L'Oréal Is Paying OpenAI to Get Found in ChatGPT
The world's largest beauty company just told the market what its top priority is. At VivaTech 2026, L'Oréal announced it will work with OpenAI to improve how its products are surfaced in ChatGPT in the United States, starting with Lancôme and Kérastase (AI News, June 22, 2026). When a brand this size pays for AI discovery, every brand below it just lost the excuse to wait.
TL;DR: L'Oréal's new OpenAI partnership puts "how do my products surface in ChatGPT" at the center of the world's biggest beauty company's strategy. The mid-market can't sign a bespoke OpenAI deal, so the path to the same outcome runs through measurement and catalog optimization, not a custom partnership.
What did L'Oréal actually announce?
L'Oréal announced a broad OpenAI partnership at VivaTech 2026 spanning consumer tools, advertising, research, and internal training. The commerce-relevant piece: L'Oréal will work with OpenAI to improve how its products are surfaced in ChatGPT in the US, naming Lancôme and Kérastase first (AI News, June 22, 2026).
Three pieces matter for any brand watching this:
- Product surfacing. L'Oréal is optimizing how Lancôme and Kérastase appear when shoppers ask ChatGPT about skincare and haircare. This is discovery, not checkout.
- A ChatGPT ads pilot. SkinCeuticals, CeraVe, and Garnier are in OpenAI's global advertising pilot, placing ads at moments of shopping intent (AI News, June 22, 2026).
- Conversational try-on. Maybelline's virtual makeup try-on now runs inside ChatGPT through L'Oréal's ModiFace AR technology.
The shopper still buys on the retailer's site. ChatGPT recommends and directs; it does not close the transaction. That model has held since OpenAI discontinued in-chat Instant Checkout in March 2026, and the open Agentic Commerce Protocol now powers product discovery and merchant redirect rather than checkout (agenticcommerce.dev).
Why a beauty giant cares about ChatGPT placement
ChatGPT now has more than 900 million weekly active users and over 50 million subscribers (OpenAI, 2026, via AI News). That is a discovery surface the size of a top-three search engine, and shoppers increasingly start product research there instead of Google.
L'Oréal also has the ecommerce base to justify the bet. Its online sales grew double digits in 2025 and passed 30% of total sales (L'Oréal, via AI News, June 22, 2026). When a third of revenue runs through digital channels and a 900M-user assistant is steering product choices, ChatGPT shopping placement stops being experimental and becomes a revenue line worth protecting.
When the world's number-one beauty company says it needs to optimize how its products surface in ChatGPT, the question is no longer whether AI discovery matters. It is whether you can answer the same question for your own catalog.
This is the real signal. Not the partnership logo. The fact that L'Oréal has decided AI product surfacing is worth a named, public commitment.
The mid-market problem L'Oréal just exposed
L'Oréal gets a bespoke, enterprise-scale OpenAI deal. A direct relationship, custom integration, and a seat in an ads pilot. A mid-market brand doing $20M to $500M cannot pick up the phone and get the same arrangement. OpenAI is not signing custom partnerships with the long tail of the market.
That gap is the whole point. The outcome L'Oréal is buying (its products getting recommended when a shopper asks ChatGPT for the best retinol or the right shampoo) is available to any brand. The mechanism is different. You get there through clean, structured, AI product recommendations data that AI systems can read and rank, not through a handshake with OpenAI.
Here is the practical translation:
| What L'Oréal gets | What the mid-market gets instead |
|---|---|
| Named OpenAI partnership | No custom deal exists |
| Custom integration team | Standards-based product feeds (ACP, UCP) |
| Ads pilot seat | Organic AI discovery via catalog quality |
| Enterprise budget | Measurement plus optimization at SaaS cost |
The brands that win the next 18 months will not be the ones with the biggest OpenAI contract. They will be the ones whose product data is good enough that AI surfaces them anyway.
