A 700-question LLM buying simulation reveals the #3 semiconductor GC in America barely exists in the AI layer now shaping $80–120B in procurement decisions.
When procurement committees ask AI for contractor recommendations, JE Dunn barely registers — while Turner Construction dominates the conversation.
We simulated a complete semiconductor fab procurement committee — 7 stakeholder personas, 100 questions each — through leading LLM platforms.
Every stakeholder on the committee encounters a different version of the gap. Priya Patel (Procurement) sees Turner in 42% of her responses while JE Dunn's carry 69% confidentiality friction. She's the gatekeeper for shortlist inclusion — and the AI is giving her every reason to pick Turner.
Across all seven personas, the pattern is consistent: Turner wins on source diversity (15+ unique third-party citations vs. 8 for JE Dunn), named project references, and clean recommendation language. JE Dunn wins on only 1 of 18 competitive visibility metrics.
JE Dunn's strongest evidence — the actual project work, the client relationships, the track record — lives behind confidentiality agreements. The AI can't access it. What the AI can access is a thin layer of first-party content that it discounts relative to the diverse, third-party citation ecosystem Turner has built.
This isn't a quality problem. It's a signal problem. The substance is there. The signal isn't.
The gap between what the #3 semiconductor GC should capture and what an invisible #3 will actually win.
A phased program targeting the specific mechanisms creating JE Dunn's visibility deficit — starting with actions JE Dunn controls entirely.
Transform existing project evidence into LLM-visible content. Named case studies, attributed project data, client-approved references. Targets the 57.6% confidentiality friction directly.
Build third-party validation through trade media, industry associations, and analyst relationships. Move beyond first-party content (36.1% of JE Dunn's sources) to the citation diversity LLMs weight most.
Optimize JE Dunn's digital infrastructure for LLM ingestion — structured data, semantic markup, and content architecture designed for AI systems, not just human readers.
This executive summary covers the headlines. The full assessment maps every stakeholder, every metric, every competitive gap — and the specific actions to close them.
Rick Nash, CEO · rick.nash@spotlightar.com