Translating a product catalogue at scale means solving two problems at the same time: a product-data problem and a language problem. The data problem is keeping thousands of products, with all their fields and variants, flowing cleanly between your systems and each market.
The language problem is making the words read as if they were written for that market, not translated into it. Handle only one and the catalogue breaks somewhere a shopper will notice, on a size chart, a price, a product name or a claim.
This piece follows a catalogue from the moment it leaves your product information system through translation, cultural adaptation and quality checks, and out to every market, and it does so for the three catalogue types that behave very differently: online, industrial and fashion.
The thread running through all of it is scale, because at a handful of products none of this is hard, and at ten thousand it is the whole job. For teams whose catalogues are also marketing assets, this sits alongside our work on marketing translation.
TL;DR
|
What changes when you translate a catalogue at scale
At scale, a catalogue stops being a document you translate and becomes a data pipeline you localise. The same term, size or claim repeats across thousands of items, so a single inconsistency is not one mistake, it is the same mistake copied into every market you sell in. Three forces make scale hard: the sheer number of products, how often they change, and how many markets they go to at once. A seasonal refresh across ten languages is not ten times the work of one language, it is ten times the chance of drift.
The rest of this piece works through the two sides in turn, the product data and the language, then looks at how trust differs by market, and finally at how three very different catalogue types are handled. The table below is the map.
| Catalogue type | Main form | The scale challenge | What matters most |
|---|---|---|---|
| Online, PIM-driven | Web store and feeds | Constant updates across many storefronts | Continuous flow and product findability |
| Industrial / manufacturing | Print and PIM | Large, spec-dense ranges | Accuracy of specifications and units |
| Fashion / seasonal | Print and online | High turnover, many markets at once | Tone, sizing and fast turnaround |
The product-data side: systems, feeds and structure
Localising a catalogue at scale starts in the data, not the copy. Your products usually live in a product information management (PIM) system, the central place that holds each product as a set of structured fields, and their images and video often live in a digital asset management (DAM) system alongside it.
The translation workflow has to read those fields, translate them, and write them back without flattening the structure, and it has to do the same for the product feed, the file that pushes your product data out to marketplaces like Amazon or Zalando. Do that by hand and it works for one small export; at scale it is where everything jams. You can connect your PIM or shop to the translation workflow with SmartConnect so the content moves automatically when it changes.
Attributes, variants and reuse
A catalogue is mostly repetition. The same material, colour, care instruction or dimension appears across hundreds of products, and each sellable variant, each stock-keeping unit (SKU, the code for a specific size or colour of a product), reuses those shared fields. Translate a shared attribute once and reuse it everywhere through translation memory, the store of what you have already translated, and a ten-thousand-SKU catalogue stops being ten thousand translation jobs. This is the single biggest lever on cost at scale.
Feeds, marketplaces and images
The catalogue is not only the product pages on your own site. It is also the feed going to each marketplace and the images in your DAM, some of which carry text or are chosen for a specific market. A localised catalogue that updates the website but leaves the feed on last season’s wording, or shows a product shot that does not suit the market, is only half localised, and at scale those gaps multiply quietly until someone notices a wrong price on a marketplace.
The language side: drawing the transcreation line field by field
The most useful decision in catalogue localisation is made field by field, not product by product. Some fields you translate literally, some you transcreate, and some you simply reformat. Transcreation means recreating the copy so it lands in the target culture, rather than translating it word for word, and it is the right treatment for the parts of a catalogue that are meant to persuade. Getting this line right is what separates a catalogue that reads as if it was written locally from one that reads as if it was fed through a machine.
| Field type | Treatment | Example |
|---|---|---|
| Dimensions, specifications, materials | Translate literally, with controlled terminology | “Stainless steel, 24 cm diameter” |
| Product names, taglines, benefit lines | Transcreate so they land in-market | A benefit or pun that has to work in the local language |
| Prices, currencies, units, sizes | Reformat to the local convention | € and comma decimals; cm to inches; EU to UK sizing |
| Regulated or claim fields | Translate with a compliance check | Health, safety or performance claims |
Terminology is what holds this together across thousands of items. A maintained termbase, the agreed list of what each product term is called in each language, keeps a material or a component named the same way everywhere, so the catalogue does not call the same fabric three different things across three categories. AdHoc maintains that through terminology management, and it is the quiet reason a large catalogue stays coherent as it grows.
Cultural drivers of trust: why the same catalogue converts differently by market
The same translated catalogue converts differently in Germany, Sweden and Italy, because shoppers in each market decide to trust a seller in different ways. Localising at scale means adapting the signals each market reads as trustworthy, not only the words: the specifications, the payment options, the returns promise and the tone all count.
