Press and Data

In one paragraph

The VIBEAU Index asked ChatGPT, Claude, Gemini, Perplexity and Grok 138 real fragrance shopping questions, none naming a brand, three times each in fresh sessions, on 16 and 17 October 2026: 2,070 answers, 198 houses named. Chanel was the most frequent first pick on all five engines but named first in only 9% of answers. Designer houses took 37% of all mentions and niche houses 26%. Lattafa, a Dubai house, was named in 12% of answers, more than any celebrity brand; Chinese houses appeared only on the prompt that asked for them. No brand pays to be in the Index.

Quotable findings

Chanel is the most frequent first pick on all five AI engines for fragrance, named in 29% of 2,070 answers, but named first in only 9%. Fragrance answers are long lists: a median of ten houses per answer on ChatGPT.

Mention rate 29%, first-mention rate 9%

ChatGPT names Dior most often (39% of its answers); Gemini names Maison Francis Kurkdjian (38%); Perplexity, Claude and Grok name Chanel.

Per-engine mention rates, 414 answers per engine

Designer houses take 37% of all fragrance brand mentions and niche houses 26%. Maison Francis Kurkdjian, Diptyque, Le Labo and Byredo are all in the top ten.

Share of 16,288 house mentions, grouped

Baccarat Rouge 540 is the most-named single fragrance in AI answers, in 324 of 2,070, ahead of Dior Sauvage (304), Glossier You (268), Creed Aventus (231) and Le Labo Santal 33 (222).

Named-fragrance counts across 2,070 answers

Lattafa, a Dubai house, is named in 12% of AI fragrance answers and its Khamrah in 172. Middle Eastern houses take 5% of all mentions; celebrity fragrances take 2%. Sabrina Carpenter, Billie Eilish and Rihanna were not named once.

Mention rates and group shares, 2,070 answers

Chinese fragrance houses (To Summer, Scent Library, Documents, Melt Season) were each named 11 to 16 times, nearly all on the single prompt asking for Chinese brands. On the other 137 questions they did not appear.

Mentions by prompt, 2,070 answers

Asked the same question three times, ChatGPT names the same fragrance house first on 28% of prompts. Gemini 42%, Claude 38%, Grok 53%, Perplexity 73%.

Share of prompts with the same first-named house across all three runs

The most cited source for fragrance is Fragrantica (490 answers). Second and third are whatscent.app (385) and scento.com (343), two sites of machine-generated fragrance descriptions. Reddit (279) and YouTube (252) are top ten.

Answers citing the domain, all five engines

Charts

Download and reuse

PNG, 2000px wide. Licensed CC BY 4.0. Credit: "VIBEAU Index, October 2026".

Named vs named firstMention rate against first-mention rate, top houses
Download PNG

Who gets the shelfShare of mentions by house group
Download PNG

Five engines, one first pickEach engine's top mentions and first pick
Download PNG

The most-named fragrancesNamed-fragrance counts, top twelve
Download PNG

Where the answers come fromMost cited domains, all five engines
Download PNG

Header image2000 × 1000, for article leads
Download PNG

Brands CSV ↓ Sources CSV ↓

Method

How to describe it

Prompts. 138 English shopping prompts across nine clusters: occasion and mood, price and value, scent profile, longevity and performance, gifting, clean and safe, bottle and ritual, generation and origin, and custom and bespoke. No prompt names a brand.
Engines. ChatGPT (gpt-5), Claude (claude-sonnet-5), Gemini (gemini-3.5-flash), Perplexity (sonar) and Grok (grok-4.7), through each provider's API with web search on where the API supports it, US locale, 16 and 17 October 2026. Three runs per prompt per engine, each in a fresh session. 2,070 answers.
Counting. Answers matched against 242 houses and their fragrance names. Mention rate is the share of answers naming the brand; first-mention rate is the share naming it before any other brand. Sponsored placements excluded; none appeared.
Independence. No brand pays to be in the Index, and no VIBEAU client is moved in it.
Limits. API answers can differ from what a signed-in consumer sees in an app. Google AI Mode has no API and is not included. One category, one month; the Index is not a ranking of brand quality.

Suggested description: "the VIBEAU Index, a monthly study of which brands AI assistants recommend, by category, published by brand intelligence platform VIBEAU".

Contact

Questions, interviews, the full dataset

Vivian Qu, founder of VIBEAU, wrote the Index and can speak to the method, the findings and what they mean for beauty brands.

vivian@vibeau.ai · LinkedIn

Replies within one working day. The full response set (all 735 answers) is available to journalists and researchers on request.

Read the edition →

Download and reuse

How to describe it

Questions, interviews, the full dataset