GEO for Wine & Spirits · Benchfolk 2026
Which wine & spirits brands dominate ChatGPT — and which are invisible.
Benchfolk runs real LLM tests across champagne, cognac, Scotch whisky and fine wine. A documented prompt panel · 4 AI systems · updated monthly. The concentration of AI visibility is sharper than you think.
capture 60% of ChatGPT recommendations (out of 16,000 producers)
go to 4 châteaux out of 6,000+
ChatGPT · Claude · Perplexity · Gemini
per sub-vertical, updated monthly
Why AI visibility matters for wine & spirits brands
Premium wine and spirits buyers are shifting their discovery behavior. Instead of searching Google for "best champagne for a gift" or "cognac recommendations," a growing share — particularly in the 35–55 affluent buyer segment — ask ChatGPT, Perplexity or Claude directly.
The problem: LLM recommendations are highly concentrated. Our benchmark data from May 2026 shows that in every category we tested, 3-4 brands capture the majority of AI citations. The other hundreds — or thousands — of producers are invisible.
Unlike Google rankings, this invisibility is not primarily about marketing budget. It is driven by structured data gaps: missing Wikidata entities, absent Wikipedia articles in key languages, no Schema.org markup, and no presence in the premium media sources (Decanter, Wine Advocate, Wine Spectator, RVF) that LLMs weight most heavily in their training corpus.
These are fixable problems. A grower-producer with 50,000 bottles per year can outrank a well-funded négociant in ChatGPT if its structured data is richer. Benchfolk measures the gap and turns it into prioritised SEO, content, data or authority work.
How visible is your brand in ChatGPT and Perplexity?
Lancer l'audit gratuit →Benchmark findings by category
out of 16,000 producers
Cristal (Roederer) and Taittinger rank 5th and 6th despite strong DR scores — a clear gap between Google authority and LLM citation.
out of 270 producers
Independent houses like Frapin, Delamain and Pierre Ferrand barely appear despite critical acclaim. The corpus skews heavily toward the four LVMH/Rémy Cointreau/Pernod brands.
out of 150+ producers
Japanese whisky (Nikka, Suntory Yamazaki) punches above its weight — strong English-language Wikipedia presence in the whisky corpus.
out of 6,000+ producers
The 1855 classification still dominates LLM responses 171 years later. Second wines and Crus Bourgeois are structurally invisible — a solvable problem.
How Benchfolk measures AI visibility
Per sub-vertical. Designed to reflect real buyer questions: gifting, pairing, connoisseur recommendations, regional discovery. Updated monthly as LLM behavior evolves.
GPT-4o, Perplexity Sonar Pro, Claude Sonnet. Each has a different training corpus and retrieval behavior — a brand can rank well on one and be invisible on another.
Citation frequency + position in response + context quality + sector benchmark. One number that tracks over time. Agent ships the actions that move it.
Frequently asked questions
What is GEO for wine and spirits brands?↓
How does Benchfolk measure AI visibility for wine and spirits?↓
Which LLMs does Benchfolk test?↓
How long does it take to improve a wine brand's LLM visibility?↓
Can smaller wine estates compete with large houses like Dom Pérignon or Hennessy?↓
Explore the benchmarks
22 houses ranked by ChatGPT citations
LVMH vs independents — who wins in AI
Speyside, Islay, Highlands, Japan
6,000+ châteaux, 4 in the core club
ChatGPT vs Perplexity vs Claude
Score your brand in 5 minutes
Find out where your brand stands in ChatGPT and Perplexity
La Carte gratuite par email : conversations perdues, concurrents cités et première priorité. Aucune carte bleue.
Tester ma marque gratuitement →