The title of E.M. Foster’s most famous novel, A Room with a View, aptly describes what it was like for STP to attend the Welocalize LocLeaders 2014 Forum. Jesper (Sandberg) and I spent a day in a room with seasoned professionals representing both Welocalize’s clients and their vendors – a unique opportunity for us to hear what clients see as today’s localization challenges and to share a vendor’s perspective with them.
One key to success, in an industry that continues to be as labor intensive as localization, is attracting and managing human excellence. Welocalize puts a lot of emphasis on finding and keeping talent, as they understand that the greatest technology and processes in the hands of uninterested or unskilled people is, at best, a waste of investment and, at worst, a recipe for disaster.
In his opening session at LocLeaders, Welocalize CEO Smith Yewell explained that a good localization company applies different resource solutions depending on what are the most critical objectives in the localization project: time, cost or quality. Welocalize places linguists, typically very experienced ones, on-site with a client when it is essential that the translator’s time and effort must be fully dedicated to that one client. When there are high-volume bursts of work and the client also requires high-quality technical and linguistic solutions, Welocalize works with supplier companies, typically single language vendors. Regular freelance translators offer a cost-efficient service when that aspect of the project is of the essence. And crowdsourcing makes it possible to complete huge projects within time frames that in the past would have been a sheer impossibility.
The second session hosted by Antoine Rey, choosing the right quality model, demonstrated that not all content is created equal when it comes to the translation review process. One client shared an example of deliberately machine-translating user-generated content (UGC) on their website without any post-editing or quality checking because the sheer volume makes a review process utterly unfeasible. Others discussed the benefits and disadvantages of having their own in-country marketing department review the translations whilst others declared that having tried such a model, they preferred to use a more rigid quality metrics system that provides them with a structured analysis rather than a seemingly endless exchange of comments and opinions between the translator and the reviewer. Ultimately, what matters to clients is efficiency that shows on their bottom line and using their staff’s time in a meaningful way contributes to that. What matters to the vendor is that the quality metrics are defined in advance and that we receive a frame of reference (style guides, term bases and a clear brief) that enables us to do the job well.
The provocative afternoon topic of “what not to translate” spurred the audience on to brainstorm scenarios where language was not needed at all (from pictograms to IKEA assembly instructions), where the text type inverted the need to translate (motto and slogans) or where the target audience indicated they did not need a translation. The Nordic countries were mentioned as a region where the English language skills of the population – or at least the IT professionals and young people – were so good that translating into the local language was hardly necessary. As a Finn, I hasten to add that the well-known principle of people preferring to use their own language for important decisions (e.g. online purchasing) applies in Finland as much as elsewhere in the world. It is true that in countries boasting superior literacy rates and high expectations for language learning among their own citizens, the readers expect first-class linguistic content and have no tolerance for poor translation. However, these small, obscure language nations invest heavily in translation themselves and appreciate the same courtesy from those wanting to do business with them.
The last panel of the day discussed machine translation (MT) and how it is changing our industry. The use of MT impacts on the localization supply chain in a number of ways; translators working in production need to learn new skills when they become post-editors – it requires a new mind-set as well as methodological training in the difference between editing human output and synthetic engine output. For LSPs who operate their own engines, there is a lot of investment in technology as well as in designing optimal solutions for data cleaning and engine training. For the industry as a whole, there is the financial impact of MT to be considered. We need to find an equitable remuneration model that is acceptable to all parties in the supply chain in order to keep motivating everyone involved in the process.