Inside a decision-grade audit: a regional P&C carrier, anonymized
The subject is one of the 38 regional P&C carriers in the US Insurance AI Discovery Index — Edition 1 cohort (June 2026) that never appeared in a name-free ChatGPT or Perplexity buyer answer — shown here with its identity removed, because negative results are private to the measured organization.
The finding
On all measured name-free buyer questions — "best business insurance provider for mid-size businesses" and its variants — ChatGPT and Perplexity answered with national brands (The Hartford, Travelers, Chubb, Nationwide) and never named the subject. Branded questions ("what does [carrier] do") were answered correctly: the models know the carrier exists; they don't select it. That distinction is the audit's core: this is a selection problem, not an awareness problem.
Why competitors win the answer
The displacement analysis showed the same few national brands winning every question, supported by the sources the engines cited: best-of listicles, review directories, and industry press where the subject has no footprint. The subject's own product pages were not extractable as direct answers to any measured question — long marketing copy, no question-shaped headings, no FAQ or Organization structured data.
What the roadmap prescribed
Five recommendation groups, each labeled with confidence and an expected mechanism: (1) entity foundation across profiles and Organization schema — best practice; (2) one answer-first page per displaced buyer question — inferred from the displacement evidence; (3) citation-ready placements in the surfaces the engines actually cited — observed; (4) crawler/extractability fixes — best practice; (5) a locked baseline and pre-agreed re-measurement windows — the only honest way to attribute movement.
What we did not claim
No revenue projection, no guaranteed mention, no promise about a third-party model's future output. The engagement's guarantee is the deliverable itself: contracted artifacts, on time, traceable to evidence — with a re-measurement that tells the truth either way.
This is what your audit looks like — with your name on it.
Fixed scope, $3,500–$7,500, ten business days, every finding traceable to a stored answer.
Request a scoped audit