11, 12, 13, 15, 16), not from kpmg.com itself. The agent cited external sources (roadtooffer.com, ie.edu, FourWeekMBA, Statista, goingconcern.com) in its sources list, not KPMG's own pages.
  • ›Step 14 fetched the KPMG International services page (kpmg.com/xx/en/what-we-do/services.html) returned 200 but was not cited in the final response and did not appear to contain extractable service listings in the truncated response. This suggests the content is either client-side rendered (inaccessible to the agent's fetch mechanism) or poorly structured.
  • ›The agent correctly identified website UX issues in its 'What Was Confusing' section: heavy JavaScript, opaque regional site structure, and lack of transparent service boundaries. The site is oriented toward internal navigation and does not expose core business positioning in machine-readable or easily fetchable form.
  • ›No pricing information was found on the site (steps 11 searched for pricing; results pointed to external consultant review sites, not KPMG's own pricing page). The agent had to infer pricing from Big Four industry norms and analogues.
  • ›The site does not appear to have a structured 'About' or 'Who We Serve' section that surface client segments, industries, or service line descriptions. Step 3 (who-we-are) returned 404, and step 5 (about.html) returned 200 but yielded no visible content in the response snippet.

Want to run your own?

Join the waitlist for early access to point your own agents at any domain, with the intents you choose.

or talk to us about agent readiness →