ba&sh runs Kleep head-to-head against True Fit, and never looks back

Founded in Paris in 2003 by childhood friends Barbara Boccara and Sharon Krief, ba&sh set out to design the wardrobe they couldn't find: effortless, feminine pieces with a distinctly Parisian ease, fluid dresses, soft tailoring, knits that move. Two decades on, that wardrobe travels far beyond France. ba&sh has grown into a global brand with boutiques across Europe, North America, the Middle East and Asia, a thriving digital flagship, and a resale programme that speaks to how long its pieces stay in circulation. The house is also unusually rigorous for a brand built on ease: it was an early mover on digital experience, and its e-commerce team treats the product page with the same care the studio gives a silhouette. Fluidity, though, is a double-edged sword online, the drape that makes a ba&sh dress effortless in person makes it genuinely hard to size from a flat chart on a screen.

ba&sh was not starting from zero. The brand already ran True Fit, one of the biggest names in the sizing category, so the question was never whether to guide shoppers on fit, it was whether the incumbent was actually the best answer available. Fluid, feminine silhouettes are among the hardest garments to recommend: wrap dresses, soft knits and relaxed tailoring depend on drape and intended ease, not just body measurements, and a tool that hedges gets ignored. And a fit tool only works if shoppers actually open it, usage is the metric that precedes every other metric. Rather than let inertia decide, ba&sh did what remarkably few brands do with an installed vendor: it put the incumbent on trial. Same storefront, same period, live traffic split between True Fit and Kleep, with usage, add-to-cart and downstream behaviour measured on both arms. Whoever won the data would keep the storefront.

Kleep deployed on one arm of the A/B test with its full stack: a short questionnaire or the two-photo body scan to build each shopper's measurement profile, per-garment calibration so a fluid wrap dress and a fitted knit are modelled as the different sizing problems they are, and a white-label interface drawn in ba&sh's own visual language so the guidance reads as the brand speaking, not a third-party widget. Recommendations return in under 250 milliseconds, keeping the flow inside the natural rhythm of browsing. Crucially, the methodology was the incumbent's own game: True Fit kept serving the other arm exactly as before, so every difference in the numbers was attributable to the engine, not to novelty or placement. ba&sh's team simply watched the dashboard as live traffic voted with its behaviour, the cleanest possible answer to the question they had asked.

The test settled a debate that could have lasted months. Against the True Fit arm, Kleep drove 20% more shoppers to actually use the fit guidance, the adoption gap that decides whether a sizing tool earns its place, and lifted add-to-cart by 88%. An 88% gap on the metric closest to purchase intent is not a rounding difference between comparable tools; it is a different product. The incumbent left, and the numbers stayed. The result also carries weight beyond ba&sh: True Fit is one of the category's global reference vendors, and ba&sh measured the difference on its own traffic rather than taking anyone's word for it. It is the same pattern Kleep has since repeated against Fit Analytics at Intersport and against Fitle at Jules, head-to-head, on live traffic, decided by the client's own data. Across more than 300 live brands, that remains the only methodology Kleep asks anyone to trust.