If a market matters, pay a translator.

Automatic translation keeps getting better. I still wouldn’t trust it with a market that matters to the business.

I’ve tried automated translations before.

I’ve also had several ecommerce websites professionally translated.

The professional version costs more.

It also gives me something automation doesn’t: confidence that somebody who actually understands the language has checked what the customer is being told.

Poor translation is expensive

A slightly awkward sentence is one thing.

A product specification, size, material, compatibility note or delivery promise translated badly is something else.

I’ve been caught by poor translations enough times that I don’t think the risk is worth it.

Especially internationally, where fixing the mistake can mean an expensive return, replacement or customer-service problem.

If the market is important enough to actively target, I’d rather pay a professional translator.

If it isn’t worth paying to translate properly, I’d seriously question whether it is worth pretending we’ve localised it at all.

English is often the smaller problem

For an English-speaking ecommerce brand, language is often one of the lower barriers to international trade.

A lot of customers can already understand enough English to shop.

Browsers such as Chrome can also translate pages for customers who want them translated.

That is not perfect either, but there is an important difference.

If the customer chooses to translate the page in their browser, they understand that a translation tool is sitting between them and the original content.

If I publish a translated product page myself, I’m presenting that wording as my own storefront content.

Translation doesn’t fix international ecommerce

You can have a perfect translation and still have a terrible international offer.

The customer still cares about price.

They still care about delivery.

They still care about how long it takes to arrive.

They still care about returns.

They still need useful product descriptions and a website that makes sense.

Letting somebody read a bad offer in their native language doesn’t make it a good offer.

They’re not going to suddenly forgive ridiculous delivery costs because the checkout button is translated perfectly.

They’re not going to ignore bad UX because the product title is in German.

Localisation has to earn its cost

Professional translation can absolutely work.

I’ve seen it work.

But the market has to justify the investment.

Prove the demand first.

Understand the margin.

Understand the delivery economics.

Work out whether customers actually want what you sell.

Then, if the market matters, localise it properly.

Shopify making automatic translation easier is useful technology.

It doesn’t change my rule.