How AI recommends beauty products and what it takes to make the cut

Reddit, Who What Wear, Wikipedia, and Sephora are the top sources ChatGPT cites when beauty consumers ask for product recommendations. This is according to findings from Novi, a data infrastructure platform that aims to help brands optimize their digital discovery strategies. 

“Our research shows products with verified trust signals, like certifications and badges, are presented to shoppers significantly more often than products without them,” says Kimberly Shenk, CEO at Novi.

Shenk explains that AI models weigh both SKU-level data and trust signals when recommending what products consumers should buy. This means that the AI does not simply recommend products because they appear on sites like Reddit or because a brand often mentions them online. Rather, it considers whether the product information is clear and consistent and if it has signals that indicate credibility and quality.

“For beauty brands, that means the most effective path to AI visibility is pairing well-structured data with verified trust signals, not chasing visibility via any single source,” she adds.

AI-powered tools are changing the way beauty consumers find their products, and brands that don’t adapt their discovery strategies risk becoming invisible. 

Bot-driven uproar

Generative AI search, through large language models (LLMs) such as ChatGPT or Gemini, yields results as fast as traditional Google searches, but replies are presented more conversationally. Consumers are increasingly using these tools for product discovery, seeking personalized routines in a saturated market.

LLMs like ChatGPT are changing the way beauty consumers find their products.Novi’s analysis examined 10.7 million citations across over 98,000 source websites that ChatGPT considers when generating responses. 

According to the data platform, 73% of consumers use AI search to learn about products, categories, brands, and services. Moreover, shoppers who were directed to a retailer’s website via an AI tool were 31% more likely to complete their purchase than the comparison group.

The figures suggest that generative engine optimization (GEO) should become a growing priority for brands that want to tap into the modern zeitgeist, alongside their traditional search engine optimization (SEO) efforts.

However, the data platform notes that brands are struggling to keep pace with the shift toward AI search, with 47% of brands not knowing if or how they are showing up in AI-generated responses. 

Categoric nuances

Novi’s analysis found that AI tools prioritize different sources depending on the category of beauty products consumers search for.

“For beauty category queries overall, Reddit ranked first in number of citations by a significant margin,” the report outlines.

For skin care-related queries, AI algorithms tend to lean more toward editorial outlets and major retailers like Sephora. But Novi adds that, while missing from the top five ranking, some brand websites like Neutrogena and La Roche Posay are also cited for responses. 

Consumers are increasingly using AI tools for product discovery, seeking personalized routines in a saturated market.Meanwhile, fragrance-related responses are more typically built on brand-owned content.

Chasing visibility through credibility

Tech companies are picking up on the shift toward AI-driven search, and multiple have stepped in to address the explosion. Personal Care Insights recently spoke with Revieve about a ChatGPT plug-in it unveiled to help guide brands and retailers on how consumers discover skin care products inside the LLM.

The AI Skin Advisor for ChatGPT allows brands to control the tone, product mapping, educational content, and personalization rules inside the AI interface.

“There’s a risk of generic or unbranded guidance if brands aren’t actively participating in these environments,” said Irina Mazur, chief commercial and marketing officer at Revieve.

Reddit was shown to inform the bulk of ChatGPT’s responses, revealing that appearing in online conversations lifts the odds of a brand surfacing in AI search results. 

However, a try-hard approach in pursuit of exposure can be counterproductive and harm a brand’s image instead of amplifying it.

“We see brands trying to do too much, show up on everything across platforms, and spread themselves too thin across many things, versus owning strategic categories well,” Emily Rose Campbell, head of performance at marketing agency Iced Media, previously told us.

Novi’s report maintains that a balanced strategy should focus on making product information clear and credible. Providing ingredient information, claims like vegan or fragrance-free, certifications, size, packaging, and the same descriptions across all sites where the product is listed helps AI models prioritize them in results. 

Moreover, signals like dermatologist-tested badges, COSMOS and Environmental Working Group certifications, high review volumes, and retailer validation — like being stocked at Sephora or Ulta — help AI systems determine whether a product is trustworthy enough to recommend.

In the age of AI-powered search, it appears brands may need to win over the trust of algorithms before they can win over the shopper.