How to translate…Decisions, decisions, decisions….
Part 3 – Expanding your translation channels and leveraging the crowd
Content repurposed courtesy of Lingotek, www.lingotek.com
In part 2 of this 5-part series we established that all content is not created equal and that you should match the translation method to the value of the content. In part 3 we will elaborate on the various channels available for translating content and how to leverage the crowd to lower the cost and increase the speed of your translations.
For low-value content, the ideal translation channel is to use a computer translation, also known as a Machine Translation (MT). The pros are that MTs are instant and low cost. The cons are that MTs aren’t always perfectly accurate and can mis-translate certain specialized words and phrases. As the Content Value Index (CVI) indicates, a Machine Translation gives you a gist of the content which is usually sufficient for the reader. If the quality of a Machine Translation is insufficient and the content being translated seems to be valuable enough it may be bumped into a higher level of the CVI.
Content deemed to be of medium value can ideally be translated by the community or crowd. Bi-lingual users may volunteer to translate certain key pieces of content themselves. With the right technologies they can be empowered to translate instantly and directly inside the website itself.
There’s a natural self-selecting effect created as your content becomes attractive to people speaking other languages. For example, if your content is attractive and relevant to a Spanish-speaking audience, you will likely attract bi-lingual’s to your content who speak Spanish and are willing to translate it for the others interested in consuming that content. On the other hand, if content isn’t relevant or interesting for people speaking German, there will be a lack of bi-lingual volunteers, but there also will be no demand to consume the content in German.
High-value content is often best translated by professionals. We discussed in part 1 that traditional translating methods are a very poor fit when translating social websites that create a constant stream of new content. There are two ways you can overcome this problem. First, by categorizing your content into low-, medium-, and high-value buckets you can greatly reduce the amount of content considered high-value and justifying a traditional translation. Second, by using the right translation platform – Lingotek’s Collaborative Translation Platform (CTP) is one example – you can have your traditional translators get away from emailing Word documents back and forth and have them translate content directly in your website with a number of powerful translation tools at their disposal. The right translation platform makes it easy to monitor the progress and ensure the quality of translations regardless of who is translating.
In fact, by using the right translation technologies you may even be able to leverage your crowd to translate most or all of your high-value content while maintaining a degree of quality, accuracy and consistency equal to traditional translators. That’s what we’ll address in part 4 so stay tuned.