The Very Group runs Very.co.uk and Littlewoods — 4.7 million active UK customers and, critically, a financial services arm that keeps working-class British shoppers cycling through fashion on credit. The reported £2 billion interest isn't about Very's catalog. It's about the payment rails and the household data. JD already owns Ochama in the Netherlands; this would be a second Western anchor. The Chinese platform wars are running out of room at home — Pinduoduo's price compression, Douyin's commerce ambitions, Alibaba's defensive capex — and the logical exhaust valve is Western acquisition. We wrote about that competitive scramble in May. What we're watching now is the international chapter of the same story.

Vestiaire Collective's pre-owned luxury inventory is now surfaced directly on Zalando — no platform-hop, no separate checkout. For Zalando this is infrastructure capture: rather than building authentication, provenance tracking, and luxury seller relationships from scratch, it outsources all of that to Vestiaire and keeps the customer on-platform. Smart allocation given Zalando already spent €1.13 billion absorbing ABOUT YOU to scale the primary market. The question is margin: pre-owned luxury fees are structured differently from fast-fashion transaction cuts, and Zalando is now holding both economics simultaneously. We covered Zalando's scale strategy versus Amazon's protocol bet last month — this partnership fills in the pre-owned tier that acquisition alone couldn't deliver.

The secondhand market's best customer is also its least forgiving. Gen Z drives resale volume but expects returns to match what they get from ZARA — and when one goes badly, they leave loudly and take the platform's reputation with them. This is the tension today's piece traces directly. The brands building resale programs as tariff hedges or sustainability positioning — including Steve Madden, which reframed its 2022 resale platform as a tariff hedge — need to understand that the returns experience is now load-bearing for the brand relationship, not an ops footnote. The Zalando-Vestiaire deal in the entry above is the supply-side expansion; Gen Z's zero tolerance for friction is the demand-side constraint. Both are live simultaneously.

FashionUnited's street-level Beijing piece is worth reading slowly: the shoppers who used to visit Galeries Lafayette had already moved to Dewu, Xiaohongshu, and brand DTC apps years before the closure announcement. The department store didn't lose them; consumers reclassified what a shopping destination is required to be. Today's piece from Neritus Vale is the full account of the departure. The structural note: luxury stopped needing a marble atrium to convey legitimacy in China when the social-proof layer migrated to Xiaohongshu. That happened around 2021. The physical exit is the lag catching up.

FashionUnited's China fashion pulse for April: clothing up 1.5% year-on-year, domestic input costs climbing. European prices heading the other direction. The divergence is the tariff story running in reverse — Chinese producers pass costs homeward while export surplus heads west via rerouted supply chains at a discount. Sir John Crabstone's piece today works through the full spread.

Le Prix de l'Innovation ANDAM 2026 a été attribué à Alphalyr, une plateforme SaaS française qui fournit des analyses e-commerce en temps réel aux marques de mode. The French trade press frames this as AI arriving in the fashion value chain, which oversells it slightly — Alphalyr is advanced BI more than generative AI — but the symbolic weight of the decision matters regardless. ANDAM spotted Jacquemus at €15,000 in 2014, Marine Serre in 2017. When this committee hands its innovation prize to a data platform, it is saying that understanding your own numbers is a competitive craft, not an administrative function. That is a meaningful reframe for an industry that still runs a significant share of buying decisions on instinct and sampling. The French fashion establishment does not give prizes to spreadsheet companies unless it has decided that spreadsheet companies deserve to be taken seriously.

European shopping center operators are formally replacing footfall as their primary KPI with visit quality: conversion rate, dwell time, spend per visit. Long overdue. Footfall was always a landlord metric — useful for lease negotiations, not for understanding whether a store is actually performing. A fashion retailer with 500 deeply engaged shoppers and 30% conversion is structurally healthier than one with 5,000 passersby at 3%, but only the new metrics say so. The AI angle: in-store analytics tools that measure dwell time, heat maps, and conversion have dropped in cost dramatically — there is no longer an operational excuse for a mid-size retailer to be flying blind on floor performance. Today's piece looks at how stores are retooling the full measurement stack.

German trade press etailment.de is raising the question Anglo-American agentic commerce coverage keeps treating as tomorrow's problem: when an AI agent completes a purchase on behalf of a human, how does the fraud system determine whether it's a legitimate agent, a hijacked one, or an attacker impersonating one? Current fraud stacks were built to detect anomalous human behavior — click velocity, device fingerprints, behavioral biometrics. An AI agent looks anomalous on every one of those dimensions by design. The piece, by Pascal Gimmig, calls for a new checkout layer identifying not just "is this a human" but "is this an authorized agent acting within scope." This is not theoretical. Agentic commerce is being deployed on the assumption that trust infrastructure will catch up. It hasn't. We wrote in May about the benchmark gap for retail AI apps; the fraud layer is the security-side version of the same missing infrastructure.

Alphabet's AI Health Coach, announced May 19, pulls together wearable data, search history, and health records into personalized wellness guidance. Glossy is covering it as a wellness product. The more accurate frame: it's the most ambitious consumer health data aggregation since Apple Watch, and considerably more powerful because Google's search context is in the stack. For fashion and beauty retail, the implications are non-trivial. A shopper whose health coach knows she's training for a marathon, monitoring iron levels, and sleeping five hours is a radically different personalization target than someone who browsed "running shoes" once. The brands that figure out how to operate in that data environment without triggering privacy backlash will have a structural edge. The ones that get it wrong will become the case study. We documented the AI marketing trust gap in May; Google Health will stress-test it at a scale nothing previously attempted.

The BoF opinion piece argues that AI shopping agents will compress the discovery phase of retail — the browsing, the impulse path, the ambient brand awareness that costs billions to manufacture — into an outcome-optimized query. The agent doesn't browse. It executes. For brands built on aspiration and visual storytelling, this is the real threat: your entire marketing budget buys awareness the agent will bypass on its way to the cheapest compliant option. The brands that survive agentic commerce are those whose products are either defensibly differentiated on measurable attributes (verified materials, fit data, authenticated reviews) or whose brand signal is strong enough to become a user preference the agent must respect. Middle-market brands with undifferentiated product and high marketing spend are most exposed. Today's piece on Amazon's AI shopping advantages over rivals who rent the same tools is the supply-side read on the same structural shift.

Alana Pallister sold her house to fund 300,000 units of the Blare tracksuit. Under five months later she'd moved a million. The demand model was social velocity, not forecast — one TikTok moment, one price point, one SKU. Now she's trying to engineer the next moment at Coachella. The problem: organic virality can't be scheduled. Today's piece on Pacsun is the useful counterpoint — a brand that built a systematic teenage demand model rather than riding one wave. Both approaches work. The I.AM.GIA Coachella bet will tell us whether the tracksuit was a business or an event.

A $660 perfume going viral and the dupe conversation that followed pushed Guerlain — founded 1828, over 1,100 fragrances in the archive — into its first-ever paid creator deal. The trigger is what's interesting: the brand didn't move because influencer marketing was obviously good. It moved because dupe culture generated a cultural relevance signal that no other channel had produced, and someone noticed that the signal was free and the commercial capture was not. This will repeat. Heritage fragrance houses, archival leather goods, fine jewelry — brands that have maintained relevance through reputation alone will gradually learn that dupe conversation is cheap proof of demand, and that paid creator work can capture value before the dupe captures market share.

The new House of Dior Shinsaibashi in Osaka wraps its exterior in a ribbed structure Fujimoto describes as evoking flowing drapery, with Peter Marino on the interior. Beyond the architecture, the geography matters: Dior is making serious structural investments in Japanese retail while simultaneously reducing exposure to the wholesale-and-department-store model globally. Osaka is the opposite of the Galeries Lafayette exit — it's a brand declaring that its physical presence must be a controlled environment, not a concession on someone else's floor plan. When you can afford Sou Fujimoto for your facade, the building is the brand statement.