How AI decides which beauty product to recommend
When a shopper asks ChatGPT "what's the best vitamin C serum for sensitive skin," the model does not run one search. It decomposes that request into 8 to 12 sub-queries about ingredients, skin type, price, and reviews, then assembles an answer from whatever product data it can retrieve. Google calls this query fan-out, and AI shopping surfaces work the same way.
The consequence: your product can be a household name and still lose. If your catalog answers the top-level query but not the sub-queries about texture, concentration, or fragrance-free formulation, a smaller competitor with richer data gets recommended instead. A Surfer SEO study of 173,902 URLs found that 68% of AI-cited pages are not in the top 10 organic results (Surfer SEO, December 2025). Ranking on Google does not guarantee you get picked by the agent.
This is why L'Oréal is investing in surfacing rather than assuming brand strength carries over. Ulta Beauty learned the same lesson when it went live on Google's commerce protocol earlier this year, proving that beauty is where structured product data gets tested first (covered here). Strong brand recognition is an input. It is not the deciding factor when an agent is reading attributes.
What to do this week
You do not need an OpenAI partnership to act on the same insight L'Oréal is acting on. Start here:
- Run a discovery check. Ask ChatGPT, Google AI Mode, and Perplexity five real category questions a shopper would use. Note whether your products appear, get named, or get shown as cards with price and image. Tools like our AI Readiness Report score this in about 30 seconds.
- Compare branded vs unbranded results. Search your brand name, then search the category without it. Measuring branded vs unbranded AI queries shows whether you only win when shoppers already know you.
- Audit your top 20 SKUs for sub-query coverage. For each, list the attributes a shopper asks about (skin type, ingredients, concentration, finish). Fill every gap in the product data.
- Track your AI share of voice against named competitors. Watch your AI share of voice on category queries, not just whether you appear in isolation.
- Treat this as an ongoing program. Answer engine optimization is continuous, the way SEO became. Surfaces and models change monthly.
Frequently Asked Questions
Does the L'Oréal deal mean shoppers buy beauty products inside ChatGPT?
No. The partnership covers product discovery, advertising, and conversational try-on. ChatGPT recommends products and directs shoppers to the merchant's site to complete the purchase. OpenAI discontinued in-chat Instant Checkout in March 2026, so the current model is discovery and redirect, not checkout.
Can a mid-market brand get the same OpenAI partnership L'Oréal got?
No. L'Oréal's arrangement is a bespoke, enterprise-scale deal. OpenAI does not sign custom partnerships with mid-market brands. The same outcome, getting recommended in ChatGPT, is reachable through standards-based product feeds and catalog optimization rather than a direct contract.
Why does product data quality matter more than brand size for AI discovery?
AI surfaces decompose a shopping query into many sub-queries and answer from retrievable product attributes. A well-known brand with thin data can lose to a smaller competitor with richer attributes. Surfer SEO found 68% of AI-cited pages are not top-10 organic, so brand authority alone does not secure the recommendation.
Which AI surfaces should beauty brands prioritize in 2026?
ChatGPT, Google AI Mode, and Perplexity are the primary discovery surfaces for US shoppers. ChatGPT still leads, but Gemini and Perplexity keep gaining ground, so optimizing for a single surface is no longer enough. Measure across all three rather than betting everything on one.
How fast can a brand improve its AI surfacing?
Measurement is immediate. Catalog enrichment, the part that moves results, takes weeks depending on SKU count and data gaps. The brands acting now are not waiting for a partnership; they are fixing product data so AI systems can read and rank them today.
L'Oréal did not announce a side project. It told the market that getting found in ChatGPT is worth a public, named commitment from the largest beauty company on earth. The brands that hear that and assume their size protects them are the ones that will quietly disappear from the answer. The ones that fix their product data will be in it.
How AI-ready are your products?
Check how ChatGPT, Google AI, and Perplexity evaluate any product page. Free score in 30 seconds.
