AI OBSERVATORY · LUXURY · NOVEMBER 2026
Which luxury houses does AI recommend?
Every month, we put the same buying questions to ChatGPT, Gemini and Claude and record the brands they cite. Here is the November 2026 reading for three worlds of luxury: leather goods, Paris hotels and prestige real estate on the French Riviera.
KEY FINDINGS
Four lessons from this reading
- Even the leader misses answersHermès is the top-ranked leather goods house, yet appears in only 79% of answers. Louis Vuitton follows at 66%.
- Each AI has its favouritesFor Paris palace hotels, ChatGPT cites the Four Seasons George V in 9 answers out of 10, while Claude favours Le Bristol and Le Meurice (8 out of 10).
- A brand can exist for one AI onlyPolène is cited in 7 of Claude’s 10 answers, but in a single Gemini answer. A gap a classic Google audit never sees.
- Being cited is not enough: rank mattersEngel & Völkers appears in 87% of prestige real estate answers, but at an average rank of 5; Sotheby’s International Realty is cited every time, on average in 2nd or 3rd place.
RANKING · NOVEMBER 2026
Luxury leather goods
| # | Brand | Score | Presence | Avg. rank | ChatGPT | Gemini | Claude |
|---|---|---|---|---|---|---|---|
| 1 | Hermès | 56.6 | 79.3% | 2.9 | 7/10 | 8/9 | 8/10 |
| 2 | Louis Vuitton | 42.8 | 65.5% | 3.1 | 7/10 | 5/9 | 7/10 |
| 3 | Goyard | 25.5 | 44.8% | 3.7 | 3/10 | 5/9 | 5/10 |
| 4 | Delvaux | 21.4 | 41.4% | 3.9 | 4/10 | 3/9 | 5/10 |
| 5 | Moynat | 20.7 | 41.4% | 3.9 | 3/10 | 4/9 | 5/10 |
| 6 | Chanel | 17.9 | 31% | 3.7 | 3/10 | 3/9 | 3/10 |
| 7 | Bottega Veneta | 16.6 | 37.9% | 5.1 | 3/10 | 4/9 | 4/10 |
| 8 | Berluti | 14.5 | 27.6% | 4.4 | 2/10 | 3/9 | 3/10 |
| 9 | Polène | 13.8 | 34.5% | 7.4 | 2/10 | 1/9 | 7/10 |
| 10 | Valextra | 13.1 | 27.6% | 5.1 | 3/10 | 4/9 | 1/10 |
29 answers analysed (one empty Gemini answer). Score: presence weighted by rank (100 = cited first in every answer).
RANKING · NOVEMBER 2026
Luxury hotels in Paris
| # | Brand | Score | Presence | Avg. rank | ChatGPT | Gemini | Claude |
|---|---|---|---|---|---|---|---|
| 1 | Le Bristol | 41.3 | 60% | 2.8 | 4/10 | 6/10 | 8/10 |
| 2 | Four Seasons George V | 37.3 | 63.3% | 3.4 | 9/10 | 4/10 | 6/10 |
| 3 | Le Meurice | 32 | 50% | 2.9 | 4/10 | 3/10 | 8/10 |
| 4 | Cheval Blanc Paris | 29.3 | 50% | 3.7 | 3/10 | 5/10 | 7/10 |
| 5 | Ritz Paris | 28 | 40% | 2.6 | 6/10 | 2/10 | 4/10 |
| 6 | Plaza Athénée | 27.3 | 50% | 3.5 | 4/10 | 4/10 | 7/10 |
| 7 | Shangri-La Paris | 26.7 | 40% | 3.3 | 4/10 | 3/10 | 5/10 |
| 8 | Crillon | 22 | 40% | 3.6 | 4/10 | 2/10 | 6/10 |
| 9 | Peninsula Paris | 16 | 30% | 4.2 | 3/10 | 3/10 | 3/10 |
| 10 | Royal Monceau | 10 | 13.3% | 3.3 | 1/10 | 0/10 | 3/10 |
30 answers analysed. Score: presence weighted by rank (100 = cited first in every answer).
RANKING · NOVEMBER 2026
Prestige real estate, French Riviera
| # | Brand | Score | Presence | Avg. rank | ChatGPT | Gemini | Claude |
|---|---|---|---|---|---|---|---|
| 1 | Sotheby’s International Realty | 72.7 | 100% | 2.5 | 10/10 | 10/10 | 10/10 |
| 2 | Barnes | 53.3 | 90% | 3.8 | 8/10 | 9/10 | 10/10 |
| 3 | John Taylor | 43.3 | 76.7% | 4.4 | 9/10 | 10/10 | 4/10 |
| 4 | Knight Frank | 35.3 | 70% | 3.9 | 6/10 | 6/10 | 9/10 |
| 5 | Engel & Völkers | 31.3 | 86.7% | 5.3 | 9/10 | 8/10 | 9/10 |
| 6 | Savills | 29.3 | 63.3% | 4.1 | 2/10 | 9/10 | 8/10 |
| 7 | Michaël Zingraf | 27.3 | 36.7% | 2.8 | 4/10 | 7/10 | 0/10 |
| 8 | Christie’s | 16 | 33.3% | 4.2 | 2/10 | 1/10 | 7/10 |
| 9 | Magrey & Sons | 11.3 | 23.3% | 3.9 | 1/10 | 6/10 | 0/10 |
| 10 | Agence des Parcs de Saint-Tropez | 4.7 | 16.7% | 7.4 | 0/10 | 1/10 | 4/10 |
30 answers analysed. Score: presence weighted by rank (100 = cited first in every answer).
METHOD
How we measure
- 10 fixed questions per sectorThe same buying questions every month (for example, « What are the best palace hotels in Paris in 2026? », asked in French), naming no brand, so readings can be compared over time.
- 3 AIs, no web browsingChatGPT, Gemini and Claude are queried through their APIs, without web search: we measure what the models have retained about brands.
- A readable scorePresence (share of answers citing the brand) weighted by average citation rank. 100 means cited first in every answer.
- No endorsementThe ranking only reflects the AIs’ answers. It is neither an opinion nor a recommendation from Million Marketing.
WHAT ABOUT YOUR HOUSE?
Find out what AI says about you
FAQ: the luxury AI observatory
How often is the observatory updated?
Every month, with the same questions, to track how cited brands evolve.
Why don’t all AIs cite the same brands?
Each model was trained on different data and weighs sources differently, which is why we measure all three separately.
Can you measure a brand that is not in this ranking?
Yes: the AIO Diagnostic puts ten buying questions from your sector to the three AIs and measures your brand’s place against your competitors.