18 Jan Machine Translation and its impact on Search Engine Rankings

There have been great improvements in Machine Translation (MT). Many people have used tools such as Google’s translate when attempting to read text in a foreign language. The mobile version of this has also come in handy when we visit foreign countries. This is made possible by using advanced algorithms. The two best-known algorithms are the rule-based methodology and the probabilistic approach.

The rule-based method uses a list of quite comprehensive linguistic rules for the sentence structure of each language and another set of rules combine the two sentences together. Probabilistic MT uses statistical theory to link source and target text. For each sentence in the source language, there is any number of sentences in the target language with some being more likely than the others. Those with the most probability of being correct are then chosen.

Despite the advances that have been made in MT, the technology still falls behind the accuracy of human translation. This is because MT places too much focus on the words in a sentence than on the meaning behind the text. There is any number of examples where MT has led to humorous results. For companies and websites that have a reputation to uphold in a global market, these mistakes can be costly.

There is, however, an even more important consideration when dealing with online content. This comes down to the impact of MT on your search engine rankings. When it comes to optimizing your content for a foreign market, one has to focus not just on the correct translation but also on appropriately localizing it. Foreign markets have unique keywords that differ from a literal translation of the word. For example, “coche” is the Spanish term word for car. Yet, in some Latin American countries, it refers to a baby stroller. Hence, your “car” keyword strategy should be adjusted when selling in different countries.

It is clear, therefore, that MT is unlikely to help your SEO strategy in other languages and markets. An even more important consideration is whether using MT content in your website will harm your rankings. In fact, in the video below by Google’s Matt Cuts, he explains that search algorithms may interpret MT content as auto-generated content.

In fact, according to Google’s definition of “Automatically Generated Content”, “Text translated by an automated tool without human review or curation before publishing” is listed as an example. This implies that your content may be considered as spam by the search engines.

One can understand Google’s aversion to presenting MT content. Their search algorithms rely on high quality, well worded content. Similarly, Google emphasizes the importance of content that adds value to website users. Poorly translated text can negatively affect your visitor’s experience. Apart from harming your rankings, negative user experiences harms your brand and may drive potential users, customers and business partners away.