How AI Search Is Changing Beauty Product Discovery

Ask ChatGPT or Google's AI Overviews for a skincare recommendation and you'll get a confident, specific answer, often naming a product before a consumer has ever visited a brand's website. That shift is happening faster than most beauty brands have adapted to. RoC Skincare's CMO Hillary Hutcheson has put a figure on it: roughly 40% of consumers are now discovering new products through generative AI search, not a traditional search engine results page.

For beauty and skincare brands, that changes what "being found" actually means.

A few brands already dominate AI recommendations

Data from the eMarketer AI Visibility Index, reported by BeautyMatter, shows just how concentrated AI-driven discovery already is. In Q1 2026, La Roche-Posay led the overall leaderboard, appearing in 22% of beauty-related AI recommendations, followed by CeraVe at 20%, with Neutrogena and Vanicream tied at 15%.

Zoom into specific categories and the gap widens further: La Roche-Posay appears in 81% of facial skincare queries and CeraVe in 71%. In bodycare, it's CeraVe at 58% and La Roche-Posay at 57%. In fragrance, Chanel and Dior share the top spot at 52% each. If your brand isn't one of a handful of names an AI model already trusts, you're largely invisible in these answers, no matter your organic rankings.

Even heritage brands are finding they can't be found

This isn't only a problem for smaller or newer brands. Glossy reported that Borghese, an Italian skincare brand with more than 70 years of history, struggled to get its hero product surfaced in AI search at all. "It took us 14 times to find it on search, on AI," said COO Dawn Hilarczyk, prompting the brand to launch an internal "Project PDP" to rebuild product pages around clear, factual, scientifically-backed copy.

RoC Skincare took a similar approach, building more than 400 Q&A sets directly into its product pages to match the way consumers actually phrase questions to AI tools, and using AI simulation tools to identify which third-party sources were shaping its generative search rankings. One finding stood out: legacy publications like Vogue and Allure, now operating as digital properties, carry significant weight in what AI models recommend.

What this means for your brand

Ranking well in Google is no longer the finish line. Being cited, understood and recommended by generative AI search depends on having clean, structured, factual product information, strong third-party coverage and a content strategy that answers the exact questions buyers are asking, whether that's on your own site or the publications AI models trust. Brands that treat this as a new discipline now will be the ones showing up in the answer, not just the search results.