60% Outcome (does AI recommend you) + 40% Readiness (is your site built for AI).
Outcome — does AI actually recommend you
Cross-provider AI visibility
simulated — add model keys for live measurement
Outcome
Mention rate
36%
CI 32–40 · n=648
Recommendation
6%
CI 5–8 · n=648
Citation rate
12%
CI 10–15 · n=648
Mention rate by provider
Share of voice vs competitors
How you surface
66% parametric · 34% retrieved
Retrieved is fixable now (content & citations); parametric is the model's trained memory.
Sentiment of mentions
+210 / ~22 / −0
Across 232 mentions.
Readiness — is your site built to be found
Turmeaus Cigars
Emerging
26/82 pts
- —Sparse semantic HTML (missing main/header/nav/article landmarks)
- —No canonical URL on homepage
- —No lang attribute on <html>
- —Expected exactly one <h1>, found 0
- —No llms.txt (speculative signal)
- —No JSON-LD structured data found
- —Missing Organization schema
- —Missing Product schema
- —Missing BreadcrumbList schema
- —No FAQPage schema
- —No LocalBusiness/Store schema
- —No named Organization entity in structured data
- —No sameAs / social profile links (weak entity graph)
- —No postal address in structured data
- —No telephone / contact point in structured data
- —No About page found
- —No Contact page found
- —No FAQ / help content found
- —No blog / guides / editorial content found
- —No shipping/returns information page found
- —No current copyright year in footer
- —No OpenAPI spec (/openapi.json)
- —No AI plugin manifest (/.well-known/ai-plugin.json)
- —No link to API / developer docs
Remediation — how to fix it
23 prioritised fixes
7 quick wins · ranked by impact
Turn mentions into recommendations
AI mentions you but rarely picks you first. That's a differentiation and proof gap — the model has no strong reason to put you at the top.
- 1.Earn ratings and reviews — volume and recency signal quality to AI.
- 2.Get positioned as a top choice in comparison and 'best of' lists.
- 3.Publish specific reasons to choose you (specialisms, provenance, guarantees) that AI can quote.
Earn citations to enter the models' memory
Your mentions come almost entirely from the models' trained memory, and there's little live retrieval picking you up. The durable fix is becoming widely cited across the web — which wins retrieval now and, over training cycles, enters the models' memory.
Unlock all 23 fixes — with copy-paste code.
Claim this brand and subscribe to AVI Pro to see every fix, step-by-step, and track your progress.