Machine translation (MT) describes a software-based translation approach that translates content from one language to another.

With the rise of technology and an increasing demand for quick, cheaper translation, technological advancements have increased the overall usability of machine translation. For that reason, many companies are turning to machine translation and away from traditional translation, leaving many to question what might become of this industry and the language professionals who have built it.

Recent technological advances

Speech recognition systems

In recent years, speech recognition has made remarkable advances, especially in regard to machine translation. Using speech recognition, software learns from a body of recordings and the transcriptions created by humans. Conversational spoken phrases are instantly translated and repeated back in the target language, such as in the case of Skype Translator.

Neural Machine Translation (NMT)

Neural Machine Translation describes an approach in which a large neural network is trained using machine learning techniques. For example, Google developed its own artificial intelligence programs and applied them to its own multilingual neural machine translation system. It has been argued that these systems can, in some cases, produce output that’s “nearly indistinguishable” from humans. However, while the quality of machine translation has improved drastically over the years, humans can still add a lot of value.

…despite improvements in machine translation technology, humans can still add tremendous value to the translation process.

Challenges with machine translation

Word-sense disambiguation

One of the main linguistic issues with any machine translation system comes down to machines being unable to account for subtleties and nuance of language at the level humans can. While the quality has improved drastically, automated translations oftentimes result in unnatural phrasing, literal translations, and overall inaccuracies.

Negative impact on SEO

Search engine algorithms have changed quite a bit from their early days. Because search engines like Google and Bing actively search for bad content and spam, they can easily recognize computer-generated texts (especially if they’re translated by their own translation tools). As a result, machine translated content is penalized and ranked lower than similar content created using human translation.

Quality varies by language pair

Another problem with machine translation is that quality varies hugely depending on the languages involved. For example, MT systems can easily translate languages with similar syntactic structures—say German and English, or Italian and French. However, it becomes more difficult when the translation involves differing syntactic structures or languages that are implicit in nature, such as Turkish or Japanese.

Brief history of machine translation


Georgetown researchers
perform the first ever public
demonstration of an early MT


The Association for Machine
Translation and Computational
Linguistics is formed in the U.S.


National Academy of Sciences
forms a committee (ALPAC) to
study MT.


The French Textile Institute
begins to translate abstracts
using an MT system.


Systran begins to translate
technical manuals.


Trados is the first to develop and
market translation memory


The first commercial MT system
between Russian, English and
German-Ukrainian is developed
at Kharkov State University.


Systran and Babelfish offers free
translation of small texts on the


Google Translate’s statistical MT
system is launched.

Looking for human translation?

Human or machine translation?


Machine translation systems have evolved from being absolutely terrible to extremely sophisticated. And while machine translation may not be the solution for technical, legal and medical content, it certainly has its value and place. However, despite improvements in machine translation technology, humans can still add tremendous value to the translation process.

Human quality

Professional human translation or post-editing is a surefire way to ensure that you capture the true essence of a brand. Automated translators generally lack the inherent human voice, and don’t take into consideration your brand’s style and tone.

Low cost, high return

When it comes to high-volume business content such as product descriptions and user reviews, machine translation isn’t your only cost-friendly option. Crowdsourced human translation platforms like Gengo are both affordable and better for customer conversion. Human translation is the best option for maximizing the return on investment for your international efforts.

Simple and swift

Companies typically experience a conversion rate boost using human translation versus machine translation, as has been the case of many of our customers. Crowdsourced translation services can translate hundreds of product listings in less than 48 hours, with minimal effort. With a technology focused partner, translating high-volume content via API makes updating multilingual content seamless.

Trusted by global companies