AI in Beauty Isn't One Story
What the giants and the independents are each teaching mid-sized maisons
Ask “how is AI used in beauty” and you get two very different answers, depending on who's asked. At L'Oréal, it looks like Beauty Genius: an agentic AI assistant on WhatsApp that scans a face, cross-references decades of formulation data across Lancôme, Garnier, Vichy and CeraVe, and recommends a routine in under a minute.¹
At Estée Lauder, it looks like ConsumerIQ and Trend Studio: two generative AI agents built with Microsoft that read consumer data across 25 brands and 150+ countries, catch a trend forming on TikTok, and draft the marketing copy before a human opens a brief.²
McKinsey estimates generative AI alone could unlock $9–10 billion in value across the global beauty industry, most of it sitting exactly where these two examples point: personalisation and speed.³
That's one story: the one that gets written up, because it's L'Oréal and Estée Lauder. It's also not one most mid-sized maisons can copy: neither company got here by buying a tool, but by having twenty-plus years of structured consumer and formulation data already sitting in a system an AI agent could query. The tool was always the easy part.
The second story is more useful, and it's told far less often. Independent and mid-sized brands are getting real, measurable results from the same underlying capability, at a fraction of the budget:
– NuNorm, a direct-to-consumer makeup brand, reports that 75% of its orders are secured through its AI shade-finder and virtual try-on tool alone, closing a confidence gap for customers who'd otherwise never risk an online shade match.⁴
– Ground AI's clients (RMS Beauty, Violette_FR, Salt & Stone) see up to 20% additional online revenue from AI-personalised marketing, without adding headcount.⁴
– Daash Intelligence gives brands like Glow Recipe and Cocokind weekly, store-level SKU visibility that used to be gatekept by the retailers themselves: the same competitive intelligence a wholesale-heavy maison usually waits a season to piece together.⁴
– Farmacy Beauty's founder puts the marketing case plainly: “If we didn't have AI, we would have to hire consultants, which can be expensive.” AI is replacing a line item, not adding one.⁴
None of these are enterprise budgets. None required a data science team. What they required was clean, structured first-party data (shade libraries, purchase history, SKU-level sales, ingredient data) organised well enough that a tool could actually use it.
What this means if you're running a mid-sized maison
Three things, roughly in order of where the return actually shows up. First, personalisation pays for itself fastest, and doesn't require a giant's dataset: a shade-matching or skin-diagnostic tool built on your own customer and product data is closing sales for brands a fraction of your size, not just the majors.
Second, the trend-detection and marketing layer is now genuinely accessible without an agency retainer. The brands seeing the clearest return aren't the ones with the biggest AI budget; they're the ones who stopped outsourcing insight and started asking their own data the question directly.
Third, and this is the one boards tend to underweight, competitive intelligence has quietly become a data problem rather than a research problem. Weekly, SKU-level visibility, once the retailer's privilege to withhold, is now something a mid-sized brand can build for itself. Wholesale partners who used to control that information asymmetry are losing it.
What none of this changes: whether it's L'Oréal's twenty years of formulation data or a two-person indie brand's shade library, every result above traces back to data that was clean and structured before AI ever touched it. The tool is never the differentiator. The foundation is.
If you're weighing where AI would actually move a number for your business, rather than just look impressive in a board deck, I'm always glad to compare notes.
Elisabeth
The Beauty of Data. Designed for Growth.
References
1. IMD, “Pure Genius: How L'Oréal Is Helping Customers Solve Their Own Problems” (on L'Oréal's Beauty Genius AI assistant). https://www.imd.org/ibyimd/strategy/pure-genius-how-loreal-is-helping-customers-solve-their-own-problems/
2. Microsoft Source, “Estée Lauder Uses AI to Reimagine Trend Forecasting and Consumer Marketing.” https://news.microsoft.com/source/features/digital-transformation/estee-lauder-uses-ai-to-reimagine-trend-forecasting-and-consumer-marketing/
3. McKinsey estimate on generative AI's value potential in beauty, as cited in Microsoft Source, “Estée Lauder Uses AI to Reimagine Trend Forecasting and Consumer Marketing.” https://news.microsoft.com/source/features/digital-transformation/estee-lauder-uses-ai-to-reimagine-trend-forecasting-and-consumer-marketing/
4. US Chamber of Commerce, CO—, “How AI Is Leveling the Beauty Industry Playing Field for Small Brands” (NuNorm, Ground AI, Daash Intelligence and Farmacy Beauty cases). https://www.uschamber.com/co/good-company/launch-pad/ai-for-small-beauty-brands
