This has far-reaching implications for site owners. What should you bear in mind when translating in 2014?
Cynical as it seems, executives should ask themselves whether they want a translation in order to grow their markets, or to give the appearance of a global service or offering to impress a domestic market. (Our experience of translation projects across the globe suggests that this is not at all far-fetched.)
In the latter case, be careful - something whipped up for internal consumption need not be that great, but poorly translated public-facing websites put readers off.
Translation implies much more than the ability to use a dictionary correctly. Each new language version of a website has the same content production cycle components as the original text and updates, and moreover introduces additional technical complexities.
As such, translation requires project, technology and market experience in the regions and industries that the website targets.
Local knowledge is a supreme differentiator. Translating a US website into German isn't as simple as paying someone on your staff.
Depending on the market you're trying to reach (Namibia has different numbering conventions from Germany), your audience (affluent Germans speak a more formal German than youth), and a range of other factors, translation can be a tricky proposition.
Given the inherent complexity of translation, companies must accept that they will have to expend very nearly the same amount of resources on each new language project - more if they undertake multiple concurrent projects.
For example, each new translation text must align with marketing strategy and also integrate with the original (to stay current with subsequent updates).
Machine translation (MT) services like Google Translate are a double-edged sword. On the one hand they can help companies cope with the sheer volume of content on the web needing translation, but on the other they can obstruct the reader's understanding.
Human translators, in turn, are expensive and their efforts do not scale. But translation services that use technology to aid human translators with translation memory databases, software integration skills and automation of certain tasks can greatly speed up and scale translation, and deliver outcomes of the highest quality.
Cloud-based ('crowd-sourced') translation services tend to be heavily funded by venture capital of late, but the value proposition of these services is questionable.
Many promise a painless experience, when translation is anything but simple, and to approach it as such would put the outcome at risk. When the hosted translation service goes down, a service level agreement will not magically meet your customers' expectations of on-time availability of new-language versions of your website. Cloud services also tend to throw any available resource at isolated portions of content, which disregards the best practice of conducting translations within their proper context.
What about the proliferation in social media content? Companies appear loath to translate social media participation by consumers, and even the Facebook and Twitter accounts under their control. In many instances, they view this as inviting complexity (having to take on more content projects and customer service capacity). As long as their domestic markets outweigh others in importance, this will not change.
Another reason is legacy marketing attitudes. Many companies want to control their social presence in the way they are used to controlling advertising. But if you want to grow your markets, you may want to give your consumers in new markets as much say as you do those in foundation markets.
Companies are fixated on significant differences in the quoted cost of a translation, when they should factor in the cost of fixing poor translation outcomes or inefficient processes. These could entail re-translating, approving and loading each language version separately every time the original changes.
But if you use a language service provider that employs computer-aided techniques, changes will propagate automatically across language versions, because of language database integration and automation. Updates further require minimal project time, because the provider's translation memory database already contains much of the core client text.
For companies with market-focused expansion plans, translation cannot be a throwaway effort. Our experience is that someone just entering into this realm may be content to use MT, crowd-sourcing or someone on their staff, but a more sophisticated understanding will lead them to professional translation services aided by technology.
