L'Oreal's Moat Isn't Its 37 Brands. It's the 3% You Never See.

In 2024, L'Oréal's chemists filed 694 patents, launched more than 3,600 new formulas, and spent more than €1.3 billion finding things out about hair, skin, and pigment.³ None of that arrives at the counter with a label on it. You buy a lipstick or a serum and read a brand name - Lancôme, Maybelline, CeraVe, Kérastase - and you assume the brand is the thing you're paying for. It isn't, or not mostly. Underneath all 37 of those brands sits a single shared machine that does the actual inventing, and it is the most underrated competitive advantage in consumer goods.¹

The official story is that L'Oréal is a brand collector - a holding company that buys and grooms beauty labels. That's the surface. The real engine is one science base, funded at a stubbornly constant share of sales for decades, whose every discovery can be poured into every brand at once. L'Oréal isn't a portfolio that happens to do research. It's a research operation that happens to sell under 37 names.

The 3% that never moves

Here is the number that tells the whole story, and it's not the big one. It's the small, stable one: roughly 3% of sales, year after year. In 2023 Research & Innovation ran at 3.1% of sales; in 2024 it was 3.1% again - €1,355 million, up 5% in absolute terms but flat as a ratio, which L'Oréal itself calls 'in line with the long-term trend.'⁴ In 2025, on €44 billion of sales, the figure held at 3.1% once more.⁸ A company that lets R&D drift down in fat years and starve in lean ones never builds anything that compounds. L'Oréal treats the science budget like rent on its own future - non-negotiable, paid in good weather and bad. The absolute spend rises because the sales rise; the discipline is that the ratio doesn't.

What does that buy? In its own filings L'Oréal counts roughly 4,000 scientists across 85 nationalities, working out of 21 research centres in 7 countries by 2024.³ The company describes this, in its own words, as 'the top R&I force in the cosmetics industry.'² That superlative is L'Oréal's self-assessment, not an independent ranking - but you don't need the trophy to see the mechanism. The mechanism is what matters.

One discovery, 37 places to use it

Imagine a lone challenger brand discovers a better way to stabilise a sunscreen molecule. It bears the full cost of that research and recoups it across one product line. Now imagine L'Oréal's central labs make the same discovery. The cost is paid once, centrally, and the result flows into the mass-market line, the dermatological line, the luxury line, the professional salon line - 3,600-plus new formulas a year drawing on the same well.³ The marginal cost of putting good science into the next brand is close to nothing, because the hard part - the basic research - was already funded by the portfolio as a whole. That is the asymmetry at the heart of the moat: L'Oréal amortises invention across 37 brands; a rival amortises it across one or a handful. Same science, wildly different economics.

When the moat went and bought a startup

The clearest proof that L'Oréal thinks like a science company, not a brand company, came in March 2018. It made its first-ever acquisition of a technology firm - ModiFace, a Toronto outfit founded by Parham Aarabi roughly eleven years earlier, with about 70 engineers and researchers, more than 30 patents, and over 200 scientific publications behind its augmented-reality try-on technology.⁵ But here's the sharp part. Before the deal, ModiFace had been the quiet supplier of that AR/AI capability to much of the rest of the industry - Estée Lauder, Sephora, Unilever, Coty among them.⁶ L'Oréal didn't just buy a tool. It bought the tool out from under its competitors. The capability moved inside the wall; the rivals who'd been renting it had to start over.

Isn't 3% just a rounding error?

The honest objection is that 3% of sales is not a lot. Big pharma spends fifteen, twenty percent; software companies bury whole P&Ls in R&D. Three percent looks like a cost line you could match over a weekend, and a sceptic would say the brands - not the labs - are what people actually pay for, and that distribution and marketing muscle, not chemistry, win the shelf. Some of that is fair. A serum is not a cancer drug; the science ceiling is lower, and a brilliant ad can outsell a better molecule for years. But the objection misreads where the advantage lives. It isn't in the height of the spend - it's in the steadiness of it and the breadth it's spread across. Three percent compounded without interruption for decades builds a base of formulations, patents and methods that a newcomer can't buy at any price, because it was accumulated over time, not bought in a year. The patent count itself is climbing - from 610 filings in 2023 to 694 in 2024 - not because L'Oréal suddenly tried harder, but because a compounding base produces more.²³ The moat isn't the depth of the trench. It's that nobody else has been digging the same one for as long, across as many brands.

L'Oréal's ownership has shifted under it - Nestlé, once holding close to 30%, sold €8.9 billion of shares back to the company to trim its stake to around 20%.⁷ The brands rotate; some are bought, some sold. What doesn't move is the 3.1%. Year after year, in good markets and bad, L'Oréal pays the same fraction of everything it earns into a machine that turns chemistry into 3,600 new things and quietly hands them to every brand at once. The brands are what you see on the shelf. The moat is the part you were never meant to notice - the small, unglamorous number that has been compounding for longer than any challenger has been alive.