AI platforms split on beauty brand recommendations

Foundation has published research on how major AI platforms recommend beauty brands, finding sharp differences by category and platform.

The agency analysed responses from ChatGPT, Gemini, Perplexity and Claude to 20 consumer-style beauty prompts, each repeated three times. It logged 1,335 brand mentions and 690 cited sources, and found that AI visibility was tied more closely to authority in a specific segment than to broad brand recognition.

Across all four platforms, the most frequently recommended brands were e.l.f., Maybelline, Charlotte Tilbury, NYX, L'Oréal Paris, NARS, Hourglass, Fenty Beauty, Rare Beauty and Merit. The picture changed when prompts focused on specific parts of the market. Budget beauty queries most often returned e.l.f., Maybelline, NYX and L'Oréal Paris.

Clean beauty prompts produced a different group of leaders: ILIA, Kosas, Merit, Saie and Tower 28. Luxury-focused questions tended to favour Giorgio Armani, Charlotte Tilbury and NARS.

Product categories also showed distinct patterns. Charlotte Tilbury, Makeup by Mario, Urban Decay and Natasha Denona ranked strongly for eye products, while Aquaphor, La Roche-Posay, Laneige and Vaseline appeared frequently in lip care recommendations.

The dataset showed limited overlap between platforms. Only 40 brand recommendations were returned consistently by all four AI systems for the same prompt, suggesting consumer exposure can vary widely depending on which assistant they use.

Platform differences

Each model drew on different source types when producing recommendations. ChatGPT cited established beauty and lifestyle publications such as Vogue, Allure and Cosmopolitan heavily.

Gemini mixed editorial coverage with official product information and reference-style directories. Perplexity leaned more on retailer listings and community-led platforms, while Claude generated the widest spread of recommendations and often surfaced brands that did not appear elsewhere in the research.

For marketers, that divergence points to a fragmented discovery environment rather than a single route to prominence. A brand that performs well on one model may not appear at all on another, even when users ask closely related questions.

Foundation also found that third-party sources carried far more weight than material published by brands themselves. Earned media from editorial digital PR accounted for 44% of all cited sources in the sample, compared with 7% for brand-owned content.

Role of Reddit

Among individual sources, Reddit was the most cited platform in the study. It ranked ahead of major beauty publications including Vogue and Allure, indicating that user discussion plays a major role in how AI tools assemble beauty answers.

Olivia Ford, Performance Lead at Foundation, highlighted one of the central conclusions. "One of the clearest findings from the research is that beauty brands don't need to win every conversation. The brands appearing most consistently in AI recommendations are those that own a particular space, whether that's affordable makeup, clean beauty, luxury cosmetics or a specific hero product category," Ford said.

The methodology was designed around prompts intended to resemble natural consumer questions for AI assistants rather than conventional search queries. Examples included requests for the best budget foundation for oily skin and advice on achieving a glowy base without looking greasy.

Each prompt was submitted to the four platforms in three separate runs to test consistency over time, across sessions and between users. Every named brand was recorded along with its position in the answer and any cited source, then grouped by brand, category and source type.

The findings add to a broader shift in digital marketing as consumer discovery moves from traditional search engines to generative AI tools. For beauty companies in particular, where peer discussion, reviews and editorial endorsement have long shaped purchasing decisions, the data suggests those signals are now being reused and amplified in AI-generated recommendations.

Ford said the results showed that visibility depended as much on outside validation as on direct brand messaging. "AI platforms rely heavily on third-party validation and although editorial coverage remains incredibly influential, now we're also seeing that community discussion is just as important. The brands earning visibility are those being talked about by trusted publishers, retailers and consumers, not simply those talking about themselves," Ford said.