What Happens When 2,000 Products Need to Speak Twelve Languages

Managing thousands of products in multiple languages sounds like a logistics nightmare. For most businesses, it is. Specifications change, translations fall out of sync, and the people responsible for keeping it all together spend their days buried in spreadsheets instead of doing work that actually moves the business forward.

The reality of managing multilingual product data

Most companies start the same way. Product information lives in a spreadsheet. Someone adds a French column, then a German one, then Dutch. Before long, the file is 15,000 rows deep with merged cells, colour-coded tabs, and a naming convention that only one person understands. When a specification changes, someone has to find that value in twelve different columns and update each one by hand.

Translations tend to be managed through email. The product team sends a batch of English descriptions to a translation agency, who return the files a week later. Those translated strings then get pasted into the spreadsheet, or into a CMS, or into both. Nobody is entirely sure which version is current. When a dealer in Italy spots an error on their product page, the fix request bounces between three departments before anything changes.

The real cost is not the translation itself. It is the coordination overhead: the chasing, the version control, the slow creep of inconsistency across every market you operate in. One wrong kilogram value in one language version and suddenly a dealer is quoting the wrong specifications to a customer.

A project that made us rethink the problem

When EP Equipment came to us, they were dealing with exactly this situation at scale. EP Equipment is one of the world's largest manufacturers of lithium-powered material handling equipment. Their catalogue spans electric forklifts, pallet trucks, stackers, and warehouse automation solutions, sold through dealer networks across more than 20 markets worldwide.

They had over 2,000 products and translations to manage across twelve languages. Product data was scattered across spreadsheets, ERP exports, and shared drives. Translations were coordinated through email threads. Launching a single new product across all language versions took weeks. The marketing team was spending more time on content administration than on actual marketing.

What struck us was that EP Equipment's problem was not unusual. They just had enough scale for the pain to be impossible to ignore. Most businesses dealing with multilingual catalogues face the same fundamental issues. EP Equipment simply could not afford to keep solving them with manual processes.

What actually made the difference

Building a PIM is the easy part to describe. You put all the product data in one place and connect it to the website. But the features that actually changed EP Equipment's day-to-day were more specific than that.

Auto-translate was the single biggest time saver. Previously, getting a new product description into twelve languages meant sending content to a translation agency, waiting for the files to come back, and then pasting each translation into the right place. With auto-translate built into the PIM, the content team can generate draft translations instantly and then route them through a human review process. The translation agency still reviews everything, but they are reviewing and refining rather than starting from scratch. A product that used to take weeks to launch across all markets now goes live in hours.

AI-powered search solved a problem we had not originally scoped for. EP Equipment's dealer network needed to find products quickly, but their catalogue is large and technical. Searching for a part number is fine if you know it. Most dealers do not. They know they need an indoor pallet truck rated for a certain weight. The contextual search we built understands that kind of query and returns the right results, not just keyword matches.

Dynamic PDF datasheets eliminated an entire category of maintenance work. Previously, every product had a separate PDF datasheet for each language. Updating a single specification meant regenerating and re-uploading dozens of files. Now, datasheets are generated on demand from live PIM data. They are always current, always branded correctly, and available in any language without anyone touching a file.

Three things we learnt the hard way

Every project teaches you something. This one taught us three things we now bring into every similar engagement.

Data migration is the hard part, not building the system. We budgeted a reasonable amount of time for cleaning and importing EP Equipment's existing product data. It was not enough. Years of spreadsheet-based management had produced duplicates, inconsistent naming, missing fields, and translations that no longer matched their source text. Cleaning that data took as long as building the PIM itself. If you are planning a similar project, double your data migration estimate and you will probably still be slightly optimistic.

Translations need workflow, not just a field. Our first instinct was to add language fields to every product attribute. Twelve languages, twelve columns, done. But that misses the point entirely. What matters is the process around those fields. Who writes the source text? Who triggers the translation? Who reviews it? What happens when the source text changes after the translation is already approved? Without a proper review workflow, you end up with twelve language versions and no confidence that any of them are correct.

The content team has to want it. We have seen beautifully engineered systems gather dust because the people who were meant to use them daily were not involved in designing them. With EP Equipment, we ran workshops with the content editors early in the project. We watched how they worked, asked what frustrated them, and built the interface around their actual process rather than an idealised one. The result is a system that the team genuinely prefers to the old way of working. That matters more than any feature list.

Is this just for manufacturers?

EP Equipment is a manufacturer, so the examples here are about product specifications and datasheets. But the underlying problem is not specific to manufacturing at all. Any business managing a large catalogue of structured content across multiple markets or languages faces the same challenges.

We have seen it with talent agencies managing thousands of performer portfolios across different regions. With dealer networks maintaining localised product listings for each territory. With franchise businesses that need consistent branding but market-specific content across dozens of locations. The details change, but the pattern is always the same: structured data, multiple audiences, and a team spending too much time on coordination.

If your team is spending more time managing content than creating it, the tools are probably the bottleneck, not the people.