- Large beauty groups now treat AI answers as a channel: Estée Lauder Companies announced a portfolio-wide GEO deployment in September 2026, and L'Oréal a ChatGPT partnership in June 2026.
- Beauty is the fastest-growing category for AI referral traffic in Similarweb's data, but still the smallest by volume.
- In one study, 96.1% of the sources ChatGPT cited for beauty recommendations were third-party sites — so beauty GEO is as much about retailers, media and experts as about your own site.
- On the supply side, almost all business buyers use AI somewhere in their buying process, and Alibaba.com's own AI sourcing agent reports more than 10 million monthly users. Beauty-specific supplier data does not exist yet.
Why this page exists
GEO is being sold with big numbers. Some are solid; many come from market-size press releases or from companies selling GEO tools. This page keeps only figures we could trace to the original publisher on 10 October 2026, says who published them, and lists what is still unknown. We sell Beauty GEO, so read it with that in mind — and check the sources.
1. Shoppers: are people asking AI what beauty products to buy?
| Finding | Source | Notes |
|---|---|---|
| 49% of consumers surveyed had received beauty product recommendations from generative AI platforms such as ChatGPT, Claude, Gemini or Copilot | NielsenIQ, 2026 | Sample size not stated on the page |
| 38% of US beauty shoppers had used AI for beauty research or purchases | Tinuiti beauty marketing study, March 2026 | 1,050 US beauty shoppers; Tinuiti is an agency |
| More than 250 million customers used Amazon's Rufus assistant in 2025; users were 60% more likely to complete a purchase | AWS Machine Learning blog, November 2025 | All categories, not beauty only |
| Sephora launched an app inside ChatGPT, piloting in the United States | Sephora newsroom, 24 March 2026 | Retailer channel, not organic answers |
Read together: a large minority of beauty shoppers already use AI assistants, and the biggest retailers are building inside them. Survey figures measure whether people have ever used AI for beauty — not how many purchases it decides.
2. Brands: who is acting?
| Company | What was announced | Source |
|---|---|---|
| Estée Lauder Companies | Deploying Profound across its full portfolio to "operationalize Generative Engine Optimization (GEO) and AI marketing at scale", including product pages, blogs and YouTube content | ELC press release, 14 September 2026 |
| L'Oréal | Partnership with OpenAI to strengthen product discovery in ChatGPT in the US for brands from Lancôme to Kérastase; SkinCeuticals, CeraVe and Garnier in the ChatGPT ad pilot | L'Oréal Finance, 17 June 2026 |
| e.l.f. Beauty | Built an internal team to make its products more visible to large language models; its chief digital officer estimated that "almost 60% of discovery" happens on LLMs (no method given) | Glossy, 1 April 2026 |
Large groups are buying enterprise tools, signing platform deals and building teams in-house. Mid-sized and independent brands — and Asian brands entering the US and Europe — usually lack those teams, which is where specialist agencies come in.
3. Traffic: how big is AI as a channel today?
- Beauty is growing fastest — from a small base. Similarweb counted about 1.0 million AI referral visits to beauty sites in May 2025 and 4.0 million in May 2026, the fastest growth of any category, while noting that beauty was last by volume and that AI favours large multi-category platforms (Similarweb, 2026).
- AI referrals are still about 1% of traffic. Across ten industries, AI referrals were 1.08% of all website visits (Conductor 2026 AEO/GEO Benchmarks; Conductor sells GEO software).
- The bigger effect is on Google clicks. Pew Research found users clicked a traditional result on 8% of visits when an AI summary appeared, against 15% without one (900 US adults, March 2025 browsing data, published 22 July 2025).
So today, beauty GEO is less about a large new traffic stream and more about not disappearing from the answers that are replacing clicks — while the channel grows.
4. Sources: what does AI rely on for beauty answers?
Avenue Z analysed ChatGPT's beauty recommendations for 60 brands (565 citations across 75 prompts) and found that 96.1% of citations came from third-party sources rather than brand sites (reported by Personal Care Insights, 21 July 2026; Avenue Z is an agency). Our own smaller snapshot linked brand sites more often — 10 of 26 sources — with retailers close behind (method and limits). Either way, retailer listings, editorial coverage and expert sources matter as much as your own pages.
5. Suppliers: do brand buyers ask AI to find factories?
| Finding | Source | Notes |
|---|---|---|
| 94% of business buyers report using AI during their buying process | Forrester, State of Business Buying 2026 | Cross-industry, technology-heavy |
| 51% of B2B software buyers now start research with an AI chatbot more often than with Google (29% a year earlier) | G2, The Answer Economy, April 2026 | 1,076 software buyers; G2 sells review services |
| Alibaba.com's Accio matches buyer requests against "factory capabilities and track records" | Alibaba.com press release, 14 November 2025 | AI sourcing inside a marketplace |
| Accio Work has more than 10 million monthly active users, according to Alibaba | Digital Commerce 360, 24 July 2026 | Company figure |
| Cosmoprof Asia 2025 drew 64,761 visitors and 2,688 exhibitors | Cosmoprof, 2025 | Trade shows remain a major sourcing channel |
Buyers clearly use AI. But much AI sourcing happens inside marketplaces, and trade shows and referrals still matter. For a beauty supplier, that means two AI surfaces to manage: the open web that ChatGPT, Gemini and Perplexity read, and marketplace listings that tools like Accio read — with the same verifiable facts on both.
6. What the data does not show
- No public data on what share of beauty purchases, or of beauty OEM/ODM enquiries, start in an AI assistant.
- No credible market-size figure for GEO services. Published estimates vary widely, and several do not add up arithmetically.
- Almost no search demand for the phrase "beauty GEO" itself: Google Trends shows too little data, while searches for "AI SEO" grew several-fold from 2024 to 2026 (our check, October 2026). Buyers describe the problem in their own words.
- No independent audit of GEO results. Most case studies are published by the vendors themselves — ours included, which is why we publish raw answers.
What we conclude
- Beauty brands: demand is real and led by large groups; the opportunity for specialist help is mid-sized brands and brands entering new markets. The work spans product data, claims, retailers and earned media.
- Beauty suppliers: early. Buyers use AI, but sourcing still runs through shows, marketplaces and referrals. GEO works best combined with marketplace listings and trade-show preparation.
- Everyone: measure with fixed prompts and raw answers, and judge GEO on whether AI names you accurately — not on promises of rankings.