Treating Europe as one market is the quickest way to leave sales on the table, and the markets really are that different. Across EU regions in 2024, the share of people who had shopped online in the past three months ranged from about 22% to 92%, according to Eurostat, so a catalogue meeting a confident shopper in the Netherlands is meeting a very different one elsewhere.
Germany and the DACH region: verify before you commit
German shoppers (and their neighbours in the DACH region, meaning Germany, Austria and Switzerland) are cautious and like to check before they part with money. In 2024 data from bevh, Germany’s e-commerce association, buying on invoice, the “Kauf auf Rechnung” tradition of paying only once the goods have arrived and been checked, was the single most-used online payment method at 38.8%, and 68% of German consumers said easy, free returns were a priority, as reported in Rapyd’s guide to German payments.
They also lean towards sellers they see as local: German shoppers showed 89% loyalty to local web shops in Nexi’s European Ecommerce Report 2024. For the catalogue, that means flawless German, exact specifications with correct units and size charts, delivery and returns information kept visible, and pay-on-invoice offered at the checkout. A converted size that is slightly off, or a spec that is vague, does more damage to trust here than almost anywhere.
The Nordics: local and mobile-first
Nordic shoppers are also loyal to local sellers, with Swedish shoppers showing 82% loyalty to local web shops in the same Nexi 2024 report, and they pay through mobile-first methods such as Swish. The catalogue needs Swedish that reads natively, the local mobile payment options at the checkout, and a layout that works on a phone first, because that is where much of the browsing happens.
Southern Europe: reassurance and values at the checkout
In Southern Europe the reassurance comes in a different form. PayPal and cash on delivery, paying in cash when the parcel arrives, remain strong there, again per the Nexi 2024 report, and Italian shoppers are among the most values-driven, with 66% wanting merchants to reduce packaging. A catalogue selling well into Italy or Spain offers PayPal and cash on delivery, warms up the product copy rather than keeping it clinical, and surfaces sustainability points in the product content itself, where they influence the decision.
The common thread: people buy in their own language
Underneath the market differences sits one constant. In CSA Research’s 2020 “Can’t Read, Won’t Buy” survey of 8,709 consumers across 29 countries, 76% preferred to buy products with information in their own language and 40% said they would never buy from a website in another language, with German respondents among the most emphatic, at 57% buying only from local-language sites, as published by CSA Research. The figure is from 2020, but it sets the floor: native-quality language is the entry ticket, and the market-specific signals above are what turn a browse into a sale.
Quality at scale, and why some markets are least forgiving
At scale, quality is judged in-market on the details a catalogue gets wrong: prices and currencies, units and measurements, size charts, regulated attributes and culturally specific claims. The control for that is a defined process rather than a hopeful one. ISO 17100, the international standard for translation services, requires a second qualified linguist to revise the first, and ISO 18587 covers the same discipline for post-edited machine translation, where a human corrects a machine draft.
On top of both sits in-country review, a check by a specialist who lives in the target market and can see when a price format, a size or a claim reads wrong to a local. AdHoc holds ISO 17100 and ISO 18587 certification and builds in-country review into the workflow.
| What catalogues get wrong | Why it costs you | The control |
|---|---|---|
| Price and currency formatting | Looks careless, breaks trust at the checkout | In-country review, locale formatting rules |
| Units and size charts | Wrong size ordered, returned, refunded | Terminology and conversion checks |
| Regulated or safety claims | Compliance exposure in-market | ISO 17100 revision, compliance check |
| Terminology drift across items | Catalogue reads inconsistent, cheapens the brand | Maintained termbase and translation memory |
This is where the cultural point becomes an operational one. The markets that verify hardest before buying, Germany and the DACH region above all, are also the least forgiving of a sloppy catalogue, so the quality bar is set by your most demanding market, not your easiest. SmartDesk, AdHoc’s project portal, gives you the visibility to see that the checks actually happened across every language, rather than trusting that they did.
Digital, PIM-driven product catalogues
Online catalogues live in a PIM or web store and change constantly, so the localisation has to run continuously, keep each product findable in its market, and push clean data back to every storefront and feed. At scale the enemy is the update: a catalogue that was perfect at launch drifts language by language every time the range changes, unless the updates flow automatically.
Continuous updates and connectors
A store that adds or edits products weekly cannot localise in quarterly batches. Connecting the store to the translation workflow means a changed product is picked up, translated with memory and terminology already applied, reviewed, and returned to every market version, without a person exporting a spreadsheet each time. That is what keeps a live catalogue current instead of perpetually a season behind in the smaller languages.
Getting found in each market
A localised product that no one can find does not sell, so search engine optimisation (SEO, helping pages surface in search results) has to be part of the catalogue work.
That means localising product titles and attributes to the words shoppers actually use in each market rather than translating your source keywords, setting hreflang tags (the signal that tells search engines which language and region a page is for) correctly, and keeping product structured data, the code that lets search engines read product details like price and availability, intact in every language. For the storefront itself, this connects to our work on website translation.
Industrial and manufacturing product catalogues
Industrial catalogues are specification-dense and usually live in two places at once, a print or PDF catalogue and a PIM, so accuracy of units, part numbers and technical terms matters far more than tone, and the same data has to feed both outputs without diverging. Here a mistranslation is not an awkward phrase, it is a wrong tolerance or a wrong part number that sends an engineer the wrong component.
Specifications, units and part numbers
The core of an industrial catalogue is precise: dimensions, tolerances, materials, standards and part numbers, translated with controlled terminology and converted to the right units for each market. This is exacting, low-tone work, and it is the part where the German market’s expectation of exactness, described earlier, bites hardest.
One source, print and online
The efficient way to run an industrial catalogue is to localise the product data once and let both the print layout and the online catalogue draw from it, rather than translating the InDesign file and the web store separately and hoping they match. Running both from one localised source is also what keeps the printed spec and the online spec identical, which for a technical buyer is a basic test of whether they can trust you.
Fashion lookbooks and seasonal catalogues
Fashion catalogues are the opposite case: they turn over every season, lean heavily on tone, and need more transcreation than translation, plus tight turnaround across many markets at once and careful handling of sizing. A lookbook, the seasonal styling catalogue that sells a mood as much as a product, only works if the copy feels native and current.
Transcreation and seasonal turnaround
Product names, styling copy and campaign lines have to be recreated so they land in each market, not translated literally, and they have to be ready across markets fast enough to launch the season everywhere at once. This is where translation memory and terminology do the heavy lifting on the repeating parts (fabrics, care instructions, sizing) so the linguists can spend their time on the copy that actually needs a creative hand.
Sizing, fit and cultural fit
Sizing systems differ by market and a wrong conversion drives returns, which are expensive and, as the German data showed, expected to be free. Beyond sizes, the styling language and imagery have to suit the market, since a claim or a look that flatters in one country can fall flat in another. Getting this right is what makes a season feel designed for the market rather than shipped to it.
Frequently asked questions
How do you translate a product catalogue at scale?
You treat it as a data pipeline, not a document. Connect the product information system to the translation workflow, apply translation memory and terminology so repeated content is translated once and reused, decide field by field what to translate, transcreate or reformat, and add in-country review so quality holds across markets. Handling the data and the language together is what makes scale manageable.
What is the difference between translating and transcreating product content?
Translation carries the meaning across accurately, which is what specifications, dimensions and materials need. Transcreation recreates the copy so it persuades in the target culture, which is what product names, taglines and benefit lines need. A good catalogue draws that line field by field rather than treating every field the same way.
How do you keep terminology consistent across thousands of products?
You maintain a termbase, an agreed list of what each product term is called in each language, and apply it across every product, supported by translation memory. That stops the same material or component being named three different ways across categories, which is what makes a large catalogue read as one coherent whole.
Can you localise both print and online catalogues from the same source?
Yes, and it is the efficient way to do it. Localising the product data once and letting both the print layout and the online catalogue draw from it keeps the two identical and avoids paying to translate the same content twice. It is especially valuable for fashion and industrial catalogues, which routinely exist in both forms.
How do you handle sizes, units and currencies in a localised catalogue?
These are reformatted to each market’s convention rather than translated: currencies and number formats, measurement units, and size systems all change per locale. Getting them right matters for conversion and for returns, since a wrong size conversion is one of the most common and expensive catalogue errors.
Should our PIM or a translation partner own catalogue localisation?
Both, in their lane. The product information system holds and moves the data; the language partner supplies the terminology, transcreation, in-country review and quality process that the system cannot. The two connect, so the data flows automatically and the linguistic quality is handled by people who do it for a living.
Book a workflow review
If your catalogue is growing faster than your team can keep it consistent across markets, a workflow review is the practical next step. Book a workflow review or see how we structure your multilingual content, and we will map your catalogue, your markets and the connections that keep them current.
Sources
Nexi and Teleperformance, European Ecommerce Report 2024 (market and payment differences). https://www.nexigroup.com/en/media-relations/news/2025/07/european-ecommerce-report/
bevh (German e-commerce association) 2024 payment data, reported in Rapyd, A Guide to Navigating Germany’s Payments Ecosystem. https://www.rapyd.net/blog/germany-payments-guide/
Eurostat, online shopping across EU regions, news article, 28 November 2025 (2024 data). https://ec.europa.eu/eurostat/web/products-eurostat-news/w/ddn-20251128-2
CSA Research, “Can’t Read, Won’t Buy” survey, 2020 (native-language preference). https://csa-research.com/l/media/Consumers-Prefer-their-Own-Language
ISO 17100:2015, Translation services; ISO 18587:2017, Post-editing of machine translation output.