So how does your organization categorize the relative value of its content for translation? How, what and when do you translate specific types of content? I propose we look at this in terms of a “Content Value Index.” Content can be divided into three levels to help distinguish the relative value and the appropriate method to translate it. These levels are automatic, community and professional.
The “automatic” level consists of content that is less important. In this level content such as comments, tweets and some blog and forum posts would be automatically translated. With such content just the “gist” of what it’s saying is good enough.
For “automatic” content, machine translation and translation memories are perfect. Machine translation technology provides an instant computer translation. You’ve likely seen or used Google’s translation tool, which is a great example of this technology. The quality of these translations can be inconsistent: sometimes they are remarkably accurate and other times they don’t make a lot of sense, but you can almost always get a basic understanding of what the content is saying. Translation memories are previously translated content that you reuse, recycle and repurpose and can be applied automatically to content, the same as machine translation.
The next level, “community,” represents content that is important to readers but doesn’t require a professional translator. A large bulk of the content on many sites falls into this category. Examples include many support pages, blog posts, some popular forum posts and comments, and wiki content.
Content in this level is an ideal candidate for community translation. Using cloud-based technology bi-lingual readers can volunteer to translate this content in real-time without leaving the site. Organizations can use community volunteers, crowdsourcing, partners or internal employees to help with the translation of content.
The final level, “professional,” is content that is very important to the organization. It may consist of product/service information such as features, benefits, and pricing. It may also include pillar blog posts including your most popular or visited content.
Because this content is so central to your site, it may make sense to have a professional translate this content. The reason I say “may” is that some studies have shown crowd-based translations can be superior to professional translations. This is because volunteers often understand the content matter and related colloquialisms even better than a professional might. In any case, cloud-based translation technologies enhance localization projects even when using professionals.
Mix for Best Results
It is possible to mix these translation workflows to create the best approach. One could have the content machine translated with translation memories and then ask its community to do a review of the translation and then have a professional translator review, modify, and approve the translations. This presents a much less time-consuming and expensive option but results in the same level of quality.
Not all content is created equal. The mindset of “all or nothing” pervades and prevents many localization projects from launching. As the translation industry starts to embrace cloud-based solutions that leverage a hybrid translation management approach, new opportunities become available for small and large companies alike to vary their approach and use the right workflow for the job to translate their content in an efficient, cost-effective manner.
Editor's Note: To read more by Rob Vandenberg: