6 lessons for scaling on European marketplaces

UFESA’s case study describes 2023 as a “learning phase”: selling on marketplaces, but informally, without defined processes. Eminence’s team frames their early marketplace work in similar terms. Singularu had “several frustrating early attempts to scale.” In all three cases, the inflection point was the same: a deliberate decision to treat marketplaces as an operational channel rather than an experiment, with the systems and metrics to match.

This is the unglamorous first move and the one that gets skipped most often. Brands run pilot marketplaces with a spreadsheet, a part-time owner and no SLA on listings. Two years later, the same setup is generating real revenue and quietly breaking every time a marketplace updates its category taxonomy.

All three brands made the same first technical move: collapse fragmented marketplace integrations into a single feed engine. Eminence is the cleanest example – 13,689 SKUs across six brands, distributed through eight French marketplaces and one Belgian platform, all managed through one tool. UFESA centralised catalogue, pricing and stock for its multi-marketplace Spanish strategy. Singularu, working without an ERP and using Shopify CMS as the source of truth, built its entire expansion on a single centralised feed enriched per channel.

The centralisation point isn’t really about technology, but about what becomes possible once the catalogue lives in one place: channel-specific rules, marketplace-specific transformations, exclusions, mappings, promotional campaigns timed across multiple platforms at once. None of this is feasible when your “system” is six CSVs and a shared inbox.

Eminence reports launching a new marketplace channel in approximately one day (when they have been accepted previously by the marketplace in question). Singularu launches new countries in hours and new channels in days or weeks. These are not vanity numbers. They translate directly into the strategic luxury of testing more, failing faster, and doubling down on what works without a six-month integration sunk cost hanging over the decision.

For comparison: a custom integration with a marketplace, done from scratch, can easily run two to three months of engineering time. A brand operating at that pace gets perhaps three marketplace bets a year. A brand operating at Singularu’s pace can run a dozen and still kill the ones that don’t perform. That is a different business.

This is the strategically richest insight from the three case studies, and it comes mainly from Singularu. The team explicitly uses marketplace performance data to decide which countries to enter on their own e-commerce site, which assortments to push, and which categories to prioritise. Marketplaces, in their model, are a low-risk, high-data way to test a market before committing direct-to-consumer infrastructure to it.

The Zalando Germany launch is the textbook case. Beyond the marketplace sales themselves, the channel “significantly increased brand visibility and drove qualified traffic back to Singularu’s own website”, accelerating first purchases on the D2C site in a market they had not yet built for. This is the halo effect that brands keep theorising about and rarely instrument. Singularu instruments it.

UFESA does a quieter version of the same thing with Buy Box and competitor data through NetRivals, watching how its own distributors price its products on Carrefour, El Corte Inglés, Fnac, MediaMarkt, PC Componentes, Worten and Amazon.

Singularu’s team breaks down its marketplaces by customer archetype: Veepee buyers are bargain- and deal-driven; El Corte Inglés shoppers are classic and selective; Amazon attracts practical, last-minute purchasers with a higher male share; Zalando draws fashion-led, digital-native consumers. This is the kind of behavioural mapping that most brands assume someone else has done for them.

The practical consequence is that assortment, pricing, copy and even photography should differ by marketplace, and not by a token A/B test, but as a fundamental operating principle. A discount-positioned hero SKU on Veepee may not be the same SKU you want fronting your Zalando shopfront. Eminence treats Black Friday campaigns this way, adapting offers simultaneously across multiple channels rather than running one promo and hoping. UFESA applies “assortment exclusions by channel” to keep low-margin products off marketplaces with unfavourable economics, and to protect new launches from premature exposure. The thread is consistent: each marketplace gets its own version of the brand.

A footnote worth surfacing because it shows up in two of the three case studies and matches what the underlying market data has been signalling all year: the European off-price model is having a moment. Eminence’s marketplace growth is “driven in particular by Veepee, which is now the leading contributor to Eminence’s marketplace revenue.” Singularu names Veepee as one of its four core channels and has built specific assortment logic around its deal-hunting audience.

For brands still in marketplace-selection mode, the implication is straightforward: Veepee, ShowroomPrivé and Privalia deserve a serious look, especially for stock rotation, promotional pushes and reaching audiences that Amazon and Zalando don’t convert. The off-price tier has quietly become a strategic layer of European marketplace strategy, not a clearance dump.

Three brands. Three categories. Three different country footprints. The same conclusion: marketplace scale is not won by adding more channels. It is won by treating the channel itself as a real product line – with centralised data, fast launches, per-channel logic, margin discipline and a feedback loop into the rest of the business.

The lessons: stop running marketplaces like a side project, get the catalogue under one roof, optimise per channel rather than at the channel, and use what marketplaces tell you about demand to inform decisions far beyond the marketplaces themselves.

For brands and retailers looking at their own marketplace P&L and wondering why it has plateaued, the answer is almost never the marketplace. It’s usually the operating model around it. UFESA, Eminence and Singularu are three different proofs of the same point.