Showroomprivé enhances online shopping with AI-powered sizing

Founded in 2006 by Thierry Petit and David Dayan, Showroomprivé is one of France's largest e-commerce platforms, a flash-sales pioneer that turned end-of-season inventory into a daily event for tens of millions of members. The model is speed at scale: branded sales open with deep discounts, run for a few days, and vanish, with thousands of brands rotating through the platform across fashion, beauty, home and travel. That machine made Showroomprivé a listed company and one of the defining names of French digital retail. It also made the company a uniquely demanding operating environment for any technology partner: catalogues appear and disappear in days, assortments span every fashion category and price point, and the volumes are among the largest in the country. Nothing survives on Showroomprivé that requires manual care and feeding, a tool either works automatically, at flash-sale speed, or it doesn't work at all.

Flash sales concentrate every sizing problem in e-commerce and add a clock. Each sale brings a new brand with its own size grid, its own cut philosophy and its own quirks, multiplied by thousands of brands a year, with no time for the manual, per-brand onboarding that legacy sizing tools assume. Shoppers, meanwhile, are buying unfamiliar labels at speed, often in categories they don't usually shop, which makes 'order your usual size' the worst possible advice. The result, at Showroomprivé's volume, was a return rate with industrial consequences: every mis-sized parcel eroding the margin that discounting had already compressed, and reverse logistics scaling with success. The platform needed sizing guidance that could calibrate to a brand-new catalogue automatically, serve heterogeneous audiences from first-time members to power users, and prove its impact on the only metric that mattered, parcels that don't come back.

Showroomprivé went live with Kleep in January 2025, deployed for the realities of flash retail. Calibration is automated at catalogue speed: as each sale's products land, Kleep's models adapt to the incoming brand's sizing logic without manual setup, so guidance is ready when the sale opens. Shoppers get a dual-model flow, matching the platform's breadth of audience, the two-photo body scan for members who want measurement-grade precision, or a quick questionnaire for those mid-rush on a phone, with recommendations returned in under 250 milliseconds so nothing slows the treasure-hunt rhythm the model depends on. In March 2026, the partnership deepened into the hardest category on the platform: a purpose-built lingerie flow, modelling band, cup and sister sizes rather than treating a bra like a T-shirt, rolled out across the relevant sales. One engine, every brand, every category, at the pace of a platform that never runs the same shop two weeks in a row.

The headline number is public: Showroomprivé's return rate fell by 11% with Kleep live, reported in the press, and industrial in its implications at the scale of one of France's largest e-commerce platforms. Fewer returned parcels means recovered margin on discounted goods, lighter reverse logistics, and inventory that stays sold, the compounding economics that decide whether a flash-sale model thrives. The partnership's direction of travel says the rest: a platform that could choose any vendor extended Kleep into lingerie in March 2026, trusting it with the category where sizing tools are most brutally exposed. Showroomprivé's deployment is also proof of the engine's range, the same technology that sizes couture pieces for Givenchy calibrates a rotating catalogue of thousands of mainstream brands, automatically. Across Kleep's portfolio of 300+ live brands, serving a sizing flow every five seconds in 2026, Showroomprivé remains the scale test: flash retail leaves nowhere for a sizing engine to hide.