Sephora

Sephora is a retailer, and the distinction matters more than anything else in this case. It does not formulate, it does not own the intellectual property in most of what it sells, and it does not capture the gross margin of a beauty brand. It buys prestige cosmetics at wholesale, marks them up roughly two times, and pays for stores, staff and samples out of the difference. That is a structurally thinner business than the brands sitting on its gondolas — and a far more durable one, because no single brand going out of fashion can kill it.

What Sephora actually owns is the format. Before 1993, buying prestige fragrance in Europe meant asking a salesperson behind a glass counter, who was usually paid by one brand to steer you toward it. Dominique Mandonnaud pulled the products out from behind the glass, arranged them alphabetically by brand rather than by supplier deal, and let customers touch everything. Open-sell converted a transaction into a visit. The visit is the asset: Sephora's customers browse for forty minutes and buy three things from three brands that were competing for the same shelf.

LVMH bought that format in 1997 and scaled it. Sephora now runs more than 2,700 stores in over 35 countries, plus a shop-in-shop network inside more than 1,100 Kohl's locations in the US. Each new door raises Sephora's leverage over the brands that need to be inside it — and each year Sephora's own private label, Sephora Collection, takes a little more of the shelf it controls. That is the tension the whole case runs on.

Sephora's moat is distribution, and it is genuinely wide because it is two-sided. A prestige brand that is not in Sephora is invisible to a large share of the people who buy prestige beauty; a customer who wants to try eleven foundations in one afternoon has almost nowhere else to go. Neither side can leave without paying a real cost, and Sephora sits in the middle collecting the toll.

Scale reinforces it. Sephora's buying volume, its sampling budget and its Beauty Insider file are things a new entrant cannot assemble at any speed. When a founder launches a beauty brand in 2026, the launch plan is usually a Sephora launch plan — which means Sephora sees the entire category's product pipeline before consumers do, and can decide what gets a full endcap and what gets a shelf edge in the back.

The honest limit: this is a moat around a low-margin business, and Sephora is not the only one who knows it. Ulta has built a comparable position in the US on a broader price ladder. Amazon Premium Beauty has authorized-seller relationships with brands that once refused it. And the brands themselves keep building DTC sites specifically to reduce their dependence. Sephora's answer — put more of its own product on its own shelf — strengthens the P&L and weakens the alliance that makes the shelf worth visiting. It cannot push private label indefinitely without becoming a competitor to its own suppliers.

Sephora does not allocate its own capital in any meaningful sense — LVMH does. That makes this section less about buybacks and more about what a conglomerate parent is for. LVMH bought a 1990s French perfumery chain for a reported quarter of a billion dollars and spent the following three decades funding store openings through recessions, a retail apocalypse and a pandemic. No independent retailer could have absorbed the 2020 store closures and then opened aggressively into 2021.

The two decisions that look best in hindsight are both distribution calls. Walking away from the JCPenney partnership in 2020 and signing Kohl's in 2021 traded a declining department-store host for a healthier one, and got 1,100 shops built in four years. Sephora Collection is the more interesting call, because it is the one with a cost: every point of private-label share improves margin and slightly degrades Sephora's standing as a neutral curator. Retailers who have pushed that lever too hard — think of any supermarket chain — end up in a permanent low-grade war with the suppliers who drive their traffic.