Key Takeaways:
Open AI’s decision in March to pull back from instant checkout quickly led some to consider agentic commerce dead. This conclusion misses the point, and simply hints at the fact that while Open AI’s flagship chatbot ChatGPT has become a highly popular recommendation tool, acting as a retailer or transaction platform is much more complex.
What remains unchanged is agentic AI’s growing influence on the purchase journey. Even without native checkout, ChatGPT drives meaningful traffic to retailers; it accounts for 15% of Target’s referral traffic and 20% of Walmart’s. Consumers trust these AI tools for recommendations, and there is no question that it has become a significant influence platform. According to research by Accenture, generative AI has officially overtaken social media as the top source for recommendations.
Retailers like Ulta Beauty have noticed. Through its innovation team and partnerships with Google and other technology leaders, Ulta is investing in AI as an infrastructure across all its operations, including the shopper journey. The commitment to agentic commerce is something the company is doubling down on, as confirmed by Mike Maresca, Chief Technology and Transformation Officer at Ulta Beauty.
“Being at the forefront of agentic commerce allows us to meet guests where they are and deliver personalized experiences that feel seamless, relevant, and undeniably Ulta Beauty,” he told BeautyMatter.
He also acknowledged that “beauty is deeply personal, and our guests are discovering and shopping in new ways every day,” highlighting how the category is exposed to the shift in discovery that agentic AI is driving. This technology can now deliver dynamic personalization, which very much fits beauty shoppers’ needs and expectations. Today, more than 60% of beauty consumers begin their journey via a guided diagnostic such as an AI skin analysis or chat, rather than browsing through a website’s entire product list.
As the discovery process evolves, the journey from awareness to exploration and conversion gets shorter and quicker. Jing Feng, co-founder and COO of Bluefish, an AI marketing platform for Fortune 500 companies, explained why recognizing this gradual shift in consumer behavior when it comes to beauty discovery is key to meeting consumer demand via these large language models (LLMs). “AI-driven shopping experiences are fundamentally different from traditional search: Intent signals are far richer, consumers move through the funnel much faster, and product recommendations are dynamically personalized in real time,” she said.
Purchase intent is indeed stronger for consumers being redirected to a beauty brand’s website after they’ve researched and explored via AI chats. Research indicates beauty brands using AI-guided journeys see conversion rates of two to three times higher than standard browsing paths.
What This Means for Beauty Brands
Brands must adapt their landing experiences to reflect differing AI consumer journeys and profiles. A user arriving after researching a serum for oily, breakout-prone skin does not have the same expectations as someone exploring a routine for sensitive skin. This requires beauty and skincare brands to anticipate and capture the full spectrum of use cases their products serve, and ensure their e-commerce touchpoints are aligned with how and why they are being recommended by LLMs.
In an industry where visuals and aesthetics drive interest and traffic, the repercussions are consequential. Ghezal Ebrat, founder and CEO of SearchBeaute.ai, an AI-powered beauty recommendation tool, told BeautyMatter, “The digital shelf, influencer marketing, retailer placements, and SEO have historically determined which products win [in beauty]. Agentic commerce shifts power away from who can capture attention and toward who best satisfies a specific user need.” In other words, context and relevance are now imperative.
Indeed, as AI agents mention brands in the context of conversations with users, meeting context and product needs becomes essential. “If an AI agent is tasked with finding the most suitable niacinamide serum for a particular skin profile and budget, brand awareness alone may no longer guarantee selection,” said Ebrat. Beauty brands need to design more personalized and adaptive content tailored to the consumers who are visiting their website with specific intent.
This also means that brands with the most clearly defined use cases could be at an advantage when seeking visibility and recommendations on such platforms.
“Traditional SEO and paid media focus on keywords, rankings, and impressions, rewarding visibility and spend rather than relevance. In an agentic world, discoverability depends on intent and context, meaning products match a user’s skin type, concerns, routine, and preferences.” This reinforces the importance of having legible, clear, and coherent product content that these algorithms can pick up on and confidently share with their users.
A simple search on ChatGPT for a skincare product based on a few personal criteria, such as sensation, skin concern, and price, typically returns just a few results, often no more than five. This compression of the selection set drastically changes the dynamics of discovery: Instead of competing for attention, brands must now compete for inclusion in highly curated shortlists of recommendations.
As competition for visibility therefore intensifies, understanding how these AI algorithms work becomes crucial. Brands cannot control how LLMs incorporate their products into chats, but they can influence the selection process.
Yogesh Chavda, founder of Y2S Consulting, which helps CMOs redesign marketing for an AI-first world, told BeautyMatter that brands need to focus on what kind of conversations consumers are having with ChatGPT. “It's about the relevancy of consumer language, and framing your brand along those lines, rather than defaulting to technical language, is key,” he said.
Feng shared a similar statement, adding that brands that clearly communicate their unique product details across all of their assets, while continually refining their baseline improvements, will be more likely to appear in agent-driven recommendations.
Concretely, brands need to optimize for how different AI agents gather information to make relevant, trustworthy recommendations. They will browse the web for customer reviews, press coverage, product information, metadata, and any other legible online source that can feed their algorithms enough information to foster trust. As Chavda put it, “The way brands show up and the type of content they put out there has to be strong enough for LLMs to trust and consider relevant.”
As AI agents become more active sources of beauty and skincare recommendations, guiding decisions, and influencing routines, the rules of brand visibility are being rewritten. Discovery is shifting the balance between intent and influence, with conversion depending less on traditional marketing and more on how clearly a product is understood by both LLMs and consumers.
While agentic AI might not be an immediate sales channel in itself just yet, its ability to drive serious intent and referral traffic to retailers is already proven. ChatGPT, Gemini, and others are growing into a new layer of discovery infrastructure, where brand visibility relies on trust, relevance, and the ability to meet specific consumer needs and criteria with precision.