Treat translations like languages and your audience like humans

Isaline says: “Treating translations like languages means that it isn’t enough to copy the text you have written in your first language and directly translate it for another audience. I call this one-on-one translation and it doesn’t work, because when you translate text for websites, you want to capture the intention. You also want to achieve similar SEO, visibility, and conversion results. Most of the time, just direct translations will not help. You will need a bit of copywriting work on top of the translation.”

Is the optimum situation to get someone in the local country who understands the idioms, dialects, etc. to do that translation for you?

“Exactly. The ideal case scenario is that you have a localised copywriter or at least a native speaker of the target audience’s language. For instance, if you hire someone from Germany to do a Swiss German website, it will not work 100%. That is because the vocabulary used in Swiss Germany might be slightly different from that used by Germans based in Germany. Of course, that depends on the subject too.

When I say treat your audience like people, it’s the same. It means people across different countries are quite different. With language comes a universe of references and concepts that vary. That becomes quite obvious when you translate expressions like ‘white as snow’, which is an expression that you typically can’t directly translate if you use it on your websites.

Therefore, you have to really think about the impression you want to give and what you are trying to achieve with your audience. Then you can deliver something with a similar impact in the country. Adapt and localise what you are trying to say for the people in the target country. Doing that means that texts also don’t need to be exactly the same. Even when you have two articles, they don’t have to be identical. Sometimes you might have a different definition or additional explanations - maybe because the people in that country don’t understand a concept that is usually known among those that speak your first language.

This applies to expressions at the level of words and concepts at the level of ideas. That’s why translation is often not enough on a website. For some articles, they can be, but sometimes you need to go further.

If you’re speaking about an institution, for instance, do you have a similar institution in the target country that people can relate to? It’s about the words, but it’s also about the ideas you are conveying, what you’re referring to, the links you use, etc.”

Should you just focus on your higher converting and profitable pages for the type of translation that you’re advocating? What about the other pages on your site?

“Realistically, a company will not have the same budget for every language. You might have your first language and realise you cannot invest the same amount of money for the second language - or at least not initially. A good plan is to start by listing all of the content. By all of the content, I also mean the architecture, the buttons, etc.

My recommendation here would be to focus not on a type of page but on a type of service you want to sell in the given country. For instance, in eCommerce, instead of translating all the products, pages, etc. - choose a category of products that is very relevant to Swiss German people. Then, there might be another category that is relevant for UK people.

The idea is to focus on a domain of expertise of service or category of projects and see how that works before you touch on other aspects of the site. Direct translation could be enough for elements that are not so important for the target group or where competition is still too high. You would not need to invest heavily everywhere.”

Is there ever a place for automated translation?

“There can be. It depends on the amount of content you have to translate for your website. In that case, you can start with the less important parts - the types of projects on your websites that might not work as well in the target country - and start with an automatic translation there. However, you should have a plan to optimise it.

Very importantly, if you are only the SEO or project manager, you should clarify to the higher-level managers that this will not convert well. That will help manage the expectations of those who say ‘no’ or call for you to translate everything. You will need to explain to them the various levels of translation that can be achieved, and their corresponding levels of profitability.

Basically, you really can’t do everything since it is expensive. Manage expectations and have a plan to do everything at the right time eventually. As you execute the plan, you should have your KPIs set. Then, examine the elements you already localised with translation and find out how well it works.

You will have the KPIs from traffic, from Search Console, and the rankings of keywords. However, also get some internal feedback from places like customer support. The customer support team are your friend in the company. Their feedback is important because they might be receiving lots of questions from a country. What are these questions? Find out. This is not something that can be seen as quantitative data. Nevertheless, you need to gather this internal feedback and know what type of questions they are always answering.”

How do you ensure that you’re covering the right keywords in another language if you don’t speak the language? Is it necessary to employ a local SEO on top of a local writer in a country?

“I would start with translating the keyword research with automation. Get an export from a tool such as Ahref and Semrush, and then maybe choose the highest-ranked keywords. Afterwards, you can have the ones you plan to use checked by the copywriter. That would be the cheapest alternative, if you can’t have a local SEO doing the keyword search.

In an ideal scenario, you have your local SEO who does the keyword research along with the copywriters - just like you do with the first language. That way, it is not just a copy and paste of the first language’s content. However, this is expensive. You can do the first part yourself with the tools you have at your disposal, and then only check the essential keywords you plan to use when you do your URL mapping.”

What shouldn’t SEOs be doing in 2023? What is seductive in terms of time, but ultimately counterproductive?

“The first thing is to be careful with the type of work the client is asking you to do. For instance, if the client asks you to translate the keyword search. Translation work is very different from idea gathering and data management of keywords. You should be mindful of settling for translation or, rather, you should it with the client. Do they want you to translate the words and quality check them, or find better alternatives involving idea gathering, keyword multiplier, etc.?

Generally, do not just settle for work that sounds like translation because, often, clients expect cheap translations with big SEO results. The SEO process is different from direct translations so you must be careful with that. One of the risks I have encountered is that you translate the keyword search and check the words. Then, the client checks on their tool, discovers that there is an alternative and starts to doubt your work. I often have to explain that they asked me to translate the keyword search, which is different from the SEO process of idea gathering.

Basically, be careful with the client’s wording and be sure of what they are asking for so that you deliver the results that they are expecting. Your keyword search should be two different documents, and then you can map the translations.”

Isaline Muelhauser is an SEO Consultant at Pilea.ch and you can find her at pilea.ch.