Where US airlines produced a settled top three, industrial machinery fractures: the five engines overlapped just 0.34 (airlines: 0.64), agreed on the #1 only a third of the time, and filled most of the list with companies that aren't American. This category is molten — the answer hasn't set.
Hover or tap any engine cell above to see the real prompts behind that number.
Scope & caveats. Rows show US companies (the index roster); the foreign firms the engines surfaced are reported in the finding above, not charted. This cut now covers all five engines (ChatGPT, Claude, Gemini, Perplexity, Grok), web-grounded, with company names canonicalized from auto-extracted output.
REAL pilot capture — ChatGPT, Claude, Gemini, Perplexity, Grok, all web-grounded, 6 prompts × 3 runs, 2026-05-27. Company names canonicalized from auto-extracted output. The Answerability Index · pilot.
Run the same instrument on industrial machinery and the airlines structure collapses. Overlap falls by nearly half (0.34 vs 0.64), only Caterpillar is a true consensus pick, and the engines disagree run-to-run even with themselves. The category's answer hasn't set — it's molten.
When no canonical answer exists, the engine improvises from whatever it can retrieve. That's why "leading US industrial companies" fills with Komatsu, Siemens, ABB, and Schneider: the global industrial corpus is far denser than the US-specific one, so the engines retrieve on topical density and quietly drop the "US" qualifier. It isn't a fact about who's American — it's a retrieval failure you can watch happen in the data.
In a molten category, the lower rungs of the ladder decide everything. With corroboration unsaturated, surfacing is up for grabs — and it goes to whoever is most retrievable (an AI crawler can reach and parse you), most cleanly resolved as an entity, and most answer-shaped for the buyer's question. Those are precisely the things a company controls — and precisely what a diagnostic measures.
Read it through the retrieval surface. Because the surface hasn't solidified, the work is to define your content territory before it sets — publish the answers your buyers are actually asking, make them retrievable and entity-clear, and you stake a claim while the field is still forming. In a frozen market the retrieval surface is already drawn and defended; in a molten one like this, it's being drawn right now — and whoever maps it first tends to get named when it hardens.
Hover any cell in the grid above to see the real prompts behind it — which buyer questions that engine surfaced the company on, and at what rank. That per-cell evidence is the retrieval surface, made legible.