Florasis花西子
Known by name, missing from the shortlist.

What we did

1. Wrote twelve questions a premium-makeup shopper asks across the whole decision: finding brands, solving a problem, looking for alternatives, checking who to trust, choosing a gift, and comparing Florasis with named rivals.
2. Asked every question on seven engines, three Western and four Chinese, and stored every answer verbatim. Florasis was counted in English and in Chinese (花西子).
3. Traced the sources behind every answer that cited any, down to the domain.
4. Read florasis.com the way an AI crawler does: robots.txt, llms.txt and the site’s structured data.
5. Checked every figure twice. A second, independent pass recomputed each number from the raw answers before the report went out.

What we found

Absent until the brand is named

Eight of the questions didn’t mention Florasis: the best premium brands, a gift, a wedding look, long wear, the most trusted names. None of the seven engines named Florasis in any of the 56 answers, and no Chinese beauty brand appeared at all. The shortlist was Western, led by Charlotte Tilbury, Dior and Chanel. The Asian brand the engines reached for was Clé de Peau Beauté.

Known, and framed around packaging

Asked about Florasis directly, all seven engines recognised it and called it a legitimate brand. They described it the same way: premium or accessible luxury, with its packaging as the reason to buy. Asked to compare it with Charlotte Tilbury, three engines picked Charlotte Tilbury overall and none picked Florasis.

AI citing unofficial storefronts

Perplexity cited two lookalike sites that present themselves as Florasis but aren’t on its official list of sales channels. It cited them alongside florasis.com, including in its answer to whether Florasis is legitimate. ChatGPT cited neither.

A loosely linked identity

florasis.com lets AI crawlers in. But its structured data doesn’t connect the English and Chinese names, and its llms.txt describes checkout rather than the brand. Those are the signals an engine uses to know that Florasis and 花西子 are one company.

What we recommended

1. Close the storefront gap. Pursue the lookalike domains, and make the official-channels page the source engines cite when asked whether Florasis is legitimate or where to buy it.
2. Tie the English and Chinese names together in the site’s structured data, and rewrite llms.txt so it describes the brand.
3. Earn a place in the coverage engines read. The most-cited sources on the unprompted questions were makeupbrands.org, Allure and Who What Wear, along with Macy’s, Harrods and Sephora.
4. Give engines evidence beyond packaging: independent wear tests and reviews, which answer the long-wear and most-trusted questions where Florasis scored zero.
5. Measure the Chinese-language market separately. This snapshot asked in English, and shoppers in China ask their own engines in Chinese.