Take the issue of translation cost savings. Often, this means dumbed-down content too devoid of any context or detail to be of any use in problem solving or task completion. Such content generally facilitates large scale “leveraging,” but it is in direct conflict with a basic of good design, that context always wins over consistency.

Or consider the controlled authoring debate as a sine qua non to facilitate the introduction of machine translation (MT). This can result in source material that is so dismal in terms of content, style and grammar that it generates support calls and does nothing for disaffected users. To heck with customer experience, though, as long as there are high fuzzy and perfect matches, and easily trained MT engines, right?

It doesn’t have to be this way. Recently, for example, the Oracle Worldwide Product Translation Group cooperated with the Oracle Applications User Experience group, facilitating mobile ethnography worldwide. The local market and cultural knowledge of the former team dovetailed with the latter’s UX science and engineering methodology to research mobile context of use in Europe. .. .