One Catalog, Twelve Markets: Localizing Product Data at Scale

Ask a brand why they're live in three markets instead of twelve and you'll hear about logistics, tax, maybe brand awareness. Push harder and the real answer surfaces: every new market means re-doing the entire catalog, and nobody can face doing it again.

The demand is usually there. The 3PLs and payment providers are solved problems. What actually caps expansion is the unglamorous grind of making 8,000 SKUs correct per market, and keeping them correct as the catalog changes daily.

What "Localized" Actually Means

Translation is the part everyone plans for, and it's maybe a third of the work.

Language, but commerce language. Machine-translating listings produces copy that reads fine and sells nothing, because product search is vocabulary-specific. A "jumper" listed in Germany needs to be a Pullover, not a literal translation, and the title needs the terms German shoppers actually type. Localizing search terms, not just sentences, is what moves rank.

Sizes and measurements. A women's US 8 is a UK 12, an EU 40, and an IT 44. A duvet is measured in inches in one market and centimetres in the next, against different standard bed sizes. Get this wrong and the cost isn't just lost sales. It's return rates. Cross-border fashion returns already run 30%+; size confusion is a leading driver.

Currency and price presentation. Not just conversion: psychological price points (€49,95 vs £44.99), VAT displayed inclusive or exclusive by law, and per-market price positioning that doesn't leak across borders.

Compliance. This is where manual processes genuinely break. EU textile listings need fibre composition in the local language. Germany requires packaging-licence registration; France mandates repairability scores on electronics; the EU's GPSR requires a responsible-person address on listings. These are per-market, per-category data requirements, they change, and a marketplace can suppress your listings for missing any of them.

Imagery and content norms. Lifestyle photography that works in Stockholm reads wrong in Riyadh. Some markets expect size charts as images; others require energy labels rendered a specific way.

The Arithmetic That Kills Expansion

Run the numbers on doing this by hand: 8,000 SKUs × 11 additional markets × translation, size mapping, price rules, and compliance fields per SKU. Even at an optimistic ten minutes per SKU per market, that's roughly 15,000 hours (seven person-years) before the first sale. So brands do the sensible manual-era thing: launch two markets, badly, and stop.

Worse, localization isn't a project. It's a treadmill. Every new product, every copy revision, every price change has to propagate to twelve localized versions. Manual processes don't just make expansion slow; they make maintaining expansion impossible, which is why market five always launches with market one's stale content.

One Master Record, Twelve Renderings

The scalable model treats each market listing not as a copy to maintain but as a rendering generated from one master record.

That's the architecture Qooly runs. The intelligence layer holds the canonical product (attributes, imagery, copy, pricing logic) and generates each market's version: translation tuned to local search vocabulary, size and unit conversion from structured attributes, per-market price formatting and tax presentation, compliance fields validated against each market's current requirements before publishing. When the master record changes, all twelve renderings update. When France adds a labelling rule, it's a validation update, not a 3,000-SKU manual audit.

The honest test of your localization model: how long would market thirteen take? If the answer is a quarter and a project team, you don't have a localization process. You have a catalog-rebuilding process with a French accent.

Expansion capacity is a product data property. Brands with one clean, structured, machine-actionable catalog can treat new markets as configuration. Everyone else pays the seven person-years, market by market, forever.