Translated or English creatives to scale abroad: how to decide

We had this exact debate not long ago with a jewellery brand we work with. We asked ourselves a question that sounds dead simple on paper. Translate the creatives market by market, or keep everything in English and run the same stuff everywhere.

And behind that, there's a real trade-off. With time, money and performance on the line. Here's how I see it.

The starting point

To really crack a market, you need creatives in the local language. Spanish creatives for Spain, Italian for Italy, German for Germany. That's the baseline principle.

The problem is, this brand didn't have any. They were running English everywhere, across every market. And when we looked at the numbers, the English creatives clearly weren't the top performers in non-English-speaking markets. No surprise there. You're talking to an Italian who only sort of gets English, of course your hook lands softer.

Quick note while we're at it, translating a creative and adapting a creative are two different things. Translating means taking your message word for word into another language. Adapting means reworking the hook so it actually speaks to the local culture. On short formats like a Meta ad, adaptation is what counts, not the literal translation. But we'll get to that below.

The thing is, the answer really comes down to two factors. First your product. Then your market.

It depends first on your product

Jewellery is a visual product you understand at a glance. People don't necessarily need a creative in their language to grasp what they're buying. They see a ring or a necklace, they get it instantly. The photo does 80% of the work.

In eCommerce, you're mostly buying the photo. You judge the product through it. And with jewellery, that's even more true. The copy becomes almost secondary.

Now take a technical cosmetic. That's a whole other story. You've got active ingredients, benefits, a routine to explain, specific vocabulary. On that kind of product, language matters enormously because you have to make people understand why your thing is different. An English creative explaining a formula to a German lands way worse.

So basically, obvious product, English is enough. Product you need to explain, switch to the customer's language or you're leaving sales on the table.

And the market changes everything

Even with the same product, the right call shifts depending on the market. Here are two concrete examples we've seen on our clients.

In Italy, Italian creatives performed clearly better. Italians respond well to content in their own language, it's plain in the numbers. In Portugal though, English creatives worked really well, because there are tons of foreigners and expats, a population already used to English. So translating into Portuguese didn't add much.

So you can't just decide in your corner that "we translate everything" or "we keep everything in English." You have to look market by market. What works in Italy can fall flat in Portugal.

Keep English or translate, how to decide creative by creative

To avoid staying stuck in the grey zone, here's how I approach it when I have to decide whether to leave a creative in English or switch it to the local language.

I start with the product. If it's visual and obvious, English usually works. Technical, I translate. Then I look at the market, in an expat-heavy country where people already live in English day to day, English holds up, elsewhere it doesn't. Then there's the tone of the hook, and that's often where it breaks, because a pun or a punchy line translated word for word ends up clunky. Either you adapt it so it still makes sense in the local language, or you switch angle entirely.

When in doubt, you test. We'll get there.

The thing everyone forgets, creative/site consistency

This one, I see it all the time and it drives me crazy.

You can have the most beautiful Italian creative in the world. If it links to a site in English, you've broken everything. Picture an Italian who doesn't speak English, they click your ad in their language, all happy, land on your product page and suddenly nothing speaks their language anymore. That's super frustrating. They bounce. And you just burned your acquisition budget for nothing.

With that jewellery brand, we found things in real time as we went. The jewellery personalisation was still in French, the language selector was buggy, and the site showed prices in dollars by default. For an Italian trying to buy, that's a lot of friction all at once.

So before you even ask yourself what language your creatives should be in, make sure your site is fully translated. The whole funnel, not just the product page. From product page to checkout, including personalisation and prices in the right currency. Even one English product page in the middle of a translated journey is enough to lose the customer. If you've translated 80% of the funnel, finish the last 20%, those are the ones costing you sales at every step.

How to decide without overthinking, the A/B test

My recommendation, instead of debating for three weeks, is to test. Straight away.

Here's the method we proposed.

Take your 5 best-sellers. Produce a few creatives in the right local language on one side. On the other, keep untranslated English creatives. Run both over a set window, say a few weeks, long enough to get usable data. Then you decide based on results.

One thing you absolutely have to respect. Don't dump 15 creatives at once. If you push everything at the same time, you get no clean learning. Add your creatives gradually, and if one doesn't work, you cut it and bring in a new one. You run a rotation. That's how you isolate what actually works and what's dragging the account down.

And test your hooks on a small audience before putting budget behind them. It costs nothing to see which one actually lands before you scale. On social formats, everything plays out in the first few seconds and on the hook visible on mobile, so that's exactly what you need to validate first.

The goal is to come out of this test with one or two winner creatives per market. Once you've found your recipe, you can comfortably ramp up budgets on it.

The workload, we forget it every single time

Producing localised creatives for each country costs a lot of time. Really. You have to translate, adapt, sometimes redo visuals, make sure the wording fits the local culture.

And while you're doing that, you might have more urgent priorities. Like getting your best-seller creatives out for France before a promo period. You have to weigh all of that. The real question isn't just whether translating works better. It's whether you have the time to do it given your workload.

A number to put it in perspective. During a promo period, this brand produced 4 creatives last year versus 2 this year. Your capacity to produce creatives is limited. The clients who perform best with us are the ones with a strong capacity to produce content. So before spreading yourself across 5 countries, ask yourself whether you already have enough volume on your main market.

A creative bonus that has nothing to do with language

Since we're talking creatives, here's something I see too rarely exploited.

The same product photo shot from different angles can create real performance differences. Closer, farther, zoomed in, worn with other jewellery or on its own. From a single shoot, you often have 5-6 genuinely usable photos. Most brands use one or two and skip the rest.

Before even diving into translation, look at what you can already squeeze out of your existing visuals. There's often dormant volume sitting right there.

What I'd actually do

If I had to sum up how I make the call.

First thing, your site. Until it's 100% translated for the target market, with no stray English product pages, you don't even touch the creatives. You translate first. The rest, product and market, just nuances where you set the dial. Jewellery that's understood at a glance can run perfectly well in English on an expat market, a technical cosmetic on a non-English market, never. And when you're still unsure, you don't decide in a meeting, you launch a clean A/B test on your best-sellers, over a set window, and you let the numbers talk.

Honestly there's nothing to agonise over. You ask the question, you let the numbers decide, and you stop wasting hours in meetings.