ETQ Amsterdam

    ETQ Amsterdam

    PREMIUM FOOTWEAR · E-COMMERCE

    452 products in 9 languages, automatically on brand voice

    From 15 minutes of manual work per product to 30 seconds, for all 9 languages at once. For roughly 1.5 cents per translation. And it gets better every month.

    INDUSTRY
    Premium footwear
    PRODUCTS
    452+
    LIVE SINCE
    Q2 2026
    STYLE
    Own tone-of-voice

    The dashboard where English product content is automatically picked up and translated into 8 languages.

    KEY RESULTS
    30 sec
    9 languages, in one go
    Was: 15 min per product
    €0,015
    Per translation
    3,600 = €54 per cycle
    Zelflerend
    Sharper every month
    Approvals = training

    What didn't scale before.

    ETQ sells in 9 languages across a catalog of 452 products. English product content had to be translated by hand or by an agency into 8 languages, and every source change kicked off the same loop again.

    Tone of voice slipped along the way. Metafields shared across dozens of products were retranslated every single time.

    A dashboard that tracks the English source and translates automatically.

    A Shopify dashboard where English product content is automatically translated by Claude into 8 languages, with ETQ's tone of voice baked into the prompt.

    Reviewers approve in one click, and the system pushes straight to Shopify Translate & Adapt, including metafields, materials and premium features.

    BEFORE· English source, Dutch empty
    AFTER· Translated, ready for approval

    Left: a product with only the English source. Right: the same product page after the system added the Dutch translation, ready for one-click approval.

    No generic translations. ETQ's voice in 8 languages.

    Brand examples, language-specific pitfalls and style guidelines are baked into the prompt. Every translation sounds like ETQ. Not like Google Translate, not like a freelancer.

    The prompt configuration with brand examples, language-specific pitfalls and style guidelines built in.

    A single material description propagates automatically across every product that shares it.

    A material description used on 33 products? Translated once.

    Shared content is translated once and propagates everywhere, across all 8 languages, on every product that uses it. Any change to the source follows automatically.

    WHAT IT DELIVERS

    Per catalog: 109 hours gone, no more translation agency, a system that sharpens itself.

    452 products × 15 min by hand = 113 hours per catalog. With the system: 30 sec per product = 3.8 hours. AI cost: €54 per catalog.

    From 113 hours to 3.8 hours

    452 products in 9 languages. Manually 15 min per product, with the system 30 sec.

    €54 in AI costs per catalog

    No translation agency, no freelancers. Roughly 1.5 cents per translation.

    Every approval is training

    Reviewers approve, the system learns ETQ's voice more deeply. Month over month, fewer corrections.

    ROADMAP

    Roadmap

    Not a multi-year project. Prove it first, then scale.

    Week 1
    Analysis

    452 products mapped, 8 target languages confirmed, every metafield inventoried.

    → Architecture and scope locked in.

    Week 2-5
    Build

    Dashboard built, prompt engineering with brand examples, Shopify integration, review flow.

    → Working system, ready for the first batch.

    Week 6
    Live

    First batch of products in production.

    → Reviewers work inside the system, translations go live.

    It used to take weeks to launch a product because we were waiting for translations. Now that happens in a morning. My team can focus on the markets themselves, not the logistics around them.

    Arjan Spakman, Head of Marketplaces at ETQ Amsterdam
    Arjan Spakman
    Head of Marketplaces · ETQ Amsterdam

    What ETQ built in 6 weeks, we can build for you too.

    Also translating product descriptions every month for a new collection?

    ETQ went from 113 hours of manual work per catalog to 3.8 hours — no translation agency. What is your team doing manually that a system could take over?