Close

Why is Google Translate (un)suitable for business communication?

Published in ADMA magazine, issue no. 48, November 2022, in the Jezikovni kotiček (Language Corner) column

Google Translate currently offers 133 languages, is free and convenient, and to a limited extent accessible without an internet connection. So are translators going to become obsolete? We will see below why, for companies that care about their reputation and want to look professional in foreign markets, machine translation should not be the first and only choice.

We all know Google Translate – the online machine translation service ‘par excellence’, used daily by more than 500 million users for translations into various languages. You can type your text, upload it as a document, or even enter the address of the website and – ‘voilà’ – in seconds your translation is ready.

If Google Translate was created 15 years ago with the aim of bridging linguistic barriers and making the world more accessible for everyone, today we are looking at a smart tool that has gone way beyond its original purpose. In addition to public use, machine translation tools have also found their role in business communication and in the translation of business materials. Machine translations: useful for a quick and economical translation, BUT…

Alongside Google, these kind of machine translation tools are also developed by other great corporations (e.g. Amazon, Microsoft, Meta – formerly Facebook, and others), as well as smaller niche companies, for private or commercial purposes; even the European Commission created its own machine translation tool. You may wonder why… Google Translate, for example, translates between 100 and 146 billion words every day, which is four times as much as all professional human translators put together would translate in a month. With the large quantities of text generated each day, machine translation services have therefore become the key to economical and financially effective translation and to understanding foreign texts.

Google regularly adds new languages to its service. In April 2021 it could handle 109 languages, but today 133 are available and counting. For each and every company this means there are virtually no languages that cannot be roughly translated and understood for the needs of their business, as the tool can even translate from or into Swahili, Uzbek or Esperanto. In addition to this, it’s free of charge, convenient and accessible practically anywhere, therefore many might wonder why we still need human translators at all. Have the machines replaced us in this industry too?

This issue is a hot topic and has been a pressing theme in translation circles for many years. Even the outside stakeholders observe that things are not as they used to be and ask themselves if translators and translation agencies will still be needed in five or ten years’ time. When such questions are raised, I respond we will be needed more than ever.

Let’s not pull the wool over our eyes – smart technology is constantly improving and changing our daily lives, including the way we work. In the case of machine translation tools, in the early stages of their development they performed best when translating shorter sentences; today, though, they can also process longer and more complex texts. However, in their raw form, machine translations do not replace those done by humans, and are not sufficient when you need high quality content. Therefore, when using them, especially for business purposes, we advise caution and following a well-thought-out process that should ideally involve an internal or external translation partner.

Google and similar machine translation services can undoubtedly help us communicate with everyday communication when abroad and with basic understanding of material and product information at an online store, etc. On the other hand, when doing business in foreign markets or designing a website in a foreign language, when preparing catalogues and similar promotional materials, and in all other cases where you have to make a professional impression and present yourself with appropriate, clear and attractive content, machine translation services should not be both your first and only choice. Let’s see why.

The quality of a machine translation also depends on the language that is being translated

The quality of a machine translation is highly dependent on the language selected. Machine translations that include English and other major languages are better than, for example, translations from Bulgarian to Slovenian. Namely, machine translators draw from existing published content, and undoubtedly more is written in English, Spanish and German compared to Bulgarian, for example – in the latter case, the translator’s “engine” is severely under-resourced and actually has little to go on for building your translation. Therefore, in such cases, translation is often done “via the English route”, meaning that the machine translator first translates your text from Bulgarian into English, and only then into Slovenian. Remember that the quality and accuracy of the translation can decline sharply in this process.

Style and terminology inconsistencies, a big weakness of machine translations

A machine translator does not work equally well in all genres of texts. They tend to work well for technical, medical, and other texts where there’s a lot of standardised content, and it can be surprisingly good for legal and administrative texts. Whenever your text requires a marketing tone, the use of specific terminology and a creative element, a machine translation tool will give poor results.

The translations are usually too literal and stylistically lacking. They often offer sentences that at first glance sound acceptable, but basically they are just an approximation of an accurate translation, both stylistically and grammatically.

Machine translations are also usually quite inconsistent – terminology, which may have a crucially important meaning, will be translated in different ways throughout a longer text. If there is a lot of terminology, the challenge is even greater. These shortcomings of machine translation are especially obvious when you follow a particular SEO strategy with specific keywords, or perhaps you are offering some kind of special service, product or technology that uses very specific terms in your industry.

Although a lot of work is done with the aim of an overall comprehension of a language, the machine translator still has difficulty understanding and translating conversational language, as well as in conveying the correct meaning or taking into account culturally specific aspects of the target language. For example, if you write a letter in Czech, the software doesn’t know that for business communication in the Czech Republic it is appropriate to address the person in question with the highest respect and with all their relevant titles. If your contact is Mr. Pokorný, who is an engineer, in Czech your letter should be addressed something like “Esteemed Mr. Engineer Pokorný”. Trust me that this way of addressing them will be much more effective than if our respected engineer were greeted in a completely ordinary way.

Conclusion: Machine translation tools may be helpful, but they should not be both our first and only choice

If you already use Google or another translation service for communication and business – no worries. Basically, you have done a lot in terms of making efficient use of your time, as you’ve translated the document in seconds, and completely free of charge, regardless of its size. However, if, as previously mentioned, you have more serious aspirations for your text, the least you could do is to trust a team of human translators to review these translations. These experts should make sure that your translation is not only grammatically correct, but also stylistically and culturally appropriate, using the proper terminology, according to the intended target audience. Not only will they improve it and put the finishing touches to it, often they will also decide whether the machine translation is usable at all. In the case of a poor machine translation, editing requires more time and effort than if the translator had translated the text from scratch. Instead of it being an effective use of time and money, such projects can drag on and, in the end, cost much more, so it is advisable to carefully decide at the outset which route to take.

Machine translations will undoubtedly become an inseparable part of the work process, but unfortunately they will sometimes let you down. Therefore, just like in other areas of business, this is a domain where you deserve to have a partner you can rely on, a translation expert in this case, who will help you when you’re considering using machine translations, and who will create high quality content that’s perfectly aligned with your wishes. If, on the other hand, you have a reliable translation team that knows how to listen to the specific needs of your business and its external business communication goals, you will surely put your message on the right track and continue to strengthen your market image. If you have already created a name for yourself in the domestic market and are now looking to expand overseas, remember that your precious products and services deserve nothing less than a truly excellent representation in a foreign language too. This is also an important reflection of your level of quality and credibility. So in this department, make sure you have the right strategy – with or without machine translation.

Sources:

CSA Research, Market research and analysis

ec.europa.eu, An official website of the European Union

Jason Wise (July 2022), How many people use Google Translate in 2022?, Earthweb, accessed from the web 1. 9. 2022

Konstantin Dranch from CustomMT

Multilingual Media LLC (May 2022), 24 languages added to Google Translate using zero-shot translation, accessed from the web 1. 9. 2022

Nimdzi Insights, Market research and international consulting

SEO Agency Twaino, Google Translate: The Complete Multilingual Translator’s Guide, accessed from the web 1. 9. 2022

Slator, Language Industry Intelligence