Multilingual content and international SEO
Lost in translation? Multilingual SEO and our daily language bread
The world was blown away by the very first success of an AI. A computer won against the reigning world chess champion in 1997. People were stunned that software was able to overtake this ingenious mind. The fact is, programs are more efficient and usually even have a recipe for all eventualities up their sleeve. Chat GPT has recently astonished and panicked our society. But there are established artificial intelligences to which the international community has long entrusted its words without suspicion: translation software like Google Translate and DeepL. These AIs have been playing some unrecognized role in our lives for some time now. Actors in movies dictate texts into their smartphones, which the app then reads out to the antagonist in translation. This is regarded without astonishment as modern-day everyday communication. The apps can even recognize and translate images that you point your smartphone at. The text can be entered or typed in by voice. Incredibly, who needs the language course for a vacation in Costa Rica! Because even in countries with a rudimentary level of development, you can now travel without a dictionary. Apps like Google Translate can also be used offline. A journey through the digital language jungle of today!
How machines became translators, the birth of Translate Apps
It started with online encyclopedias, such as Leo.dict. In further development, the first Translation Management Systems (TMS) were released that could translate sentences word by word. Subsequently, the computer learned to adjust the sentence order and inflection of words according to rules defined in the program. This trend took off especially in the 1980s. It was hoped to achieve a certain level of quality with AI. To outrank human translators in the long run. Software development was based on the analysis of multilingual texts on the web and the manual maintenance of databases - which of course could hardly be complete. The project of translating entire text corpora into several languages and establishing them as, albeit bumpy, functioning aids to communication succeeded from the 2000s onwards. The pioneer on the digital market was, as so often, Google with Google Translate. This appeared on the global screen in 2006, as particularly fast and adaptable software. The new translation AI was born, far ahead of anything that had gone before in terms of time and cost - whether human or computer. Today there are a variety of free translation programs, such as DeepLwhich are continuously fed with new material and are placed at a relatively high level in terms of their linguistic accuracy. Thanks to this neural machine translation, it is now possible to translate relatively precisely. The software networks information in a similar way to the human brain. The whole thing is nothing more than mathematics in the cyber cosmos. Vocabulary and sentences are strung together according to certain systems.
Foreign language out-sourced: Can DeepL completely replace human translators?
If we take a closer look at the topic, the question arises as to whether the foreign language system can be applied error-free by apps without human intervention in the future? Is a human being even comparable with AI? As far as foreign language skills are concerned, a comparison is entirely inappropriate. When a baby learns its native language, it does so primarily on the basis of life experience and emotional attachment to its environment, with which it wants to interact. When learning a language, however, the AI does not have knowledge of the world. Its linguistic competence is based primarily on statistics and a rule system hard-wired into the software. Unfortunately, this also leads to the fact that the AI can quickly get its words wrong when translating texts. What homo sapiens like to formulate through the flower in order to empathically approach and emotionally pick up their counterparts is blatantly trumpeted here. Basically, the race for intelligence with programs is idle. As far as perfection and structural conformity in language are concerned, we will never be able to compete with AI. In terms of social intelligence, which expresses itself through language, however, we have better cards.
Nice program, bad output - how good is the translation AI really?
In 2020, the Product test foundation 15 popular translation apps. None of these free variants scored more than satisfactory. Even shorter sentences lacked grammar, and the translations were sometimes alienated from meaning. Especially in specialized industries, the translation tools fail miserably. Here, specialized knowledge is needed to translate the terms not only in a generalized way. The following also applies to multilingual homepages: a poorly translated call-to-action can also backfire! The same applies to product descriptions. If you want to sell to the international market but present the wrong details on the website, you won't win any new customers. Navigating companies through industry trends with multilingual content requires more than simple translation software. Especially when it comes to texts with a poetic touch, the program reaches its limits. A poem or a song lyric is unrecognizable after being translated by the machine. While some translation bugs are amusing, they are very counterproductive for customer communication: If, for example, you enter the gender specification "divers" in German when translating a homepage form, you are specified as a "diver" on the English page. Without human correction, this is a fatal error. Another example is the use of the term "public viewing" in the English-speaking world. This does not, as in German, refer to the public watching of events on big screens, but to "the laying out of a corpse at funerals." Ouch! So without prior cultural knowledge of people, you don't stand a chance with the translation app.
An unstoppable trend, multilingual SEO and content for an international audience
Circulating texts in multiple language versions across the Internet is an obvious trend of the digital age. Many German celebrities and influencers write and speak in English on their social channels in order to reach an international audience. In doing so, they show themselves vulnerable and expose their complete humanity in words and images. Linguistic errors are tolerated, because the human flaw is the reason why their followers absorb any form of content from them. The demand for multilingual content is increasing year after year. Bonds are especially established and maintained in our networked world when you can read info on the Internet in your native language. However, the English version of the content should not be considered the philosopher's stone! Especially those who want to conquer the market in southern countries like France, Spain and Italy cannot assume perfect English language skills. If users find information in their local language when they search, the probability of winning them over as new customers increases. For global online marketing, it is now essential to offer content in different languages. This is the only way for companies to climb up the Google rankings in other countries and to take their business beyond the German borders.
Multilingual Websites: SEO best practices when optimizing for different countries
If you want to improve your ranking on Google in other countries, you have to make sure that users find the right pages for their location. For international companies, multilingualism should be part of a quality content marketing strategy. To avoid being penalized by Google for duplicate content, your URL should include a language indicator. This can either be a subdomain (e.g. www.fr.content.com) or as subdirectory (e.g. www.coco-content-marketing.de/en/) happen. This allows international search engines and customers to recognize the language already through the URL. Apart from optimizing the URL parameters in the different languages, only one language should be chosen per page. A mixture will confuse readers and Google as well. An important part of your multilingual SEO is the translation of the metadata. Here you definitely need a person with very good language skills to match. Metadata is a digital quintessence of your website where you can't afford any misunderstandings. International keyword research must also be adapted to the language of the target audience. Those who implement these tips properly will ensure more organic traffic to their website from other countries. We as Content agency will gladly take over the implementation of these measures for you, as well as the high-quality translation of your content into the local languages.
Lifehack for the future: Pimp your languages!
Every day, people all over the world, from Beijing to Mexico City, willingly push their texts through the translation AI. Often without realizing what kind of gibberish is being spit out for the target group in the other language. In online marketing, communication in different languages has become a must in order to hold one's own in the globalized market. If you want to play the digital game worldwide, you will not be able to avoid translating your texts into different languages. Translation programs are good tools for this! One thing is certain, however: even the most complex software cannot and will never be able to laugh. Consequently, any app is at a loss when it comes to interpreting word wit, a certain linguistic style and cultural knowledge! The computer cannot read between the lines. This is especially true for translations with emotional messages. Linguistic content that grabs our attention or makes us smile requires a social life with knowledge of its social trophies and vulnerabilities. Intensive post-editing by humans will continue to be necessary in the long run!
